This document presents a vision-based algorithm to localize and segment nutrition fact tables (NFTs) on product packages using smartphone cameras. The algorithm first localizes the NFT using vertical and horizontal projections to detect boundaries. It then segments the localized NFT into single- or multi-line text chunks. The algorithm has been implemented on Android platforms. Initial experiments localizing and segmenting NFTs are discussed. The goal is to extract nutrition information using computer vision to help with nutrition management applications.
Usability analysis of sms alert system for immunization in the context of ban...eSAT Journals
Abstract Both the market and academia strongly encourage the development of usable systems, and they do so by relying on a number of standards, guide-lines, research and good practice streams. Unfortunately, in the health sector, whilst being the owner of standards under many purposes and topics, seems still falling and running behind as the conceptual issues and practical implications of usability are concerned. In this study, it was found that rapid growth of mobile applications through SMS increases in a significant way in developing countries particularly in Bangladesh. Public satisfaction was highly shown in mobile health services through SMS. In our paper, usability has been analytically investigated throughout a simulated health oriented action setting and against a prototype of SMS based health services in Bangladesh, and several provoking conclusions in terms of “rethinking usability” applied to academic actions and decision making have been derived. Various health institutes can be influenced by this study to challenge existing difficulties against usability potential. Keywords: ICT, mHealth, Mobile applications, SMS, Usability
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Electronic prescribing or E-Prescribing (e-Rx) is electronic transmission of prescriptions from physician to pharmacists using computer and other mobile devices, such as cell phone and tablets. E-Prescribing system has replaced the phone, paper and fax based method of prescription. The system improves patient safety by reducing prescribing errors due to various reasons, such as illegible handwriting and ambiguous abbreviations. It also reduces healthcare costs over the paper based prescription systems. It permits the physician and other healthcare professionals to regenerate a new prescription, when any prescription error occurs during pharmacy operation.
This study was conducted to provide a possible solution to the unceasing problem of finding willing blood donors with the use of mobile technology and location based services. The blood donor finder application aimed to provide a means to easily locate a willing blood donor through a mobile application. Using the mixed method of conducting a research, the researchers were able to develop the application. Two groups of randomly chosen respondents participated in this study to assess the technical quality and quality of using the application. This study utilized and followed the Spiral Model with the following stages Planning, Risk Analysis, Engineering, and Evaluation. After the development of the project, the application was subjected to the assessment of Information Technology IT professionals and residents in one of the municipality in the province of Nueva Ecija, Philippines. The assessment on the technical qualities were based on the following criteria Usability, Effectiveness, Efficiency, Accessibility, and Assistive Technology. On the other hand, the assessment on the quality of using the application was based from the following criteria Perception of the respondents on the usefulness of the application, perception of the respondents on the ease of using the application, and the intentions of the respondents in using the application. In general, the application was successfully developed following the stages of the Spiral Model and the application passed the assessment made by the IT professionals and the residents with some suggested enhancements and improvements for the betterment of the application. The results of this study proved the possibility of having a blood donor finder application and such application can aid in locating a possible willing blood donors. Cris Norman P. Olipas | Elizor M. Villanueva "Dug-Uhay: A Blood Donor Finder Application" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-1 , December 2019, URL: https://www.ijtsrd.com/papers/ijtsrd29678.pdf Paper URL: https://www.ijtsrd.com/computer-science/other/29678/dug-uhay-a-blood-donor-finder-application/cris-norman-p-olipas
Usability analysis of sms alert system for immunization in the context of ban...eSAT Journals
Abstract Both the market and academia strongly encourage the development of usable systems, and they do so by relying on a number of standards, guide-lines, research and good practice streams. Unfortunately, in the health sector, whilst being the owner of standards under many purposes and topics, seems still falling and running behind as the conceptual issues and practical implications of usability are concerned. In this study, it was found that rapid growth of mobile applications through SMS increases in a significant way in developing countries particularly in Bangladesh. Public satisfaction was highly shown in mobile health services through SMS. In our paper, usability has been analytically investigated throughout a simulated health oriented action setting and against a prototype of SMS based health services in Bangladesh, and several provoking conclusions in terms of “rethinking usability” applied to academic actions and decision making have been derived. Various health institutes can be influenced by this study to challenge existing difficulties against usability potential. Keywords: ICT, mHealth, Mobile applications, SMS, Usability
IJRET : International Journal of Research in Engineering and Technology is an international peer reviewed, online journal published by eSAT Publishing House for the enhancement of research in various disciplines of Engineering and Technology. The aim and scope of the journal is to provide an academic medium and an important reference for the advancement and dissemination of research results that support high-level learning, teaching and research in the fields of Engineering and Technology. We bring together Scientists, Academician, Field Engineers, Scholars and Students of related fields of Engineering and Technology
Electronic prescribing or E-Prescribing (e-Rx) is electronic transmission of prescriptions from physician to pharmacists using computer and other mobile devices, such as cell phone and tablets. E-Prescribing system has replaced the phone, paper and fax based method of prescription. The system improves patient safety by reducing prescribing errors due to various reasons, such as illegible handwriting and ambiguous abbreviations. It also reduces healthcare costs over the paper based prescription systems. It permits the physician and other healthcare professionals to regenerate a new prescription, when any prescription error occurs during pharmacy operation.
This study was conducted to provide a possible solution to the unceasing problem of finding willing blood donors with the use of mobile technology and location based services. The blood donor finder application aimed to provide a means to easily locate a willing blood donor through a mobile application. Using the mixed method of conducting a research, the researchers were able to develop the application. Two groups of randomly chosen respondents participated in this study to assess the technical quality and quality of using the application. This study utilized and followed the Spiral Model with the following stages Planning, Risk Analysis, Engineering, and Evaluation. After the development of the project, the application was subjected to the assessment of Information Technology IT professionals and residents in one of the municipality in the province of Nueva Ecija, Philippines. The assessment on the technical qualities were based on the following criteria Usability, Effectiveness, Efficiency, Accessibility, and Assistive Technology. On the other hand, the assessment on the quality of using the application was based from the following criteria Perception of the respondents on the usefulness of the application, perception of the respondents on the ease of using the application, and the intentions of the respondents in using the application. In general, the application was successfully developed following the stages of the Spiral Model and the application passed the assessment made by the IT professionals and the residents with some suggested enhancements and improvements for the betterment of the application. The results of this study proved the possibility of having a blood donor finder application and such application can aid in locating a possible willing blood donors. Cris Norman P. Olipas | Elizor M. Villanueva "Dug-Uhay: A Blood Donor Finder Application" Published in International Journal of Trend in Scientific Research and Development (ijtsrd), ISSN: 2456-6470, Volume-4 | Issue-1 , December 2019, URL: https://www.ijtsrd.com/papers/ijtsrd29678.pdf Paper URL: https://www.ijtsrd.com/computer-science/other/29678/dug-uhay-a-blood-donor-finder-application/cris-norman-p-olipas
Managing Binge Eating Disorder with iTakeControliTakeControl
Mobile health applications are still in their infancy but present a great opportunity for the healthcare industry. They enable patients to have a greater role in their health, empower providers to make data-driven decisions, allow researchers to gain greater insight into patient and disease populations, and give payers a new window into how patients are doing on treatment.
In this paper, we dive into iTakeControl’s Binge application and look into some of the data we have gathered so far.
March 19, 2011 presentation at the Annual conference for the Association for Prevention Teaching and Research on opportunities for students to be engaged with mHealth.
The explosion in the number of applications (apps) designed for the medical and wellness sectors has been noted by many. Recently we have seen increased presence of truly medical apps, in addition to consumer health and wellbeing apps, designed for clinical professionals and patients with medical conditions.
Consumer based mHealth apps typically allow people to do old things in new ways, such as recording health measures digitally rather than on paper. We see this also with medical apps, where increases in the quality and efficiency of existing health care models provide clinical staff with digital tools that replace paper based documentation. In rare and exciting cases we are also seeing mHealth applications that are doing things in entirely new ways to drive real innovation in health care delivery through mobile devices.
The aim of the tutorial is to highlight real world, high impact mobile research that is relevant to the key discipline of Mobile HCI. Thus, the tutorial will be application rather than academically focused. The tutorial will highlight the wide range of mHealth applications available that go far beyond trackers and behavior change tools and encourage researchers to look beyond consumer applications in their research. Four key areas of mHealth applications will be covered including Apps for the HealthyWell, mHealth in Hospitals, Practice and Clinical Apps and Patient Apps and will cover applications for health assessment, treatment and triage, behavior change, chronic illness, mental health, adolescent health, rehabilitation and age care with a focus on the need for rigorous evaluation and efficacy analysis.
A Mobile-Cloud based Context-Aware and Interactive Framework for Diabetes Man...IJERA Editor
One of the biggest preoccupations of any healthcare provider is trying to eliminate the mistakes during treatment. Using Cloud computing permits to host all information in one place and make it accessible anywhere, anytime, and any channel, especially when it comes to the disease diabetes mellitus. Diabetes mellitus is a group of diseases characterized by an elevated blood glucose level (hyperglycemia) resulting from defects in insulin secretion, in insulin action, or both. It is, today, the most challenging syndrome in the world. In the latest survey, the world’s 65% of the population is suffering from either Type 1 or Type 2 diabetes mellitus. The patient’s blood glucose level is not the same 24x7 hours in most of the cases and take medication 24x7 hours is impossible. Cloud Computing is so the best solution to check in the patient’s blood glucose control and try to balance it, especially at remote areas where healthcare services aren't easily available.
Android Based Application to Ensure Medical Adherence: CareWiseDr. Amarjeet Singh
In this fast-paced world, it is difficult to balance one’s domestic and professional life. Often, we have seen our grandparents forget things that are a part of their routine for like their medication. Many of them need to take their medicines at a fixed time, due to their age they often forget their scheduled medicines which can sometimes have consequences on their health. Therefore, we feel that there is need for an application to help our fellow senior citizens with their medication by the usage of image and a general description. By doing so we feel that this application can act as a helping hand to better monitor their health. This application will have features which will help them sort out medicines based on their name, image and description. A scheduled calendar will help them plan their medicine more efficiently. Reminders ensure that they don’t forget the medications. This application will provide a very easy to use interface that the elderly can easily navigate through without any qualms.
DANES: Diet and Nutrition Expert System for Meal Management and Nutrition Cou...rahulmonikasharma
“Your body is your temple” As people across the globe are becoming more health conscious, eating more healthy food and avoiding junk food, a system that can measure calories and nutrition in every day meals can be very useful for maintaining one’s health. Food calorie and nutrition measurement system is very beneficial for dieticians and patients to measure and manage their daily food intake. We also know that it’s difficult to find an affordable nutritionist or a dietician across the street; therefore, we have proposed a system – DIET AND NUTRITION EXPERT SYSTEM. The proposed system is a responsive android application which contains the knowledge and data regarding the fitness of a person and nutrition content values. This application consists of the user interface which will be publicly displayed on the application i.e. the basic information regarding the fitness and nutrition values such as how to maintain good health by adapting healthy eating habits which includes the intake of calories, proteins and carbohydrates etc. in proper proportion. A dietician consults a person based on his schedule, body type, height and weight. The system too asks all this data from the user and processes it. It asks about how many hours the user works, his height, weight, age etc. The system stores and processes this data and then calculates the nutrient value needed to fill up users’ needs.
Managing Binge Eating Disorder with iTakeControliTakeControl
Mobile health applications are still in their infancy but present a great opportunity for the healthcare industry. They enable patients to have a greater role in their health, empower providers to make data-driven decisions, allow researchers to gain greater insight into patient and disease populations, and give payers a new window into how patients are doing on treatment.
In this paper, we dive into iTakeControl’s Binge application and look into some of the data we have gathered so far.
March 19, 2011 presentation at the Annual conference for the Association for Prevention Teaching and Research on opportunities for students to be engaged with mHealth.
The explosion in the number of applications (apps) designed for the medical and wellness sectors has been noted by many. Recently we have seen increased presence of truly medical apps, in addition to consumer health and wellbeing apps, designed for clinical professionals and patients with medical conditions.
Consumer based mHealth apps typically allow people to do old things in new ways, such as recording health measures digitally rather than on paper. We see this also with medical apps, where increases in the quality and efficiency of existing health care models provide clinical staff with digital tools that replace paper based documentation. In rare and exciting cases we are also seeing mHealth applications that are doing things in entirely new ways to drive real innovation in health care delivery through mobile devices.
The aim of the tutorial is to highlight real world, high impact mobile research that is relevant to the key discipline of Mobile HCI. Thus, the tutorial will be application rather than academically focused. The tutorial will highlight the wide range of mHealth applications available that go far beyond trackers and behavior change tools and encourage researchers to look beyond consumer applications in their research. Four key areas of mHealth applications will be covered including Apps for the HealthyWell, mHealth in Hospitals, Practice and Clinical Apps and Patient Apps and will cover applications for health assessment, treatment and triage, behavior change, chronic illness, mental health, adolescent health, rehabilitation and age care with a focus on the need for rigorous evaluation and efficacy analysis.
A Mobile-Cloud based Context-Aware and Interactive Framework for Diabetes Man...IJERA Editor
One of the biggest preoccupations of any healthcare provider is trying to eliminate the mistakes during treatment. Using Cloud computing permits to host all information in one place and make it accessible anywhere, anytime, and any channel, especially when it comes to the disease diabetes mellitus. Diabetes mellitus is a group of diseases characterized by an elevated blood glucose level (hyperglycemia) resulting from defects in insulin secretion, in insulin action, or both. It is, today, the most challenging syndrome in the world. In the latest survey, the world’s 65% of the population is suffering from either Type 1 or Type 2 diabetes mellitus. The patient’s blood glucose level is not the same 24x7 hours in most of the cases and take medication 24x7 hours is impossible. Cloud Computing is so the best solution to check in the patient’s blood glucose control and try to balance it, especially at remote areas where healthcare services aren't easily available.
Android Based Application to Ensure Medical Adherence: CareWiseDr. Amarjeet Singh
In this fast-paced world, it is difficult to balance one’s domestic and professional life. Often, we have seen our grandparents forget things that are a part of their routine for like their medication. Many of them need to take their medicines at a fixed time, due to their age they often forget their scheduled medicines which can sometimes have consequences on their health. Therefore, we feel that there is need for an application to help our fellow senior citizens with their medication by the usage of image and a general description. By doing so we feel that this application can act as a helping hand to better monitor their health. This application will have features which will help them sort out medicines based on their name, image and description. A scheduled calendar will help them plan their medicine more efficiently. Reminders ensure that they don’t forget the medications. This application will provide a very easy to use interface that the elderly can easily navigate through without any qualms.
DANES: Diet and Nutrition Expert System for Meal Management and Nutrition Cou...rahulmonikasharma
“Your body is your temple” As people across the globe are becoming more health conscious, eating more healthy food and avoiding junk food, a system that can measure calories and nutrition in every day meals can be very useful for maintaining one’s health. Food calorie and nutrition measurement system is very beneficial for dieticians and patients to measure and manage their daily food intake. We also know that it’s difficult to find an affordable nutritionist or a dietician across the street; therefore, we have proposed a system – DIET AND NUTRITION EXPERT SYSTEM. The proposed system is a responsive android application which contains the knowledge and data regarding the fitness of a person and nutrition content values. This application consists of the user interface which will be publicly displayed on the application i.e. the basic information regarding the fitness and nutrition values such as how to maintain good health by adapting healthy eating habits which includes the intake of calories, proteins and carbohydrates etc. in proper proportion. A dietician consults a person based on his schedule, body type, height and weight. The system too asks all this data from the user and processes it. It asks about how many hours the user works, his height, weight, age etc. The system stores and processes this data and then calculates the nutrient value needed to fill up users’ needs.
POST EACH DISCUSSION SEPARATELYThe way patient data is harvested.docxLacieKlineeb
POST EACH DISCUSSION SEPARATELY
The way patient data is harvested and used is rapidly changing. Patient data itself has become quite complex.
In the future
, patient data will be combined with financial data, product or drug data, socioeconomic factors, social patterns, and social determinants of health. Cognitive behavior and artificial intelligence will be applied to the data to help prevent and depict rather than cure disease.
Evaluate the future of Healthcare information technology.
Include the following aspects in the discussion:
Find two articles related to the future of information systems (IS) in healthcare
Include telehealth, wearable technology, patient portals, and data utilization
Analyze potential benefits from advances
Discuss, from your own perspective, the advantages and disadvantages of having a system where the patient manages their own data
REPLY TO MY CLASSMATE’S DISCUSSION TO THE ABOVE QUESTIONS AND EXPLAIN WHY YOU AGREE. MINIMUM OF 150 WORDS EACH
Classmate’s Discussion 1
The technological advancements that have occurred in the field of healthcare have greatly changed the way people view and interact with the healthcare system. They have also led to the reduction of costs and the increasing efficiency of the system. We expect that the future of healthcare will continue to be influenced by information technology.
Due to the technological advancements that have occurred in the field of healthcare, physicians are now able to spend less time with their patients. This has allowed them to provide more effective and efficient care to their patients. In the future, we can expect that the increasing number of specialists who can delegate their work to other doctors will have a significant impact on the healthcare system. The increasing efficiency of doctors is expected to have a significant impact on the shortage of specialist physicians in the future. This issue could be solved using technology. Hopefully, the use of information technology can help boost the number of specialist physicians (Patric, 2022).
Electronic health records have revolutionized the way healthcare is done. Despite the progress that has been made in terms of keeping and tracking these records, they are still not widely used yet. This means that the kind of growth that was expected from the adoption of these records has not materialized. Although the adoption of electronic health records has been made in various parts of the world, it’s still not widely used in all areas. This means that the ability to keep track of one’s medical history is still very important (Patric, 2022).
The increasing importance of information technology in healthcare has led to the prediction that the cost of healthcare will eventually come down. Various factors such as better accessibility and efficiency will help make healthcare more affordable and more effective.
It’s widely believed that keeping one's health is much cheaper and easier than treating a.
The future of healthcare: when mobile disappearsMatteo Penzo
In today’s digital world, mobile devices are the powerful bridges between a connected ecosystem of healthcare professionals, caregivers and patients. New developments in big data, wearable sensors and the application of social layers are shifting an industry that used to focus on curing diseases to one that emphasizes health and wellness. But the mass adoption of connected healthcare will only happen when solutions are designed to be intuitive and technologies are forgotten. The future of healthcare will happen when mobile disappears into the background, placing the patient in the center and in control of their lives.
Ligia Alexandra Gaspar - bachelor thesis in collaboration with OAMK Digital Patient project I developed a prototype of a health application that contains 3D human body modeling and is inspired from doctor Peter D'Adamo's work on nutrition and blood types
Similar to Vision-Based Localization & Text Chunking of Nutrition Fact Tables on Android Smartphones (20)
Digitizing Buzzing Signals into A440 Piano Note Sequences and Estimating Fora...Vladimir Kulyukin
Digitizing Buzzing Signals into A440 Piano Note Sequences and Estimating Forager Traffic Levels from Images in Solar-Powered, Electronic Beehive Monitoring
Many problems in information retrieval and related fields depend on a reliable measure of the distance or similarity between objects that, most frequently, are represented
as vectors. This paper considers vectors of bits. Such data structures implement entities as diverse as bitmaps that indicate the occurrences of terms and bitstrings indicating the presence
of edges in images. For such applications, a popular distance measure is the Hamming distance. The value of the Hamming distance for information retrieval applications is limited by the
fact that it counts only exact matches, whereas in information retrieval, corresponding bits that are close by can still be considered to be almost identical. We define a "Generalized
Hamming distance" that extends the Hamming concept to give partial credit for near misses, and suggest a dynamic programming algorithm that permits it to be computed efficiently.
We envision many uses for such a measure. In this paper we define and prove some basic properties of the :Generalized Hamming distance," and illustrate its use in the area of object
recognition. We evaluate our implementation in a series of experiments, using autonomous robots to test the measure's effectiveness in relating similar bitstrings.
Adapting Measures of Clumping Strength to Assess Term-Term SimilarityVladimir Kulyukin
Automated information retrieval relies heavily on statistical regularities that emerge as terms are deposited to produce text. This paper examines statistical patterns expected of a pair of
terms that are semantically related to each other. Guided by a conceptualization of the text generation process, we derive measures of how tightly two terms are semantically associated.
Our main objective is to probe whether such measures yields reasonable results. Specifically, we examine how the tendency of a content bearing term to clump, as quantified by previously
developed measures of term clumping, is influenced by the presence of other terms. This
approach allows us to present a toolkit from which a range of measures can be constructed.
As an illustration, one of several suggested measures is evaluated on a large text corpus built from an on-line encyclopedia.
Vision-Based Localization and Scanning of 1D UPC and EAN Barcodes with Relaxe...Vladimir Kulyukin
V. Kulyukin & T. Zaman. "Vision-Based Localization and Scanning of 1D UPC and EAN Barcodes with Relaxed Pitch, Roll, and Yaw Camera A lignment Constraints." International Journal of Image Processing (IJIP), V olume (8) : Issue (5) : 2014, pp. 355-383.
Toward Blind Travel Support through Verbal Route Directions: A Path Inference...Vladimir Kulyukin
The work presented in this article continues our investigation of such assisted navigation solutions where the
main emphasis is placed not on sensor sets or sensor fusion algorithms but on the ability of the travelers to interpret and
contextualize verbal route directions en route. This work contributes to our investigation of the research hypothesis that
we have formulated and partially validated in our previous studies: if a route is verbally described in sufficient and
appropriate amount of detail, independent VI travelers can use their O&M and problem solving skills to successfully
follow the route without any wearable sensors or sensors embedded in the environment.
In this investigation, we temporarily put aside the issue of how VI and blind travelers successfully interpret route
directions en route and tackle the question of how those route directions can be created, generated, and maintained by
online communities. In particular, we focus on the automation of path inference and present an algorithm that may be used
as part of the background computation of VGI sites to find new paths in the previous route directions written by online
community members, generate new route descriptions from them, and post them for subsequent community editing.
Wireless Indoor Localization with Dempster-Shafer Simple Support FunctionsVladimir Kulyukin
A mobile robot is localized in an indoor environment
using IEEE 802.11b wireless signals. Simple support
functions of the Dempster-Shafer theory are used to combine evidence
from multiple localization algorithms. Emperical results
are presented and discussed. Conclusions are drawn regarding
when the proposed sensor fusion methods may improve performance
and when they may not.
RFID in Robot-Assisted Indoor Navigation for the Visually ImpairedVladimir Kulyukin
We describe how Radio Frequency Identification
(RFID) can be used in robot-assisted indoor navigation for
the visually impaired. We present a robotic guide for the
visually impaired that was deployed and tested both with
and without visually impaired participants in two indoor
environments. We describe how we modified the standard
potential fields algorithms to achieve navigation at moderate
walking speeds and to avoid oscillation in narrow spaces.
The experiments illustrate that passive RFID tags deployed
in the environment can act as reliable stimuli that trigger local
navigation behaviors to achieve global navigation objectives.
RoboCart: Toward Robot-Assisted Navigation of Grocery Stores by the Visually ...Vladimir Kulyukin
This paper presents RoboCart, a proof-of-concept
prototype of a robotic shopping assistant for the visually
impaired. The purpose of RoboCart is to help visually impaired
customers navigate a typical grocery store and carry purchased
items. The hardware and software components of the system
are presented. For localization, RoboCart relies on RFID tags
deployed at various locations in the store. For navigation, Robo-
Cart relies on laser range finding. Experiences with deploying
RoboCart in a real grocery store are described. The current
status of the system and its limitations are outlined.
This paper examines the appropriateness of natural language dialogue (NLD) with assistive robots. Assistive robots are defined in terms of an existing human-robot interaction taxonomy. A
decision support procedure is outlined for assistive technology
researchers and practitioners to evaluate the appropriateness of
NLD in assistive robots. Several conjectures are made on when
NLD may be appropriate as a human-robot interaction mode.
Ergonomics-for-One in a Robotic Shopping Cart for the BlindVladimir Kulyukin
Assessment and design frameworks for human-robot teams
attempt to maximize generality by covering a broad range of
potential applications. In this paper, we argue that, in assistive
robotics, the other side of generality is limited applicability: it is
oftentimes more feasible to custom-design and evolve an
application that alleviates a specific disability than to spend
resources on figuring out how to customize an existing generic
framework. We present a case study that shows how we used a
pure bottom-up learn-through-deployment approach inspired by
the principles of ergonomics-for-one to design, deploy and
iteratively re-design a proof-of-concept robotic shopping cart for
the blind.
Accelerate your Kubernetes clusters with Varnish CachingThijs Feryn
A presentation about the usage and availability of Varnish on Kubernetes. This talk explores the capabilities of Varnish caching and shows how to use the Varnish Helm chart to deploy it to Kubernetes.
This presentation was delivered at K8SUG Singapore. See https://feryn.eu/presentations/accelerate-your-kubernetes-clusters-with-varnish-caching-k8sug-singapore-28-2024 for more details.
Generating a custom Ruby SDK for your web service or Rails API using Smithyg2nightmarescribd
Have you ever wanted a Ruby client API to communicate with your web service? Smithy is a protocol-agnostic language for defining services and SDKs. Smithy Ruby is an implementation of Smithy that generates a Ruby SDK using a Smithy model. In this talk, we will explore Smithy and Smithy Ruby to learn how to generate custom feature-rich SDKs that can communicate with any web service, such as a Rails JSON API.
GraphRAG is All You need? LLM & Knowledge GraphGuy Korland
Guy Korland, CEO and Co-founder of FalkorDB, will review two articles on the integration of language models with knowledge graphs.
1. Unifying Large Language Models and Knowledge Graphs: A Roadmap.
https://arxiv.org/abs/2306.08302
2. Microsoft Research's GraphRAG paper and a review paper on various uses of knowledge graphs:
https://www.microsoft.com/en-us/research/blog/graphrag-unlocking-llm-discovery-on-narrative-private-data/
State of ICS and IoT Cyber Threat Landscape Report 2024 previewPrayukth K V
The IoT and OT threat landscape report has been prepared by the Threat Research Team at Sectrio using data from Sectrio, cyber threat intelligence farming facilities spread across over 85 cities around the world. In addition, Sectrio also runs AI-based advanced threat and payload engagement facilities that serve as sinks to attract and engage sophisticated threat actors, and newer malware including new variants and latent threats that are at an earlier stage of development.
The latest edition of the OT/ICS and IoT security Threat Landscape Report 2024 also covers:
State of global ICS asset and network exposure
Sectoral targets and attacks as well as the cost of ransom
Global APT activity, AI usage, actor and tactic profiles, and implications
Rise in volumes of AI-powered cyberattacks
Major cyber events in 2024
Malware and malicious payload trends
Cyberattack types and targets
Vulnerability exploit attempts on CVEs
Attacks on counties – USA
Expansion of bot farms – how, where, and why
In-depth analysis of the cyber threat landscape across North America, South America, Europe, APAC, and the Middle East
Why are attacks on smart factories rising?
Cyber risk predictions
Axis of attacks – Europe
Systemic attacks in the Middle East
Download the full report from here:
https://sectrio.com/resources/ot-threat-landscape-reports/sectrio-releases-ot-ics-and-iot-security-threat-landscape-report-2024/
Transcript: Selling digital books in 2024: Insights from industry leaders - T...BookNet Canada
The publishing industry has been selling digital audiobooks and ebooks for over a decade and has found its groove. What’s changed? What has stayed the same? Where do we go from here? Join a group of leading sales peers from across the industry for a conversation about the lessons learned since the popularization of digital books, best practices, digital book supply chain management, and more.
Link to video recording: https://bnctechforum.ca/sessions/selling-digital-books-in-2024-insights-from-industry-leaders/
Presented by BookNet Canada on May 28, 2024, with support from the Department of Canadian Heritage.
Epistemic Interaction - tuning interfaces to provide information for AI supportAlan Dix
Paper presented at SYNERGY workshop at AVI 2024, Genoa, Italy. 3rd June 2024
https://alandix.com/academic/papers/synergy2024-epistemic/
As machine learning integrates deeper into human-computer interactions, the concept of epistemic interaction emerges, aiming to refine these interactions to enhance system adaptability. This approach encourages minor, intentional adjustments in user behaviour to enrich the data available for system learning. This paper introduces epistemic interaction within the context of human-system communication, illustrating how deliberate interaction design can improve system understanding and adaptation. Through concrete examples, we demonstrate the potential of epistemic interaction to significantly advance human-computer interaction by leveraging intuitive human communication strategies to inform system design and functionality, offering a novel pathway for enriching user-system engagements.
GDG Cloud Southlake #33: Boule & Rebala: Effective AppSec in SDLC using Deplo...James Anderson
Effective Application Security in Software Delivery lifecycle using Deployment Firewall and DBOM
The modern software delivery process (or the CI/CD process) includes many tools, distributed teams, open-source code, and cloud platforms. Constant focus on speed to release software to market, along with the traditional slow and manual security checks has caused gaps in continuous security as an important piece in the software supply chain. Today organizations feel more susceptible to external and internal cyber threats due to the vast attack surface in their applications supply chain and the lack of end-to-end governance and risk management.
The software team must secure its software delivery process to avoid vulnerability and security breaches. This needs to be achieved with existing tool chains and without extensive rework of the delivery processes. This talk will present strategies and techniques for providing visibility into the true risk of the existing vulnerabilities, preventing the introduction of security issues in the software, resolving vulnerabilities in production environments quickly, and capturing the deployment bill of materials (DBOM).
Speakers:
Bob Boule
Robert Boule is a technology enthusiast with PASSION for technology and making things work along with a knack for helping others understand how things work. He comes with around 20 years of solution engineering experience in application security, software continuous delivery, and SaaS platforms. He is known for his dynamic presentations in CI/CD and application security integrated in software delivery lifecycle.
Gopinath Rebala
Gopinath Rebala is the CTO of OpsMx, where he has overall responsibility for the machine learning and data processing architectures for Secure Software Delivery. Gopi also has a strong connection with our customers, leading design and architecture for strategic implementations. Gopi is a frequent speaker and well-known leader in continuous delivery and integrating security into software delivery.
Builder.ai Founder Sachin Dev Duggal's Strategic Approach to Create an Innova...Ramesh Iyer
In today's fast-changing business world, Companies that adapt and embrace new ideas often need help to keep up with the competition. However, fostering a culture of innovation takes much work. It takes vision, leadership and willingness to take risks in the right proportion. Sachin Dev Duggal, co-founder of Builder.ai, has perfected the art of this balance, creating a company culture where creativity and growth are nurtured at each stage.
Securing your Kubernetes cluster_ a step-by-step guide to success !KatiaHIMEUR1
Today, after several years of existence, an extremely active community and an ultra-dynamic ecosystem, Kubernetes has established itself as the de facto standard in container orchestration. Thanks to a wide range of managed services, it has never been so easy to set up a ready-to-use Kubernetes cluster.
However, this ease of use means that the subject of security in Kubernetes is often left for later, or even neglected. This exposes companies to significant risks.
In this talk, I'll show you step-by-step how to secure your Kubernetes cluster for greater peace of mind and reliability.
Encryption in Microsoft 365 - ExpertsLive Netherlands 2024Albert Hoitingh
In this session I delve into the encryption technology used in Microsoft 365 and Microsoft Purview. Including the concepts of Customer Key and Double Key Encryption.
From Siloed Products to Connected Ecosystem: Building a Sustainable and Scala...
Vision-Based Localization & Text Chunking of Nutrition Fact Tables on Android Smartphones
1. IPCV 2013
Vision-Based Localization and Text Chunking of
Nutrition Fact Tables on Android Smartphones
Vladimir Kulyukin
Department of Computer Science
Utah State University
Logan, UT, USA
vladimir.kulyukin@usu.edu
Aliasgar Kutiyanawala
Department of Computer Science
Utah State University
Logan, UT, USA
aliasgar.k@aggiemail.usu.edu
Tanwir Zaman
Department of Computer Science
Utah State University
Logan, UT, USA
tanwir.zaman@aggiemail.usu.edu
Stephen Clyde
MSDC Corporation
Salt Lake City, UT, USA
swc@mdsc.com
Abstract—Proactive nutrition management is considered by
many nutritionists and dieticians as a key factor in reducing and
controlling cancer, diabetes, and other illnesses related to or
caused by mismanaged diets. As more and more individuals
manage their daily activities with smartphones, smartphones
have the potential to become proactive diet management tools.
While there are many vision-based mobile applications to process
barcodes, there is a relative dearth of vision-based applications
for extracting other useful nutrition information items from
product packages, e.g., nutrition facts, caloric contens, and
ingredients. In this paper, we present a vision-based algorithm to
localize aligned nutrition fact tables (NFTs) present on many
grocery product packages and to segment them into text chunks.
The algorithm is a front end to a cloud-based nutrition
management system we are currently developing. The algorithm
captures frames in video mode from the smartphone’s camera,
localizes aligned NFTs via vertical and horizontal projections,
and segments the NFTs into single- or multi-line text chunks. The
algorithm is implemented on Android 2.3.6 and Android 4.2.
Pilot NFT localization and text chunking experiments are
presented and discussed.
Keywords—computer vision; image processng; vision-based
nutrition information extraction; nutrition management
I. Introduction
According to the U.S. Department of Agriculture, U.S.
residents have increased their caloric intake by 523 calories
per day since 1970. A leading cause of mortality in men is
prostate cancer. A leading cause of mortality in women is
breast cancer. Mismanaged diets are estimated to account for
30-35 percent of cancer cases [1]. Approximately 47,000,000
U.S. residents have metabolic syndrome and diabetes.
Diabetes in children appears to be closely related to increasing
obesity levels. Many nutritionists and dieticians consider
proactive nutrition management to be a key factor in reducing
and controlling cancer, diabetes, and other illnesses related to
or caused by mismanaged or inadequate diets.
Surveys conducted by the American Dietetic Association
(http://www.eatright.org/) demonstrate that the role of
television and printed media as sources of nutrition
information has been steadily falling. In 2002, the credibility
of television and magazines as sources of nutrition
information were estimated at 14% and 25%, respectively. In
contrast, the popularity of the Internet increased from 13% to
25% with a perceived credibility of 22% in the same time
period. Since smartphones and other mobile devices have, for
all practical purposes, become the most popular gateway to
access the Internet on the go, they have the potential to
become proactive diet management tools and improve public
health.
Numerous web sites have been developed to track caloric
intake (e.g., http://nutritiondata.self.com), to determine caloric
contents and quantities in consumed food (e.g.,
http://www.calorieking.com), and to track food intake and
exercise (e.g., http://www.fitday.com). Unfortunately, many
such sites either lack mobile access or, if they provide it,
require manual input of nutrition data. Manual input
challenges on smartphones are well documented in the
literatures (e.g., [2], [3]).
One smartphone sensor that can alleviate the problem of
manual input is the camera. Currently, the smartphone
cameras are used in many mobile applications to process
barcodes. There are free public online barcode databases (e.g.,
http://www.upcdatabase.com/) that provide some product
descriptions and issuing countries’ names. Unfortunately,
since production information is provided by volunteers who
are assumed to periodically upload product details and to
associate them with product IDs, almost no nutritional
information is available and some of it may not be reliable.
Some applications (e.g., http://redlaser.com) provide some
nutritional information for a few popular products.
2. IPCV 2013
While there are many vision-based applications to process
barcodes, there continues to be a relative dearth of vision-
based applications for extracting other types of useful nutrition
information from product packages such as nutrition facts,
caloric contents, and ingredients. If successfully extracted,
such information can be converted it into text or SQL via
scalable optical character recognition (OCR) methods and
submitted as queries to cloud-based sites and services.
Another problem and challenge for mobile computing is
eyes-free access to nutrition information for visually impaired
(VI), blind, and low vision smartphone users. One tool that is
frequently mentioned in the literature for eyes-free access to
print matter is the K-NFB reader (www.knfbreader.com). The
K-NFB reader is a mobile OCR software tool for Nokia
mobile phones. Given lower incomes of many VI and blind
individuals, the cost of this technology ($2,500 per phone
installation), quite possibly, puts it out of reach for many VI
users. K-NFB users are required to learn to effectively align
print matter with the camera, which may not be a problem for
dedicated users but may dissuade others from adopting this
technology. More importantly, K-NFB users are required to
use small mobile phone keys for navigation and input. The
speaker volume is too low for use in outdoors and noisy places
such as shopping malls.
In a series of evaluation experiments conducted by the K-
NFB system’s developers and published at the company’s web
site, the system accurately identified simple black on white
text but did not perform well on documents with color
graphics and images, large signs, mixed and italic fonts. The
current version of the system cannot read round containers
such as cans or products with colored fonts and images and
can read flat top boxes only if the text is plain black on white,
which is a serious limitation for grocery products, because
most of grocery product packages contain colorful images and
variable fonts.
The Utah State University (USU) Computer Science
Assistive Technology Laboratory (CSATL) is currently
developing a mobile vision-based nutrition management
system for smartphone users. The system will enable
smartphone users to specify their dietary profiles securely on
the web or in the cloud. When they go shopping, they will use
their smartphones to extract nutrition information from
product packages with their smartphones’ cameras. The
extracted information includes not only barcodes but also
nutrition facts, such as calories, saturated fat, sugar content,
cholesterol, sodium, potassium, carbohydrates, protein, and
ingredients.
Our ultimate objective is to match the extracted
information to the users’ dietary profiles and to make dietary
recommendations to effect behavior changes. For example, if
a user is pre-diabetic, the system will estimate the amount of
sugar from the extracted ingredients and will make specific
recommendations to the user. The system, if the users so
choose, will keep track of their long-term buying patterns and
make recommendations on a daily, weekly or monthly basis.
Dieticians will also be able to participate in and manage the
system’s data flow. For example, if a user exceeds his or her
total amount of saturated fat permissible for the specified
profile, the system will notify the user and, if the user’s profile
has appropriate permissions, the user’s dietician.
In this paper, we present a vision-based algorithm to
localize aligned NFTs and to segment them into single- or
multi-line text chunks. The algorithm captures frames in video
mode from the phone camera, localizes aligned NFTs via
vertical and horizontal projections, and segments text chunks
from localized NFTs. The latter part is referred to in this paper
as text chunking. These segmented text chunks can
subsequently be input into OCR engines. However, scalable
mobile OCR is beyond the scope of this paper. The algorithm
has been implemented and tested on the Android 2.3.6 and
Android 4.2 platforms.
The remainder of our paper is organized as follows. Section
2 presents related work. Section 3 discusses the localization of
aligned NFTs. Section 4 covers how single- or multi-line text
chunks are segmented from localized NFTs. Section 5
discusses NFT localization and text chunking experiments. In
Section 6, the experimental findings are discussed and several
future work directions are outlined.
II. Related Work
Many current R&D efforts aim to utilize the power of
mobile computing to improve proactive nutrition management.
In [4], the research is presented that shows how such mobile
applications can be designed for supporting lifestyle changes
among individuals with type 2 diabetes and how these changes
were perceived by a group of 12 patients during a 6-month
period. In [5], an application is presented that contains a
picture-based diabetes diary that records physical activity and
photos taken with the phone camera of eaten foods. The
smartphone is connected to a glucometer via Bluetooth to
capture blood glucose values. A web-based, password-secured
and encrypted short message service (SMS) is provided to
users to send messages to their care providers to resolve daily
problems and to send educational messages to users.
The presented NFT localization algorithm is based on
vertical and horizontal projections used by numerous
computer vision researchers for object localization. For
example, in [6], projections are used to successfully detect and
recognize Arabic characters. The presented text chunking
algorithm also builds on and complements multiple projects in
mobile computing and mobile computer vision that capitalize
on the ever increasing processing capabilities of smartphone
cameras. In [7], a system is presented for mobile OCR on
mobile phones. In [8], an interactive system is presented for
text recognition and translation.
3. IPCV 2013
III. NFT Localization
A. Vertical and Horizontal Projections
Images captured from the smartphone’s video stream can
be divided into foreground and background pixels. In general,
foreground pixels are defined as content-bearing units in a
domain-dependent manner. For example, content can be
defined as black pixels, white pixels, pixels with specific
luminosity levels, specific neighborhood connection patters
(e.g., 4-connected, 8-connetected), etc. Background pixels are
those that are not foreground.
Horizontal projection of an image (HP) is a sequence of
foreground pixel counts for each row in an image. Vertical
projection of an image (VP) is a sequence of foreground pixel
counts for each column in an image. Figure 1 shows horizontal
and vertical projections of a black and white image with three
characters.
Suppose there is an m x n image I where foreground pixels
are black, i.e., ,0, yxI and the background pixels are
white, i.e., .255, yxI Then the horizontal projection of
row y and the vertical projection of column x can defined as
yf and xg , respectively:
1
0
1
0
.,255
;,255
m
y
n
x
yxIxg
yxIyf (1)
For the discussion that follows it is important to keep in
mind that the x axis in the image is the column dimension
whereas the y axis is the row dimension. In other words, the
vertical projections computed by xg along the x axis are
used in computing the vertical boundaries of NFTs while the
horizontal projections computed by yf along the y axis are
used in computing the NFTs’ horizontal boundaries.
B. Horizontal Line Filtering
In detecting NFT boundaries, three assumptions are
currently made: 1) a NFT is present in the image; 2) the NFT
present in the image is not cropped; and 3) the NFT is
horizontally or vertically aligned. Figures 2 shows
horizontally and vertically aligned NFTs. The detection of
NFT boundaries proceeds in three stages. Firstly, the first
approximation of vertical table boundaries is computed.
Secondly, the vertical boundaries computed in the first stage
are extended to the left and to the right. Thirdly, the upper and
lower horizontal boundaries are computed.
The objective of the first stage is to detect the approximate
location of the NFT along the horizontal axis ., ''
es xx This
approximation starts with the detection of horizontal lines in
the image, which is accomplished with a horizontal line
detection kernel (HLDK) that we developed in our previous
research and described in our previous publications [9]. It
should be noted that other line detection techniques (e.g.,
Hough transform [10]) can be used for this purpose. Our
HLDK is designed to detect large horizontal lines in images to
maximize computational efficiency. On rotated images, the
kernel is used to detect vertical lines. The left image of Figure
3 gives the output of running the HLDK filter on the left
image shown in Figure 2.
C. Detection of Vertical Boundaries
Let HLFI be a horizontally line filtered image, i.e., the image
put through the HLDK filter or some other line detection filter.
Let HLFIVP be its vertical projection, i.e., the projections
of white pixels computed by for each column of HLFI. The
right image in Figure 2 shows the vertical projection of the
HLFI on the left. Let VP be a threshold, which in our
application is set to the mean count of the white foreground
pixels in columns. In Figure 2 (right), VP is shown by a red
line. It can be observed that the foreground pixel counts in the
Figure 3. HLFI of Fig. 2 (left); its VP (right).
Figure 2. Vertically & Horizonally Aligned Tables.Figure 1. Horizontal & Vertical Projections.
4. IPCV 2013
columns of the image region with the NFT are greater than the
threshold. Once the appropriate value of the threshold is
selected, the vertical boundaries of an NFT are computed as
follows:
.&|max
;|min
'''
'
rlVPxr
VPxl
xxcgcx
cgcx
(2)
The pairs of the left and right boundaries that are two close to
each other, where ‘too close’ is defined as the percentage of
the image width covered by the distance between the right and
left boundaries. It has been experimentally found that the first
approximation along the vertical boundaries are often
conservative (i.e., text is cropped on both sides) and must be
extended left, in the case of '
lx , and right, in the case of '
rx .
To put it differently, the left boundary is extended to the
first column to the left of the current left boundary, for which
the projection is at or above the threshold, whereas the right
boundary is extended to the first column to the right of the
current right boundary, for which the vertical projection is at
or above the threshold. Figure 4 (left) shows the initial vertical
boundaries (VBs) extended left and right.
D. Detection of Horizontal Boundaries
The computation of the horizontal boundaries of the NFT
is confined to the image region vertically bounded by the
extended vertical boundaries ., rl xx Let HLFIHP be the
horizontal projection of the HLFI in Figure 2 (left) and let HP
be a threshold, which in our application is set to the mean
count of the foreground pixels in rows, i.e.,
.0| yfyfmeanHP Figure 4 (right) shows the
horizontal projection of the left HLFI in Figure 2. The red line
shows .HP
The horizontal boundaries of the NFT are computed in a
manner similar to the computation of its vertical boundaries
with one exception – they are not extended after the first
approximation is computed. There is no need to extend the
horizontal boundaries up and down, because the horizontal
boundaries do not have as much impact on subsequent OCR of
segmented text chunks as vertical boundaries. The horizontal
boundaries are computed as follows:
.&|max
;|min
uHPl
HPu
rrrfrr
rfrr
(3)
Figure 5 (left) shows the nutrition table localized via
vertical and horizontal projections and segmented from the left
image in Figure 2.
IV. Text Chunking
A typical NFT includes text chunks with various caloric
and ingredient information, e.g., “Total Fat 2g 3%.” To
optimize the performance of subsequent OCR, which is
beyond the scope of this paper, these text chunks are
segmented from localized NFTs. This approach is flexible in
that segmented text chunks can be wirelessly transmitted to
cloud servers for OCR. As can be seen in Figure 5 (left), text
chunks are separated by black colored separators. Formally,
text chunks are defined as text segments separated by
horizontal black separator lines.
Text chunking starts with the detection of these separator
lines. Let N be a binarized image with a segmented NFT and
let ip denote the probability of image row i containing a
black separator. If such probabilities are reliably computed,
text chunks can be localized. Toward that end, let jl be the
length of the j-th consecutive run of black pixels in row i
above a length threshold l . If m be the total number of such
runs, then ip is computed as the geometric mean of
.,...,, 10 mlll The geometric mean is more indicative of the
central tendency of a set of numbers than the arithmetic mean.
If is the mean value of all positive values of normalized
by the maximum value of ip for the entire image, the start
Figure 4. VB Extension (left); HP of Left HFLI in
Fig. 2 (right).
Figure 5. Localized NFT (left); Text Chunks (right).
5. IPCV 2013
and end coordinates, sy and ey , respectively, of every
separator along the y axis can be computed by detecting
consecutive rows for which the normalized values are above
the threshold as follows:
.&1|
;&1|
jpjpjy
ipipiy
e
s
(4)
Once these coordinates are identified, the text chunks can
be segmented from either the binarized or grayscale image,
depending on the requirements of the subsequent of OCR. As
can be seen from Figure 5 (right), some text chunks contain
single text lines while others have multiple text lines.
V. Experiments
The NFT localization algorithm was implemented on
Android 2.3.6 and Android 4.2. Forty five images were
captured on a Google Nexus One (Android 2.3.6) in a local
supermarket. The average running time of the algorithm is
approximately one second per frame. All images were checked
by a human judge to ensure that an NFT is present in the
image, is not rotated, and is not cropped along any of its four
sides. These images were then placed on a Google Nexus One
smartphone (Android 2.3.6) and on a Galaxy Nexus (Android
4.2) smartphone with the installed application to obtain images
with segmented NFTs and save them on the smartphones’
SDK cards. The processed images were then analyzed by a
human judge. On each original image, the four corners of an
NFT were manually marked to obtain the ground truth to
evaluate the segmentation process.
Figures 6 and 7 show the error histograms for the starting
and ending positions of the segmented NFTs along the
images’ x-axis, respectively. In both figures, the x-axis
encodes the error as a fraction of the NFT width while the y-
axis encodes the number of images with a specific error value.
Figures 8 and 9 show the error histograms for the starting and
ending positions of the segmented NFTs along the images’ y-
axis, respectively. In Figure 8 and 9, the x-axis encodes the
error as a fraction of the NFT height. Positive errors occur in
segmented images where segmented NFTs contain extra
background pixels whereas negative errors occur when NFTs
are cropped.
In general, positive errors are better for our purposes than
negative ones because negative errors signal information loss
that may result in subsequent OCR or image classification
errors. It should be observed that the performance of the NFT
localization algorithm has a mean error of 1% on the sample
of images. There was one notable outlier, for which the start
position on the x-axis error was 12.5% and the end position
error on the x-axis was 14%. The same image was the outlier
for the segmentation errors of the start and end positions along
the y-axis.
Figure 10 shows the outlier image that caused the
segmentation errors along both axes. It can be observed that
the NFT in this image lacks a black colored bounding box that
is usually present around nutrition fact tables. It is the absence
of this box that caused the algorithm to fail to obtain the exact
location of the NFT in the image.
Figure 6. Start Position Errors along X-axis.
Figure 7. End Position Errors along X-axis.
6. IPCV 2013
Figure 8. Start Position Errors along Y-axis.
Figure 9. End Position Errors along Y-axis.
The performance of the NFT localization algorithm along
the y-axis has the mean errors of 5% and 7% for the start and
end positions, respectively. Most errors, along both axes, are
caused by NFTs that are not bounded by boxes, one of which
is shown in Figure 10.
The preliminary evaluation of the text segmentation
algorithm was done on a set of 15 NFT images. A total of 303
text chunks (text segments between separator bars) were
manually identified. Of these manually detected chunks, the
algorithm detected 163, which gives a detection rate of 53.8%.
The average running time of the text chunking algorithm is
approximately half a second per localized NFT.
A statistical analysis of text chunk segmentation was
executed. All text chunks readable by a human judge were
considered as true positives. There were no true negatives
insomuch as all text chunks had text. Text chunks which could
not be read by a human judge were reckoned as false
positives. False positives also included detection of separator
bars between text chunks. Figure 11 contains a true positive
example and two examples of false positives.
There were no false negatives in our sample of images,
because all text chunks either contained some text or did not
contain any text. The performance of the text chunking
algorithm was evaluated via precision, recall, specificity and
accuracy were calculated. The average values for these
measures over the entire sample are given in Table 1.
VI. Discussion
The NFT localization algorithm had a mean error of 1% on
the sample of NFT images. The average accuracy of the text
chunking algorithm on the sample of images with localized
NFTs is 85%. While we readily acknowledge that these results
must be interpreted with caution due to small sample sizes, we
believe that the approaches presented in the paper show
promise as a front end vision-based nutrition information
extraction module of a larger nutrition management system.
One limitation of the presented approach to NFT
localization is that an image is assumed to contain a
horizontally or vertically aligned NFT. We are currently
working on relaxing this constraint to localize skewed NFTs in
captured frames.
Table I. Average Recall, Precision, Specificity, Accuracy.
Precision Recall Specificity Accuracy
0.70158 0.9308 0.88978 0.8502
The detection of skewed NFTs will make the system more
accessible to users that require eyes-free access to visual
information. However, it should be noted in passing that,
Figure 10. NFT Localization Outlier.
7. IPCV 2013
Figure 11. True Positive (top); Two False Positives (middle,
bottom).
while visually impaired, low vision, and blind populations
continue to be a major target audience of our R&D efforts,
nutrition management is of vital importance to millions of
sighted individuals who will not have a problem aligning their
smartphone cameras with NFTs on product packages.
Another important improvement is to couple the output of
the text chunking algorithm to an OCR engine (e.g., Tesseract
(http://code.google.com/p/tesseract-ocr) or OCRopus
(https://code.google.com/p/ocropus)). We have integrated the
Android Tesseract library into our application and run several
tests but were unable to analyze the collected data before the
submission deadline. We plan to publish our findings in a
future publication.
Finally, we would like to bring the combined frame
processing time to under one second per frame. This will
likely be accomplished by moving current code bottlenecks to
the Android NDK or using OpenCV Android libraries.
Acknowledgment
This project has been supported, in part, by the MDSC
Corporation. We would like to thank Dr. Stephen Clyde,
MDSC President, for supporting our research and
championing our cause.
References
[1] Anding, R. Nutrition Made Clear. The Great Courses,
Chantilly, VA, 2009.
[2] Kane S, Bigham J, Wobbrock J. (2008). Slide Rule:
Making M obile Touch Screens Accessible to Blind
People using Multi-Touch Interaction Techniques. In
Proceedings of 10-th Conference on Computers and
Accessibility (ASSETS 2008), October, Halifax, Nova
Scotia, Canada 2008; pp. 73-80.
[3] Kulyukin, V., Crandall, W., and Coster, D. (2011).
Efficiency or Quality of Experience: A Laboratory Study
of Three Eyes-Free Touchscreen Menu Browsing User
Interfaces for Mobile Phones. The Open Rehabilitation
Journal, Vol. 4, pp. 13-22, 2011, DOI:
10.2174/1874943701104010013.
[4] Årsand, E., Tatara, N., Østengen, G., and Hartvigsen, G.
2010. Mobile Phone-Based Self-Management Tools for
Type 2 Diabetes: The Few Touch Application. Journal of
Diabetes Science and Technology, 4, 2 (March 2010), pp.
328-336.
[5] Frøisland, D.H., Arsand E., and Skårderud F. 2010.
Improving Diabetes Care for Young People with Type 1
Diabetes through Visual Learning on Mobile Phones:
Mixed-methods Study. J. Med. Internet Res. 6, 14(4),
(Aug 2012), published online; DOI= 10.2196/jmir.2155.
[6] Al-Yousefi, H., and Udpa, S. 1992. Recognition of Arabic
Characters. IEEE Transactions on Pattern Analysis and
Machine Intelligence, 14, 8 (August 1992), pp. 853-857.
[7] Bae, K. S., Kim, K. K., Chung, Y. G., and Yu, W. P.
2005. Character Recognition System for Cellular Phone
with Camera. In Proceedings of the 29th Annual
International Computer Software and Applications
Conference, Volume 01, (Washington, DC, USA, 2005),
COMPSAC '05, IEEE Computer Society, pp. 539-544.
[8] Hsueh, M. 2011. Interactive Text Recognition and
Translation on a Mobile Device. Master's thesis, EECS
Department, University of California, Berkeley, May
2011.
[9] Kulyukin, V., Kutiyanawala, A., and Zaman, T. 2012.
Eyes-Free Barcode Detection on Smartphones with
Niblack's Binarization and Support Vector Machines. In
Proceedings of the 16-th International Conference on
Image Processing, Computer Vision, and Pattern
Recognition, Vol. 1, (Las Vegas, Nevada, USA), IPCV
2012, CSREA Press, July 16-19, 2012, pp. 284-290;
ISBN: 1-60132-223-2, 1-60132-224-0.
[10] Duda, R. O. and P. E. Hart. 1972. Use of the Hough
Transformation to Detect Lines and Curves in Pictures.
Comm. ACM, Vol. 15, (January 1972), pp. 11-15.