This document summarizes a system for indoor augmented reality using locative technologies. The system uses GPS to locate rooms in a building and then displays augmented objects based on the device's orientation and markers in each room. Two methods are proposed - one that uses stored location data from GPS to identify rooms, and another that uses computer vision techniques to generate an indoor map and determine locations. The system is intended to allow anyone to augment rooms in a building by storing object and location data.
Goal location prediction based on deep learning using RGB-D camerajournalBEEI
In the navigation system, the desired destination position plays an essential role since the path planning algorithms takes a current location and goal location as inputs as well as the map of the surrounding environment. The generated path from path planning algorithm is used to guide a user to his final destination. This paper presents a proposed algorithm based on RGB-D camera to predict the goal coordinates in 2D occupancy grid map for visually impaired people navigation system. In recent years, deep learning methods have been used in many object detection tasks. So, the object detection method based on convolution neural network method is adopted in the proposed algorithm. The measuring distance between the current position of a sensor and the detected object depends on the depth data that is acquired from RGB-D camera. Both of the object detected coordinates and depth data has been integrated to get an accurate goal location in a 2D map. This proposed algorithm has been tested on various real-time scenarios. The experiments results indicate to the effectiveness of the proposed algorithm.
RECOGNIZING AND TRACKING OUTDOOR OBJECTS BY USING ARTOOLKIT MARKERSijcsit
We created an augmented reality platform for spatial exploration that recognizes buildings facades and displays various multimedia for different time points. In order to provide the user with the best user experience fast recognition and stable tracking are the key elements of any augmented reality app. In an outdoor environment, lighting, reflective surfaces and occlusion can drastically affect the user experience. In a setup where these conditions are similar, marker creation methodology and the app parameters are key. In this paper we focus on resizing the photo prior marker creating and the importance of camera calibration and resolution and their effect on the recognition speed and quality of tracking outdoor objects.
Depth Estimation from Defocused Images: a SurveyIJAAS Team
An important step in 3D data generation is the generation of depth map. Depth map is a black and white image which has exactly the same size of the original captured 2D image that indicates the relative distance of each pixel from the observer to the objects in the real world. This paper presents a survey of Depth Perception from Defocused or blurs images as well as image from motion. The change of distance of the object from the camera has direct relation with the amount of blurring of object in the image. The amount of blurring will be calculated with a comparison in front of the camera directly and can be seen with the changes at gray level around the edges of objects.
Interactive Full-Body Motion Capture Using Infrared Sensor Network ijcga
Traditional motion capture (mocap) has been well-studied in visual science for the last decades. However the field is mostly about capturing precise animation to be used in specific applications after intensive post processing such as studying biomechanics or rigging models in movies. These data sets are normally captured in complex laboratory environments with sophisticated equipment thus making motion capture a
field that is mostly exclusive to professional animators. In addition, obtrusive sensors must be attached to actors and calibrated within the capturing system, resulting in limited and unnatural motion. In recent year the rise of computer vision and interactive entertainment opened the gate for a different type of motion capture which focuses on producing optical markerless or mechanical sensorless motion capture. Furthermore a wide array of low-cost device are released that are easy to use for less mission critical applications. This paper describes a new technique of using multiple infrared devices to process data from multiple infrared sensors to enhance the flexibility and accuracy of the markerless mocap using commodity
devices such as Kinect. The method involves analyzing each individual sensor data, decompose and rebuild
them into a uniformed skeleton across all sensors. We then assign criteria to define the confidence level of
captured signal from sensor. Each sensor operates on its own process and communicates through MPI.
Our method emphasizes on the need of minimum calculation overhead for better real time performance
while being able to maintain good scalability.
Goal location prediction based on deep learning using RGB-D camerajournalBEEI
In the navigation system, the desired destination position plays an essential role since the path planning algorithms takes a current location and goal location as inputs as well as the map of the surrounding environment. The generated path from path planning algorithm is used to guide a user to his final destination. This paper presents a proposed algorithm based on RGB-D camera to predict the goal coordinates in 2D occupancy grid map for visually impaired people navigation system. In recent years, deep learning methods have been used in many object detection tasks. So, the object detection method based on convolution neural network method is adopted in the proposed algorithm. The measuring distance between the current position of a sensor and the detected object depends on the depth data that is acquired from RGB-D camera. Both of the object detected coordinates and depth data has been integrated to get an accurate goal location in a 2D map. This proposed algorithm has been tested on various real-time scenarios. The experiments results indicate to the effectiveness of the proposed algorithm.
RECOGNIZING AND TRACKING OUTDOOR OBJECTS BY USING ARTOOLKIT MARKERSijcsit
We created an augmented reality platform for spatial exploration that recognizes buildings facades and displays various multimedia for different time points. In order to provide the user with the best user experience fast recognition and stable tracking are the key elements of any augmented reality app. In an outdoor environment, lighting, reflective surfaces and occlusion can drastically affect the user experience. In a setup where these conditions are similar, marker creation methodology and the app parameters are key. In this paper we focus on resizing the photo prior marker creating and the importance of camera calibration and resolution and their effect on the recognition speed and quality of tracking outdoor objects.
Depth Estimation from Defocused Images: a SurveyIJAAS Team
An important step in 3D data generation is the generation of depth map. Depth map is a black and white image which has exactly the same size of the original captured 2D image that indicates the relative distance of each pixel from the observer to the objects in the real world. This paper presents a survey of Depth Perception from Defocused or blurs images as well as image from motion. The change of distance of the object from the camera has direct relation with the amount of blurring of object in the image. The amount of blurring will be calculated with a comparison in front of the camera directly and can be seen with the changes at gray level around the edges of objects.
Interactive Full-Body Motion Capture Using Infrared Sensor Network ijcga
Traditional motion capture (mocap) has been well-studied in visual science for the last decades. However the field is mostly about capturing precise animation to be used in specific applications after intensive post processing such as studying biomechanics or rigging models in movies. These data sets are normally captured in complex laboratory environments with sophisticated equipment thus making motion capture a
field that is mostly exclusive to professional animators. In addition, obtrusive sensors must be attached to actors and calibrated within the capturing system, resulting in limited and unnatural motion. In recent year the rise of computer vision and interactive entertainment opened the gate for a different type of motion capture which focuses on producing optical markerless or mechanical sensorless motion capture. Furthermore a wide array of low-cost device are released that are easy to use for less mission critical applications. This paper describes a new technique of using multiple infrared devices to process data from multiple infrared sensors to enhance the flexibility and accuracy of the markerless mocap using commodity
devices such as Kinect. The method involves analyzing each individual sensor data, decompose and rebuild
them into a uniformed skeleton across all sensors. We then assign criteria to define the confidence level of
captured signal from sensor. Each sensor operates on its own process and communicates through MPI.
Our method emphasizes on the need of minimum calculation overhead for better real time performance
while being able to maintain good scalability.
An Assessment of Image Matching Algorithms in Depth EstimationCSCJournals
Computer vision is often used with mobile robot for feature tracking, landmark sensing, and obstacle detection. Almost all high-end robotics systems are now equipped with pairs of cameras arranged to provide depth perception. In stereo vision application, the disparity between the stereo images allows depth estimation within a scene. Detecting conjugate pair in stereo images is a challenging problem known as the correspondence problem. The goal of this research is to assess the performance of SIFT, MSER, and SURF, the well known matching algorithms, in solving the correspondence problem and then in estimating the depth within the scene. The results of each algorithm are evaluated and presented. The conclusion and recommendations for future works, lead towards the improvement of these powerful algorithms to achieve a higher level of efficiency within the scope of their performance.
Enhanced Optimization of Edge Detection for High Resolution Images Using Veri...ijcisjournal
dge Detection plays a crucial role in Image Processing and Segmentation where a set of algorithms aims
to identify various portions of a digital image at which a sharpened image is observed in the output or
more formally has discontinuities. The contour of Edge Detection also helps in Object Detection and
Recognition. Image edges can be detected by using two attributes such as Gradient and Laplacian. In our
Paper, we proposed a system which utilizes Canny and Sobel Operators for Edge Detection which is a
Gradient First order derivative function for edge detection by using Verilog Hardware Description
Language and in turn compared with the results of the previous paper in Matlab. The process of edge
detection in Verilog significantly reduces the processing time and filters out unneeded information, while
preserving the important structural properties of an image. This edge detection can be used to detect
vehicles in Traffic Jam, Medical imaging system for analysing MRI, x-rays by using Xilinx ISE Design
Suite 14.2.
Interactive full body motion capture using infrared sensor networkijcga
Traditional motion capture (mocap) has been
well
-
stud
ied in visual science for
the last decades
. However
the fie
ld is mostly about capturing
precise animation to be used in
specific
application
s
after
intensive
post
processing such as studying biomechanics or rigging models in movies. These data set
s are normally
captured in complex laboratory environments with
sophisticated
equipment thus making motion capture a
field that is mostly exclusive to professional animators.
In
addition
, obtrusive sensors must be attached to
actors and calibrated within t
he capturing system, resulting in limited and unnatural motion.
In recent year
the rise of computer vision and interactive entertainment opened the gate for a different type of motion
capture which focuses on producing
optical
marker
less
or mechanical sens
orless
motion capture.
Furtherm
ore a wide array of low
-
cost
device are released that are easy to use
for less mission critical
applications
.
This paper
describe
s
a new technique of using multiple infrared devices to process data from
multiple infrared sensors to enhance the flexibility and accuracy of the markerless mocap
using commodity
devices such as Kinect
. The method involves analyzing each individual sensor
data, decompose and rebuild
them into a uniformed skeleton across all sensors. We then assign criteria to define the confidence level of
captured signal from
sensor. Each sensor operates on its own process and communicates through MPI.
Our method emphasize
s on the need of minimum calculation overhead for better real time performance
while being able to maintain good scalability
Video surveillance Moving object detection& tracking Chapter 1 ahmed mokhtar
Our Project was a security project with CNN (Machine learning ) By using detection of human and tracking and camera start record after detection a human then make alarm for the owner .
Parameterized Image Filtering Using fuzzy LogicEditor IJCATR
The principal source of blur in digital images arise during image acquisition (digitization) or transmission. The
performance of imaging sensors is affected by a variety of factors, such as the environmental conditions during image acquisition.
Blurry images are the result of movement of the camera during shooting (not holding it still) or the camera not being capable of
choosing a fast enough shutter speed to freeze the action under the light conditions. For instance, in acquiring images with a camera,
light levels and sensor temperature are major factors affecting the amount of blur in the resulting image.
Blur was implemented by first creating a PSF filter in MatLab that would approximate linear motion blur. This PSF was then
convolved with the original image to produce the blurred image. Convolution is a mathematical process by which a signal, in this case
the image, is acted on by a system, the filter, in order to find the resulting signal. The amount of blur added to the original image
depended on two parameters of the PSF: length of blur (in pixels), and the angle of the blur. This thesis work is going to provide a
new, faster, and more efficient noise reduction method for images corrupted with motion blur. This new filter has two separated steps
or phases: the detection phase and the filtering phase. The detection phase uses fuzzy rules to determine whether a image is blurred or
not. When blurry image is detected, Then we use fuzzy filtering technique focuses only on the on the real blurred pixels.
ABSTRACT Feature extraction plays a vital role in the analysis and interpretation of remotely sensed data. The two important components of Feature extraction are Image enhancement and information extraction. Image enhancement techniques help in improving the visibility of any portion or feature of the image. Information extraction techniques help in obtaining the statistical information about any particular feature or portion of the image. This presented work focuses on the various feature extraction techniques and area of optical character recognition is a particularly important in Image processing. Keywords— Image character recognition, Methods for Feature Extraction, Basic Gabor Filter, IDA, and PCA.
International Journal of Engineering Inventions (IJEI) provides a multidisciplinary passage for researchers, managers, professionals, practitioners and students around the globe to publish high quality, peer-reviewed articles on all theoretical and empirical aspects of Engineering and Science.
The peer-reviewed International Journal of Engineering Inventions (IJEI) is started with a mission to encourage contribution to research in Science and Technology. Encourage and motivate researchers in challenging areas of Sciences and Technology.
An Assessment of Image Matching Algorithms in Depth EstimationCSCJournals
Computer vision is often used with mobile robot for feature tracking, landmark sensing, and obstacle detection. Almost all high-end robotics systems are now equipped with pairs of cameras arranged to provide depth perception. In stereo vision application, the disparity between the stereo images allows depth estimation within a scene. Detecting conjugate pair in stereo images is a challenging problem known as the correspondence problem. The goal of this research is to assess the performance of SIFT, MSER, and SURF, the well known matching algorithms, in solving the correspondence problem and then in estimating the depth within the scene. The results of each algorithm are evaluated and presented. The conclusion and recommendations for future works, lead towards the improvement of these powerful algorithms to achieve a higher level of efficiency within the scope of their performance.
Enhanced Optimization of Edge Detection for High Resolution Images Using Veri...ijcisjournal
dge Detection plays a crucial role in Image Processing and Segmentation where a set of algorithms aims
to identify various portions of a digital image at which a sharpened image is observed in the output or
more formally has discontinuities. The contour of Edge Detection also helps in Object Detection and
Recognition. Image edges can be detected by using two attributes such as Gradient and Laplacian. In our
Paper, we proposed a system which utilizes Canny and Sobel Operators for Edge Detection which is a
Gradient First order derivative function for edge detection by using Verilog Hardware Description
Language and in turn compared with the results of the previous paper in Matlab. The process of edge
detection in Verilog significantly reduces the processing time and filters out unneeded information, while
preserving the important structural properties of an image. This edge detection can be used to detect
vehicles in Traffic Jam, Medical imaging system for analysing MRI, x-rays by using Xilinx ISE Design
Suite 14.2.
Interactive full body motion capture using infrared sensor networkijcga
Traditional motion capture (mocap) has been
well
-
stud
ied in visual science for
the last decades
. However
the fie
ld is mostly about capturing
precise animation to be used in
specific
application
s
after
intensive
post
processing such as studying biomechanics or rigging models in movies. These data set
s are normally
captured in complex laboratory environments with
sophisticated
equipment thus making motion capture a
field that is mostly exclusive to professional animators.
In
addition
, obtrusive sensors must be attached to
actors and calibrated within t
he capturing system, resulting in limited and unnatural motion.
In recent year
the rise of computer vision and interactive entertainment opened the gate for a different type of motion
capture which focuses on producing
optical
marker
less
or mechanical sens
orless
motion capture.
Furtherm
ore a wide array of low
-
cost
device are released that are easy to use
for less mission critical
applications
.
This paper
describe
s
a new technique of using multiple infrared devices to process data from
multiple infrared sensors to enhance the flexibility and accuracy of the markerless mocap
using commodity
devices such as Kinect
. The method involves analyzing each individual sensor
data, decompose and rebuild
them into a uniformed skeleton across all sensors. We then assign criteria to define the confidence level of
captured signal from
sensor. Each sensor operates on its own process and communicates through MPI.
Our method emphasize
s on the need of minimum calculation overhead for better real time performance
while being able to maintain good scalability
Video surveillance Moving object detection& tracking Chapter 1 ahmed mokhtar
Our Project was a security project with CNN (Machine learning ) By using detection of human and tracking and camera start record after detection a human then make alarm for the owner .
Parameterized Image Filtering Using fuzzy LogicEditor IJCATR
The principal source of blur in digital images arise during image acquisition (digitization) or transmission. The
performance of imaging sensors is affected by a variety of factors, such as the environmental conditions during image acquisition.
Blurry images are the result of movement of the camera during shooting (not holding it still) or the camera not being capable of
choosing a fast enough shutter speed to freeze the action under the light conditions. For instance, in acquiring images with a camera,
light levels and sensor temperature are major factors affecting the amount of blur in the resulting image.
Blur was implemented by first creating a PSF filter in MatLab that would approximate linear motion blur. This PSF was then
convolved with the original image to produce the blurred image. Convolution is a mathematical process by which a signal, in this case
the image, is acted on by a system, the filter, in order to find the resulting signal. The amount of blur added to the original image
depended on two parameters of the PSF: length of blur (in pixels), and the angle of the blur. This thesis work is going to provide a
new, faster, and more efficient noise reduction method for images corrupted with motion blur. This new filter has two separated steps
or phases: the detection phase and the filtering phase. The detection phase uses fuzzy rules to determine whether a image is blurred or
not. When blurry image is detected, Then we use fuzzy filtering technique focuses only on the on the real blurred pixels.
ABSTRACT Feature extraction plays a vital role in the analysis and interpretation of remotely sensed data. The two important components of Feature extraction are Image enhancement and information extraction. Image enhancement techniques help in improving the visibility of any portion or feature of the image. Information extraction techniques help in obtaining the statistical information about any particular feature or portion of the image. This presented work focuses on the various feature extraction techniques and area of optical character recognition is a particularly important in Image processing. Keywords— Image character recognition, Methods for Feature Extraction, Basic Gabor Filter, IDA, and PCA.
International Journal of Engineering Inventions (IJEI) provides a multidisciplinary passage for researchers, managers, professionals, practitioners and students around the globe to publish high quality, peer-reviewed articles on all theoretical and empirical aspects of Engineering and Science.
The peer-reviewed International Journal of Engineering Inventions (IJEI) is started with a mission to encourage contribution to research in Science and Technology. Encourage and motivate researchers in challenging areas of Sciences and Technology.
Location Provider with Privacy Using Localized Server and GPS Editor IJCATR
Maps are an essential part of any handheld device and use constantly used for navigation and other resources by application for providing location based data which can be used for customized examination outcomes, however these data are conservatively stowed on a L.B.S Server which are susceptible to attacks and misuse as these data’s are not usually have any significant security so these data’s can be sold or misused by some other parties. We try to eliminate the problem as well as provided added functionality to the conventional maps by providing a customizable map which has the added functionality of offline use mode in addition to the online mode
In this project, we describe a unique architecture for indoor navigation that integrates behavior recognition, multisensory indoor localization, and path-planning in order to pro-actively provide directions without direct input from users. To our knowledge, this is the first architecture that attempts to integrate the core navigation components of path planning and localization with intent prediction towards a more refined navigation solution. The system comprises of three core components: augmented reality, map representation and route planning, and plan recognition.
To achieve effective localization, we provide pre-built maps using QR code scanning distributed at various places of the indoor location. We are using Augmented Reality to make an intuitive and user friendly interface which uses QR codes for identification of various maps that are pre uploaded in the QR codes for the ease of users.
A smartphone from Google ATAP which creates a live 3D image of your nearby space, such that you can access those data anywhere and anytime.
For any queries contact me at : akhilanair94@gmail.com
Augmented reality (AR) is a technology which provides real time integration of digital content with the
information available in real world. Augmented reality enables direct access to implicit information
attached with context in real time. Augmented reality enhances our perception of real world by enriching
what we see, feel, and hear in the real environment. This paper gives comparative study of various
augmented reality software development kits (SDK’s) available to create augmented reality apps. The
paper describes how augmented reality is different from virtual reality; working of augmented reality
system and different types of tracking used in AR.
Design of Image Projection Using Combined Approach for TrackingIJMER
Over the years the techniques and methods that have been used to interact with the
computers have evolved significantly. From the primitive use of punch cards to the latest touch screen
panels we can see the vast improvement in interaction with the system. There are many new ways of
projection and interaction technologies that can reshape our perception and interaction
methodologies. Also projection technology is very useful for creating various geometric displays. In
earlier generations, the projector technology was used for projecting images and videos on single
screen, using large and bulky setup. To overcome the earlier limitations we are designing “Wireless
Image Projection Tracking”, which is a system that uses IR (Infrared) technology to track the body in
the IR range and uses their movements for image orientation and manipulations like zoom, tilt/rotate,
and scale. We are presenting a method of mapping IR light source position and orientation to an
image. By using this system we can also track single and multiple IR light source positions and also it
can be used effectively to see the image projection in 3D view. Extension in this technology can further
be useful for future tracking capabilities to implement the touch screen feature for commercial
applications.
Extraction of Buildings from Satellite ImagesAkanksha Prasad
Buildings are termed as important components for various applications. Building extraction is defined as a sub-problem of Object Recognition. Though, numerous building extraction techniques have been proposed in the literature. But still they often exhibit limited success in the real scenarios. The main purpose of this research is to develop an algorithm which is able to detect and extract buildings from satellite images. In the proposed approach feature-based extraction process is used to extract buildings from satellite images. The overall system is tested and high performance detection is achieved which shows the effectiveness of proposed approach.
Similar to Architecture for Locative Augmented Reality (20)
1. Architecture for Locative Augmented Reality
Chinar Patil
Rochester Institute of Technology
Abstract
In this paper, I present a new system for indoor augmented reality.
GPS is used to locate the rooms and then the augmented objects will
be displayed based on the orientation of the device and markers in
the room. The placement of the augmented objects in the scene
based on the physics is not in the scope of this study. The output
of this system is an object based on the location of the room, the
orientation of the device and a marker to define the object.
1 Introduction
Augmented reality has not been explored for indoor locations as
much as for outdoor locations. The current existing systems for in-
door augmented reality are specific to a particular use. The most
common is the marker based augmented reality. There are also
markerless systems which use either some plane, the orientation
of the camera or just project something when the camera starts.
This paper presents a generalized system with predefined targets
and locations. I propose two methods, one of which is simple as it
uses the data saved in the database. The second system is different
in terms of the data acquired. It is not tested and is just a method
which might be possible to implement.It basically uses computer
vision techniques to get the location.
The first system is tested on an Android phone to locate the rooms.
The challenge in this study was to get the locations as GPS nav-
igation does not provide indoor navigation. But, it still gives us
the GPS data. These locations are used to define the rooms in a
buildings.
2 Related Work
Augmented reality is not a new field and a lot of progress has been
made in this field. As of today, each and every industry is aiming to
use augmented reality to create a great user experience.
IKEA is using this technology to visualize furnitures which helps
the consumers to see the products in 3D in their rooms. This helps
to design the house as required and without any regrets as you have
already seen how the room would look.
ConstructAide is a system which uses augmented reality to show
the progress of a construction site. It gives information regarding
the progress in terms of the time of the project, whether the progress
is on schedule, etc. It uses the construction site as the marker, along
with the orientation of the camera to give different views from dif-
ferent angles. There is another application which uses indoor vir-
tual landmarks but not markers. This paper uses planes to display
objects. It also handles the orientation of the objects i.e. the way in
which is it displayed. GPS has been used for such systems but for
outdoor locations. Previous work has been enormous in the field of
augmented reality including gamification in the real world but there
is no generalized system which can be used by everyone to display
stuff in their rooms just by storing the database, an object in the
room which will act as a marker and the orientation of the camera.
3 System Design
3.1 Hardware and Software
The system requires a device such as a phone or Google Glass
which can get GPS data accurately and includes gyroscopes. In
order to get the GPS data, it is important to have a cellular network
as well. This is because using WiFi only does not work correctly.
Also, there can be some false positives. A web service is needed to
store the database. Some augmented reality software such as Vufo-
ria, AR Toolkit, etc. I have used Vuforia and it works very well for
marker based systems. To get the GPS data initially to define the
rooms, a software called GPS status is used. It provides each and
every data required for the implementation of this system, not just
the GPS data but also the phone data. One thing to note is that since
it is indoor navigation, the location coordinates should considered
upto 6 to 8 digits after decimal.
Figure 1: GPS Status software
2. Figure 2: System Architecture
4 Implementation
The first step is to gather the data related to each object to be aug-
mented. This step is required for both methods. The data collected
should be same but the rooms location data will differ. This is be-
cause the systems calculate the locations differently. The first sys-
tem uses GPS and the second has a model of its own which will be
discussed further.
The markers are the objects in the room. These can be sockets, ta-
bles, or just a plain wall. When a person enters the room and holds
his phone towards the markers, he will see the corresponding ob-
jects. But there arises another problem. There are multiple sockets,
multiple tables in one room. You don’t want same objects for each
of them. This is why I decided to use the orientation of the camera.
The direction in which the camera is pointing will do the job in such
situations. This was about displaying the objects. Before this the
objects should be defined and linked to the markers.
I used Vuforia SDK and it provides a database of its own to define
the markers. This database is downloaded as an xml file which
will be used to select the appropriate marker based on the above
described conditions. In case of same markers, they are identified
by unique names. Currently, this system is limited to local usage.
It is not yet generalized.
To have a generalized system, there needs to be some modification
as it would be a platform instead of a running application. A plat-
form would allow an administrator, i.e a person who wants to use
this system for his building or house, to insert the data he gath-
ered, in the database. Also, I suggest to use independent Vuforia
databases for every administrator. A Vuforia database is where the
markers are stored. Having this database independent will allow
less or no conflicts between marker names. So, as the number of
administrators increase, the number of vuforia databases will also
increase. This arises another need, selecting the vuforia database.
This problem will be solved with the help of the GPS location of
the building. This GPS location is implemented in both the meth-
ods as this belongs to the outdoor location and you can easily locate
a building, unlike a room.
The generalized application would just be a framework to develop
upon for whoever is interested in augmenting their room.
4.1 Using GPS
I used Google Maps API to get the location coordinates of each
room. Although GPS does not provide indoor navigation, it does
provide the location coordinates. This gives the latitude, longitude
which is stored in the database. Four locations of four corners are
saved for every room. This can be changed based on the shape of
the room and applying the corresponding equations to check if the
person is inside the room. For example, a circular room can have
only its center and radius stored.
There can be many ways to design the database in order to access
the correct marker. One of which can be storing the shape of room
and based on the shape it will be checked if the device is inside the
room. The code will continue to check and match whether you are
inside a room. This selects the room and then comes the selection
of the markers. This is implemented by first checking angle or ori-
entation of the phone and based on the orientation, get the name of
the marker. Now, since the first check locates the building and then
we locate the room corresponding to that building, having duplicate
names for markers by different administrators won’t be a problem.
After getting the marker name from the database, hand it over to
vuforia.
Vuforia will do the image recognition and display the correspond-
ing objects. This can be modified a bit by using the orientation
after the image recognition step. This will give better performance
as compared to the other way. Image recognition shows the aug-
mented object as long as the marker is visible to the camera but
if the orientation is checked first, there is a possibility that even if
the marker is in range, it won’t display as it never gets into image
recognition. One thing to manage is the orientation range. Database
has only one value but when checking in the code, there should be
a range in all four direction - left, right, up, down.
4.2 Create Your Map
This is a difficult method to implement. It can be a project of its
own. This is mainly for indoor navigation. I propose the use of
computer vision techniques in this system to create a map of the
building so you can navigate inside and get the location values. I
won’t be discussing the techniques to be used for getting the map.
It is just a possibility as I haven’t implemented. I will talk about the
results that can be obtained from this technique and how the data
can be used. There is no GPS in this case for indoor navigation, it’s
only for locating the building.
In order for the administrator to provide the location specific data
inside the building, he needs a floor map. The floor map will be
scanned by a system running computer vision algorithms to get the
boundaries of the rooms. There is also a need to get the dimensions
of the rooms. For this, the floor map should have the dimensions
mentioned on it. Now, the algorithm should be able to read these
numbers and build a virtual map of its own. This is the method
I thought about at the beginning of the study. But eventually it
turned out that there is no need for the map, since the next step I
had thought of was can be independent of this step.
All that is needed is some location specific data to identify the
rooms. For this, I propose the use of accelerometers to get the
distance of the device from some predefined point in the building
on that floor. Accelerometer doesn’t give the direct distance data.
Some algorithm should be implemented to do this. There are many
applications which give the distance covered by a person. But this
won’t do the job. This system needs the distance from some pivot
point and it should decrease if the person approaches the pivot. This
is very difficult and turns out that it is much more challenging when
you try to implement it. In order to the reverse motion i.e. going
back to the pivot, again the orientation needs to be used. Thus mon-
itoring the device at every point and calculating the distance from
the pivot. This is a flawed concept as of now but there should be
some way to make it possible.
3. I will assume that we have what we want and go on with the next
step. So, as of now, we have a system that measures the distance
from some predefined point. The user needs to start the application
at this point in order to make it work. Thats the way in which he can
get the correct locations. Also, every floor in the building will have
this system independently and there has to be some way to figure
out the floor number. Currently, this can be entered by the user.
The administrator of that building will find the corresponding data
and have it in the database, just like the GPS method. Everything
else will be similar to the GPS method.
The question arises that why would anyone try to implement such a
complicated system which is flawed. The answer to this question is
because I had thought of this method before I went to GPS. I have
been researching in indoor navigation system systems, as GPS does
not provide navigation, which lead me to this path. I realized that
GPS does give data inside the building and that is when I switched
to the other method.
Although it is complicated, it sounds like a good concept to me and
can be refined and implemented with some other objectives.
Figure 3: Vuforia image target using marker
5 Discussion
I have implemented the system to check if it works for the room
identification. I used an Android Phone, Vuforia SDK and a
MySQL database. I have tested the augmented reality and room
detection parts independently. The system is not implemented com-
pletely. It works fine and can be easy to implement the whole
system using GPS. GPS has one problem that it does not update
quickly sometimes. The connection can be a problem at times.
A cellular network should help to overcome this problem. I have
tested this only with WiFi which also works good.
In case of the second method, it is just a technique which can be ex-
plored further. With the advent of new technologies such as Google
Tango and Microsoft HoloLens, it seems everything has been ex-
plored. It is my personal opinion that the second system can be
used to build onto something.
6 Conclusion
I have proposed a system architecture for indoor augmented reality.
In order to achieve this system, I have proposed two methods, one
using GPS as the main element to identify locations and the other
is using mobile device’s accelerometer data or some other distance
calculation technique to get the location of the device. The second
method is not tested at any extent and is not feasible for this sys-
tem. Though it can be used to implement another indoor navigation
system. The GPS method works very good and is tested to set the
Figure 4: Room Detection using GPS coordinates. The coordinates
are latitude and longitude
location coordinates and get the rooms based on these coordinates.
The markers are tested for image targets using Vuforia SDK. Thus,
the system can be easily built by maintaining a database and getting
the device orientation data and storing everything in the database.
7 Future Work
For future work, interactions can be added to the objects. The ob-
jects can be made clickable which would do some animation. As
the person approaches a particular room, some pop ups can be given
on the camera providing information about the content in that room.
Augmented reality does not have any limits. It can go on with your
imagination. Whatever you imagine can be augmented into the real
world. Another task is to make the objects a part of the real world.
For this to be possible, There is a need to include lighting in the aug-
mented world. This lighting should correspond to the lighting in the
real world. So, another research topic comes out of this which is to
detect the light sources in the room. Lighting is very important to
make objects look real. This is because the human mind perceives
the objects based on the shadows.
4. 8 References
[1] Exploring the Evolution of Mobile Augmented Reality for Fu-
ture Entertainment Systems: Klen Copic Pucihar, Lancaster Uni-
versity, School of Computing and Communications, UK; Paul
Colton, Lancaster University, School of Computing and Commu-
nications, UK; June 2013.
[2] ConstructAide: Analyzing and Visualizing Construction Sites
through Photographs and Building Models: Kevin Karsch, Univer-
sity of Illinois; Mani Golparvar-Fard, University of Illinois; David
Forsyth, University of Illinois; November 2014.
[3] Orientation control for indoor virtual landmarks based on hy-
brid based markerless augmented reality: Fadhil Noer Afif, Ah-
mad Hoirul Basori; Faculty of Computing, Universiti Teknologi
Malaysia, 81310 UTM Skudai, Johor Darul Takzim, Malaysia;
November 2013.
[4] Qualcomm Vuforia SDK to implement the augmented reality
system