UNIT-3 SENSORS AND MACHINE VISION
Requirements of a sensor, Principles and Applications of the
following types of sensors- Position sensors – Piezo Electric
Sensor, LVDT, Resolvers, Optical Encoders, pneumatic Position
Sensors, Range Sensors Triangulations Principles, Structured
Lighting Approach, Time of Flight, Range Finders, Laser Range
Meters, Touch Sensors ,binary Sensors, Analog Sensors, Wrist
Sensors, Compliance Sensors, Slip Sensors, Camera, Frame
Grabber, Sensing and Digitizing Image Data- Signal Conversion,
Image Storage, Lighting Techniques, Image Processing and
Analysis-Data Reduction, Segmentation, Feature Extraction,
Object Recognition, Other Algorithms, Applications- Inspection,
Identification, Visual Serving and Navigation.
Introduction
 Machine Vision is the application of computer vision
industry and manufacturing.
 It is requires also digital input or output devices and
Computers network to control other manufacturing
equipment such as robot arms.
 Application of machine vision is the inspection of
manufactured goods such as semiconductor
chips,automobiles,food pharmaceuticals.
 Machine vision systems use for Digital cameras, smart
camera, image processing software to perform similar
inspections.
Introduction
 Define Vision:
 Vision is the ability to see and recognize objects by
collecting the light reflected of these objects into an image
and processing that image.
 Define machine Vision:
 Machine vision also known as industrial image
processing. It is important tool for the optimization and
automatic monitoring of production process.
 Types of Machine vision system:
 1.Robot vision system.
 2.Computer vision system.
 3.Machine vision system.
Introduction
 Basic types of Machine vision:
 1.Image formation 2.Processing of image
 3.Analysis of image. 4.interpretation of image.
 Uses of Machine vision:
1.Mistake proofing 2.Tolerance Monitoring
3.Dimensional measurement 4.Part presence and orientation.
5.Robot guidance and alignment.
6.Traceability and marking Verification.
Various of components of machine vision system:
1.Lighting 2.Lenses 3.Vision Processing 4.Image sensor
5.Communication.
Introduction
 Major components of machine vision system:
Camera
 Camera capture important information store and archive it
and allow users or software to make decisions based on
image information.
 It produces analog signal to digitized by frame grabber.
 Captures individual still image.
 Examples: Measure and count product, calculate the
product weight and volume
 Digital Camera: Record images using Digital technology.
 Digital camera have a built in computer and all of them
record images electronically.
 Types of Camera:
 1.Vidicon Camera
 2.Charge coupled device 3.Charge injection Device(CID)
 4.Complementary medal oxide semiconductor(CMOS)
Camera
 Vidicon Camera:
 Which is used to work based on photo conductive
properties of semi conductors. Decrease in resistance with
the amount of incident light.
 Construction and Working:
Vidicon Camera
 This camera technique used is same as in television.
 Change in electrode is proportional to amount of light
received.
 It does not break image in pixel as CCD camera.
 Metallic film very thin so as to be transparent. it is facing
with cathode.
 Side is scanned by electron beam.
 Electron beam for scanning is formed by combination of
cathode 1.Control grid 2.Acceleration grid 3.Anode grid
 Photo conductive material is created by 30V.
 Before the next scanning which may be done after interval
of 1/50 or 1/25 sec.
Vidicon Camera
 Advantages:
 1.Low Cost 2.More flexible 3.Sensitivity
 4.Resolution is high 5.Long life and small size
 Disadvantages:
 1.Analog output 2.High image lag 3.High dark current
 4.Imposes constrains on system designer.
 Applications:
 Education
 Medicine
 Automobile
 Aerospace
 Oceanography.
 Television Industry.
Charge Coupled Device(CCD)
 It is a charge transfer device.
 CCD are sensors used in digital camera and video camera
to record still and moving images. Its capture lights and
converts digital data that is recorded by the camera.
 Construction and Working:
Charge Coupled Device(CCD)
 It consists of Image,Lens,Target plate, Video Camera ,TV
Monitor.
 Electrical charge is converted into a voltage.
 Semiconductor are converted
Into a sequence of voltages.
 This sequence of voltage is
Sampled digitized and than
Stored in memory.
 In case of digital devices such
As digital camera.
Charge Coupled Device(CCD)
 Advantages:
 Compact size
 More Reliable
 High Dynamic Range
 More sensitive at low light.
 Disadvantage:
 Silicon processing seems to be variation in light sensitivity from
pixel to pixel.
 Applications:
 Cam
 Fax Machines.
 Security surveillance Camera.
 Dentistry X ray
Charge injection Device(CID)
 A solid state imaging device utilizing an image sensor
composed of a two dimensional array of coupled MOS.
 (Metal Oxide Semiconductor) charged storage capacitors
and designed to convert near infrared energy to electrical
signals providing broad grey shade.
Charge injection Device(CID)
 Which is used Single pixel imaging device and two dimensional
arrays ,Large pixel charge capacities, Wide dynamic range.
 Recently developed CID images and camera system, new on
chip architectures and process technology.
 CID images posses structural characteristics and improve
radiation tolerance (-10^4 gamma).
 Advantages:
 Reduce Dark current Reduce background noise
 Enhanced detection limits.
 Disadvantages:
 No direct pixel access Slower readout
 Applications:
 Missile Tracking
 Factory Inspection
Complementary medal oxide semiconductor(CMOS)
 Camera sensor is an electronic chip that converts photons
to electrons for digital processing.
 It is basically working principle of photo electric effect.
Complementary medal oxide semiconductor(CMOS)
 CMOS sensor works converts incident photons into
electric charge.
 The generated voltage is proportional to brightness and
exposure time.
 CMOS is more resistant to smearing or blooming than a
CCD.
Complementary medal oxide semiconductor(CMOS)
 Advantages:
 High Speed
 Power Consumption.
 High Sensitivity.
 Disadvantage:
 Pattern noise.
 Applications:
 Digital Camera.
 Digital Video Camera.
 Digital CCTV Camera.
Frame Grabber
 A frame grabber is a device to acquire [grab] and convert
analog to digital images. Modern FG have many additional
features like more storage, multiple cam
 Its play a crucial role in the capture of high resolution, high
quality image in a variety of setting.
 Smart camera are capable
of capture digital still
image without the uses
of a frame grabber.
 Smart camera capture still
images for low end application.
 Mega pixels resolution of
cameras used.
Frame Grabber
 A typical frame grabber consists of – a circuit to recover the horizontal and
vertical synchronization pulses from the input signal;
 An analog to digital converter a color decoder circuit, a function that can
also be implemented in software
 some memory for storing the acquired image (frame buffer) a bus interface
through which the main processor can control the acquisition and access the
data
 Types of Frame Grabber:
 Analog Frame Grabber
 Digital Frame Grabber.
 Circuitry common to both analog and digital grabber.
 Purpose:
 High speed image acquisition
 High resolution.
 Application:
 Broadcasting
 Industrial inspection.
Sensing and Digitizing Image Data
Sensing and Digitizing Image Data
Sensing Image Data
Sensing and Digitizing Image Data
 Digitalized image Data.
Sensing and Digitizing Image Data
I. Signal Conversion
Analog to Digital conversion.
 Analog Signal :A signal could be an analog quantity that
means it is defined with respect to the time. It is
continuous signals.
Sensing and Digitizing Image Data
 Digital signals:
 Discrete samples of an analog signals taken over.
 Digital signals are discontinuous signals.
Sensing and Digitizing Image Data
 Sampling:
 Sampling is process of disonetizing the analog signal to
time.
Sensing and Digitizing Image Data
 Quantization:
 The process of digitizing the amplitude value is called
quantization.
 It is depends upon variable.
 Mathematical equation Y=Sin(X)
Sensing and Digitizing Image Data
 Image Quantitation
 Image Digitization.
 Image resolution
 Grey level resolution
 Encoding:
 Encoding is the process of converting data from one from
to another.
Sensing and Digitizing Image Data
 II. Image Storage/ frame grabber.
 A frame grabber is a device to acquire [grab] and convert
analog to digital images. Modern FG have many additional
features like more storage, multiple cam
 Its play a crucial role in the capture of high resolution, high
quality image in a variety of setting.
 III. Lighting Structured
 Machine vision system creates images by analyzing the
reflected light from an object, not by analyzing the object itself.
 Types:
 Structural Lighting Technique.
 Illumination Technique.
Sensing and Digitizing Image Data
 Structural Lighting Technique.
 Use of multiple light source.
 Use polarized light.
 Redirected light to see top.
 Different colored light.
 Slight color differences,
Presence or absence of oil films,
 More complex part recognition
Task.
Sensing and Digitizing Image Data
 Illumination Techniques:
 Object detection and extraction is enhanced.
 Types:
1.Diffuse surface.
2.Condensorprojectors.
3.Flood or spot projectors.
4.Collimators.
5.Images.
Sensing and Digitizing Image Data
 Illumination Techniques:
1.Back Light:
Detecting presence or
Absence of holes and gaps
Part placement or
Orientation for measuring
Objects.
Sensing and Digitizing Image Data
 Axial Diffuse Lighting:
 Structured Light:
Sensing and Digitizing Image Data
 Dark Field illumination:
 Bright field illumination:
Sensing and Digitizing Image Data
 Diffused dome Lighting:
 Strobe lighting or Flash Lighting:
 Polarized Light:
Image Processing and Analysis
Image Processing and Analysis
 Which is used to enhance or otherwise alter an image and
to prepare it for image analysis.
 Not extracted from the image.
 Image Data reduction:
 Is to reduce the volume of data.
 Windowing:
 Windowing involves using only a portion of the total
image stored in the frame buffer for image processing and
analysis.
 Digital Conversion:
 Digital conversion reduces the number of grey levels used
by the machine vision system.
Image Processing and Analysis
 Segmentation:
 Is the separation of one or more region or objects is an
image based on a discontinuous or similar criterion.
 Purpose is to convert images in to smaller entities.
 Example: boundaries or origiones
 Segmentation Techniques:
 Structural Segmentation Technique
 Stochastic Segmentation Technique
 Hybrid Technique
Image Processing and Analysis
 Segmentation Method:
Image Processing and Analysis
 1.Therosholding Methods:
 2.Edge Detection Method:
 Edge are significant local changes of intensity in an
image. Sudden changes of discontinuous in an image are
called edges
Image Processing and Analysis
3.Region Based Methods:
 Main purpose to use is to partition an image into regions.
 Boundaries between regions based on discontinuous in
grey levels or color properties.
Image Processing and Analysis
4.Clustering based Methods:
 It is based on division into homogenous clusters.
 Which used to convert image into clusters having pixels
with similar characteristics.
 Types:
1.Hard Clustering 2.Soft Clustering.
5.Partial differential equations:
Non linear isotropic diffusion filter is used to enhance edges.
It is an time critical application.
Faster method.
Better detect the edges and boundaries.
Image Processing and Analysis
6.Artificial Neural network based methods:
 Simulate the learning strategies of human brain for the
purpose of decision making.
 In this the problem is converted to issues which are solved
using neural network.
 No need to write complex programs.
 FEATURE EXTRACTION:
 To determine the features based on the area and
boundaries of the object by using thresholding and Edge
detection.
 Feature extraction actually transform the attributes.
Image Processing and Analysis
 FEATURE EXTRACTION:
 Geometry:
1.Size 2.Shape 3.Colour 4.Coimbination of them.
 Machine vision include area, diameter and perimeter can be
used to measure.
 Extract features values for two dimensional cases.
 Example: Character recognization.
 Discard redundant information.
 A 64 * 64 images 4096 dimensional feature space.
 Used to enhance speed effectiveness of supervised learning.
 Used to extract the themes of document collections.
Image Processing and Analysis
 Object Recognition:
 Is a computer vision technique for identifying objects
in images.
 It is the key output of deep learning and machine
learning algorithms.
Image Processing and Analysis
 Template matching technique:
 Which is used to finding small parts of an image.
 Templates are frequently used for recognition characteristics, numbers, objects.
 It can pixel to pixel to pixel matching or features based.
Similarly
 Feature detection technique:
 Most features are based on either regions or boundaries in an image.
 It is assumed that region or a closed boundary corresponds to an entity that is either
an object or a part of an object.
 Application:
 Robotics
 Biometric recognition
 Medical analysis
 Intelligent vehicle system.
 Industrial inspection.
 Survelliance.
 Driver less cars.
Image Processing and Analysis
Image Processing and Analysis
 Pattern Recognition:
 Recognition=Re+cognition
 Cognition means act or the process of knowing an entity.
 A pattern is an object ,process or event.
 A class is a set of patterns that share common attributes
usually from the same information source.
 Recognition classes are assigned to the objects.
Image Processing and Analysis
 Histogram:
 It makes task easier to identify different data.
 It helps to visualize the distribution of the data.
 It is like drawn like a bar chart. But often has bars of
unequal widths.
Applications- Inspection, Identification, Visual
Serving and Navigation.
 Object counting:
 Example : Identifying the presence or absence of the
object.
Applications- Inspection, Identification, Visual
Serving and Navigation.
 Machine vision system for bottle inspection:
 Original bottle image with no
defect.
 Bottle cap defected in top view.
 Bottle present but without cap
Condition.
 Bottle present with lower level
of beverage.
Applications- Inspection, Identification, Visual
Serving and Navigation.
 Identification:
Applications- Inspection, Identification, Visual
Serving and Navigation
 Machine vision system for Egg volume prediction:
Problems
Problems

Unit 3 machine vision

  • 1.
    UNIT-3 SENSORS ANDMACHINE VISION Requirements of a sensor, Principles and Applications of the following types of sensors- Position sensors – Piezo Electric Sensor, LVDT, Resolvers, Optical Encoders, pneumatic Position Sensors, Range Sensors Triangulations Principles, Structured Lighting Approach, Time of Flight, Range Finders, Laser Range Meters, Touch Sensors ,binary Sensors, Analog Sensors, Wrist Sensors, Compliance Sensors, Slip Sensors, Camera, Frame Grabber, Sensing and Digitizing Image Data- Signal Conversion, Image Storage, Lighting Techniques, Image Processing and Analysis-Data Reduction, Segmentation, Feature Extraction, Object Recognition, Other Algorithms, Applications- Inspection, Identification, Visual Serving and Navigation.
  • 2.
    Introduction  Machine Visionis the application of computer vision industry and manufacturing.  It is requires also digital input or output devices and Computers network to control other manufacturing equipment such as robot arms.  Application of machine vision is the inspection of manufactured goods such as semiconductor chips,automobiles,food pharmaceuticals.  Machine vision systems use for Digital cameras, smart camera, image processing software to perform similar inspections.
  • 3.
    Introduction  Define Vision: Vision is the ability to see and recognize objects by collecting the light reflected of these objects into an image and processing that image.  Define machine Vision:  Machine vision also known as industrial image processing. It is important tool for the optimization and automatic monitoring of production process.  Types of Machine vision system:  1.Robot vision system.  2.Computer vision system.  3.Machine vision system.
  • 4.
    Introduction  Basic typesof Machine vision:  1.Image formation 2.Processing of image  3.Analysis of image. 4.interpretation of image.  Uses of Machine vision: 1.Mistake proofing 2.Tolerance Monitoring 3.Dimensional measurement 4.Part presence and orientation. 5.Robot guidance and alignment. 6.Traceability and marking Verification. Various of components of machine vision system: 1.Lighting 2.Lenses 3.Vision Processing 4.Image sensor 5.Communication.
  • 5.
    Introduction  Major componentsof machine vision system:
  • 6.
    Camera  Camera captureimportant information store and archive it and allow users or software to make decisions based on image information.  It produces analog signal to digitized by frame grabber.  Captures individual still image.  Examples: Measure and count product, calculate the product weight and volume  Digital Camera: Record images using Digital technology.  Digital camera have a built in computer and all of them record images electronically.  Types of Camera:  1.Vidicon Camera  2.Charge coupled device 3.Charge injection Device(CID)  4.Complementary medal oxide semiconductor(CMOS)
  • 7.
    Camera  Vidicon Camera: Which is used to work based on photo conductive properties of semi conductors. Decrease in resistance with the amount of incident light.  Construction and Working:
  • 8.
    Vidicon Camera  Thiscamera technique used is same as in television.  Change in electrode is proportional to amount of light received.  It does not break image in pixel as CCD camera.  Metallic film very thin so as to be transparent. it is facing with cathode.  Side is scanned by electron beam.  Electron beam for scanning is formed by combination of cathode 1.Control grid 2.Acceleration grid 3.Anode grid  Photo conductive material is created by 30V.  Before the next scanning which may be done after interval of 1/50 or 1/25 sec.
  • 9.
    Vidicon Camera  Advantages: 1.Low Cost 2.More flexible 3.Sensitivity  4.Resolution is high 5.Long life and small size  Disadvantages:  1.Analog output 2.High image lag 3.High dark current  4.Imposes constrains on system designer.  Applications:  Education  Medicine  Automobile  Aerospace  Oceanography.  Television Industry.
  • 10.
    Charge Coupled Device(CCD) It is a charge transfer device.  CCD are sensors used in digital camera and video camera to record still and moving images. Its capture lights and converts digital data that is recorded by the camera.  Construction and Working:
  • 11.
    Charge Coupled Device(CCD) It consists of Image,Lens,Target plate, Video Camera ,TV Monitor.  Electrical charge is converted into a voltage.  Semiconductor are converted Into a sequence of voltages.  This sequence of voltage is Sampled digitized and than Stored in memory.  In case of digital devices such As digital camera.
  • 12.
    Charge Coupled Device(CCD) Advantages:  Compact size  More Reliable  High Dynamic Range  More sensitive at low light.  Disadvantage:  Silicon processing seems to be variation in light sensitivity from pixel to pixel.  Applications:  Cam  Fax Machines.  Security surveillance Camera.  Dentistry X ray
  • 13.
    Charge injection Device(CID) A solid state imaging device utilizing an image sensor composed of a two dimensional array of coupled MOS.  (Metal Oxide Semiconductor) charged storage capacitors and designed to convert near infrared energy to electrical signals providing broad grey shade.
  • 14.
    Charge injection Device(CID) Which is used Single pixel imaging device and two dimensional arrays ,Large pixel charge capacities, Wide dynamic range.  Recently developed CID images and camera system, new on chip architectures and process technology.  CID images posses structural characteristics and improve radiation tolerance (-10^4 gamma).  Advantages:  Reduce Dark current Reduce background noise  Enhanced detection limits.  Disadvantages:  No direct pixel access Slower readout  Applications:  Missile Tracking  Factory Inspection
  • 15.
    Complementary medal oxidesemiconductor(CMOS)  Camera sensor is an electronic chip that converts photons to electrons for digital processing.  It is basically working principle of photo electric effect.
  • 16.
    Complementary medal oxidesemiconductor(CMOS)  CMOS sensor works converts incident photons into electric charge.  The generated voltage is proportional to brightness and exposure time.  CMOS is more resistant to smearing or blooming than a CCD.
  • 17.
    Complementary medal oxidesemiconductor(CMOS)  Advantages:  High Speed  Power Consumption.  High Sensitivity.  Disadvantage:  Pattern noise.  Applications:  Digital Camera.  Digital Video Camera.  Digital CCTV Camera.
  • 18.
    Frame Grabber  Aframe grabber is a device to acquire [grab] and convert analog to digital images. Modern FG have many additional features like more storage, multiple cam  Its play a crucial role in the capture of high resolution, high quality image in a variety of setting.  Smart camera are capable of capture digital still image without the uses of a frame grabber.  Smart camera capture still images for low end application.  Mega pixels resolution of cameras used.
  • 19.
    Frame Grabber  Atypical frame grabber consists of – a circuit to recover the horizontal and vertical synchronization pulses from the input signal;  An analog to digital converter a color decoder circuit, a function that can also be implemented in software  some memory for storing the acquired image (frame buffer) a bus interface through which the main processor can control the acquisition and access the data  Types of Frame Grabber:  Analog Frame Grabber  Digital Frame Grabber.  Circuitry common to both analog and digital grabber.  Purpose:  High speed image acquisition  High resolution.  Application:  Broadcasting  Industrial inspection.
  • 20.
  • 21.
    Sensing and DigitizingImage Data Sensing Image Data
  • 22.
    Sensing and DigitizingImage Data  Digitalized image Data.
  • 23.
    Sensing and DigitizingImage Data I. Signal Conversion Analog to Digital conversion.  Analog Signal :A signal could be an analog quantity that means it is defined with respect to the time. It is continuous signals.
  • 24.
    Sensing and DigitizingImage Data  Digital signals:  Discrete samples of an analog signals taken over.  Digital signals are discontinuous signals.
  • 25.
    Sensing and DigitizingImage Data  Sampling:  Sampling is process of disonetizing the analog signal to time.
  • 26.
    Sensing and DigitizingImage Data  Quantization:  The process of digitizing the amplitude value is called quantization.  It is depends upon variable.  Mathematical equation Y=Sin(X)
  • 27.
    Sensing and DigitizingImage Data  Image Quantitation  Image Digitization.  Image resolution  Grey level resolution  Encoding:  Encoding is the process of converting data from one from to another.
  • 28.
    Sensing and DigitizingImage Data  II. Image Storage/ frame grabber.  A frame grabber is a device to acquire [grab] and convert analog to digital images. Modern FG have many additional features like more storage, multiple cam  Its play a crucial role in the capture of high resolution, high quality image in a variety of setting.  III. Lighting Structured  Machine vision system creates images by analyzing the reflected light from an object, not by analyzing the object itself.  Types:  Structural Lighting Technique.  Illumination Technique.
  • 29.
    Sensing and DigitizingImage Data  Structural Lighting Technique.  Use of multiple light source.  Use polarized light.  Redirected light to see top.  Different colored light.  Slight color differences, Presence or absence of oil films,  More complex part recognition Task.
  • 30.
    Sensing and DigitizingImage Data  Illumination Techniques:  Object detection and extraction is enhanced.  Types: 1.Diffuse surface. 2.Condensorprojectors. 3.Flood or spot projectors. 4.Collimators. 5.Images.
  • 31.
    Sensing and DigitizingImage Data  Illumination Techniques: 1.Back Light: Detecting presence or Absence of holes and gaps Part placement or Orientation for measuring Objects.
  • 32.
    Sensing and DigitizingImage Data  Axial Diffuse Lighting:  Structured Light:
  • 33.
    Sensing and DigitizingImage Data  Dark Field illumination:  Bright field illumination:
  • 34.
    Sensing and DigitizingImage Data  Diffused dome Lighting:  Strobe lighting or Flash Lighting:  Polarized Light:
  • 35.
  • 36.
    Image Processing andAnalysis  Which is used to enhance or otherwise alter an image and to prepare it for image analysis.  Not extracted from the image.  Image Data reduction:  Is to reduce the volume of data.  Windowing:  Windowing involves using only a portion of the total image stored in the frame buffer for image processing and analysis.  Digital Conversion:  Digital conversion reduces the number of grey levels used by the machine vision system.
  • 37.
    Image Processing andAnalysis  Segmentation:  Is the separation of one or more region or objects is an image based on a discontinuous or similar criterion.  Purpose is to convert images in to smaller entities.  Example: boundaries or origiones  Segmentation Techniques:  Structural Segmentation Technique  Stochastic Segmentation Technique  Hybrid Technique
  • 38.
    Image Processing andAnalysis  Segmentation Method:
  • 39.
    Image Processing andAnalysis  1.Therosholding Methods:  2.Edge Detection Method:  Edge are significant local changes of intensity in an image. Sudden changes of discontinuous in an image are called edges
  • 40.
    Image Processing andAnalysis 3.Region Based Methods:  Main purpose to use is to partition an image into regions.  Boundaries between regions based on discontinuous in grey levels or color properties.
  • 41.
    Image Processing andAnalysis 4.Clustering based Methods:  It is based on division into homogenous clusters.  Which used to convert image into clusters having pixels with similar characteristics.  Types: 1.Hard Clustering 2.Soft Clustering. 5.Partial differential equations: Non linear isotropic diffusion filter is used to enhance edges. It is an time critical application. Faster method. Better detect the edges and boundaries.
  • 42.
    Image Processing andAnalysis 6.Artificial Neural network based methods:  Simulate the learning strategies of human brain for the purpose of decision making.  In this the problem is converted to issues which are solved using neural network.  No need to write complex programs.  FEATURE EXTRACTION:  To determine the features based on the area and boundaries of the object by using thresholding and Edge detection.  Feature extraction actually transform the attributes.
  • 43.
    Image Processing andAnalysis  FEATURE EXTRACTION:  Geometry: 1.Size 2.Shape 3.Colour 4.Coimbination of them.  Machine vision include area, diameter and perimeter can be used to measure.  Extract features values for two dimensional cases.  Example: Character recognization.  Discard redundant information.  A 64 * 64 images 4096 dimensional feature space.  Used to enhance speed effectiveness of supervised learning.  Used to extract the themes of document collections.
  • 44.
    Image Processing andAnalysis  Object Recognition:  Is a computer vision technique for identifying objects in images.  It is the key output of deep learning and machine learning algorithms.
  • 45.
    Image Processing andAnalysis  Template matching technique:  Which is used to finding small parts of an image.  Templates are frequently used for recognition characteristics, numbers, objects.  It can pixel to pixel to pixel matching or features based. Similarly  Feature detection technique:  Most features are based on either regions or boundaries in an image.  It is assumed that region or a closed boundary corresponds to an entity that is either an object or a part of an object.  Application:  Robotics  Biometric recognition  Medical analysis  Intelligent vehicle system.  Industrial inspection.  Survelliance.  Driver less cars.
  • 46.
  • 47.
    Image Processing andAnalysis  Pattern Recognition:  Recognition=Re+cognition  Cognition means act or the process of knowing an entity.  A pattern is an object ,process or event.  A class is a set of patterns that share common attributes usually from the same information source.  Recognition classes are assigned to the objects.
  • 48.
    Image Processing andAnalysis  Histogram:  It makes task easier to identify different data.  It helps to visualize the distribution of the data.  It is like drawn like a bar chart. But often has bars of unequal widths.
  • 49.
    Applications- Inspection, Identification,Visual Serving and Navigation.  Object counting:  Example : Identifying the presence or absence of the object.
  • 50.
    Applications- Inspection, Identification,Visual Serving and Navigation.  Machine vision system for bottle inspection:  Original bottle image with no defect.  Bottle cap defected in top view.  Bottle present but without cap Condition.  Bottle present with lower level of beverage.
  • 51.
    Applications- Inspection, Identification,Visual Serving and Navigation.  Identification:
  • 52.
    Applications- Inspection, Identification,Visual Serving and Navigation  Machine vision system for Egg volume prediction:
  • 53.
  • 54.