Image processing

 Definition
 Key Components
 Functions
 Image Aquistion
 Vidicon Tube
 Charged Couple Device
 Analog to Digital Conversion
 Sampling
 Quantization
 Encoding
 Image Processing Issues and Methods
 Image Histograms

Image Enhancement
• Spatial Domain Methods
• Frequency Domain Methods
• Smoothing
• Sharpening
 Image Restoration
 Image Analysis
• Image data reduction
• Segmentation
• Feature Extraction and
• Object Recognition
 Object Recognition
 Pattern Recognition
• Template Pattern
• Weighing Pattern
 References

A process
involving image based
automatic inspection,
process control and
robot guidance in an industry.

Charged Couple Device
Its a device used for the movement of
electrical charge that are converted into
digital image.

1. SAMPLING
It is the process of
discretizing the
analog signal in
time

2. QUANTIZATION
In this process each samplified discret-time
voltage level is assigned to a finite no of
definite amplitude levels.

no of quantization levels = 2^n

3. ENCODING
It involves in representing an amplitude level by
binary digit.
Votage range, V Binary Gray Scale
0-0.0195 0000 0000 0(black)
0.0195-0.0390 0000 0001 1 (dark gray)
. . .
. . .
4.9805-5.0 1111 0000 255 (White)

1. Image Histograms
2. Image Enhancement
3. Image Restoration and
4. Image Analysis

Image Histograms
An image histogram acts
as a graphical representation
of the tonal distribution
in a digital image.
Helps in obtaining
threshold value
Converts image into
binary image
using the threshold value.

Image Enhancement
It is to improve image interpretability
or perception of the image for better
input to the automated
image proocessing techniques.
Image Enhancement Techniques
1. Spatial Domain Methods
2. Frequency Domain Methods
3. Smoothing
4. Sharpening

Spatian Domin Methods
These techniques are based on direct
manipulation of pixels in an image.
This process will be denoted by the expression
g(x,y) = T[f(x,y)]

Frequency Domain Method
In this we compute the
fourier transform of the image.
This process will be denoted by the expression
g(u,v) = h[f(u,v)]

Smoothing
Smoothing operations are iterative
in nature and are used to
improve quality of the image.
One of the smoothing techniques are
neighbourhood averaging.
This techique generates a
smooth image I’(x,y), from image I (x,y).

Sharpening
It refers to any enhancement
technique that highlights edges and
fine details in an image.
Image sharpening is widely used for
increasing the local contrast and
sharpening the images.

Image Restoration
Image restoration attempts
to restore images that
have been degraded.

1. Image data reduction
2. Segmentation
3. Feature Extraction and
4. Object Recognition

Image data reduction
It is used to reduce the volume of data. There are two
types of techniques we use .
1. Digital Conversion:
It is used to reduce no of grey levels
used by Machine Vision System.
Example:
For a 8 bit register for each pixel there would be 2^8 =
256 grey levels
For 4 bits it is 2^4 = 16 grey levels

Image data reduction (cont.,)
2. Windowing:
It uses only a portion of the
image stored in frame buffer
for image processing and analysis.
Contrast will be visible only for the
pixel values that are within the
selected windowand rest
display no contrast.

Segmentation
It is the process of partionining
a digital image into multisegments.
In this process a label is assigned
to every pixel in an image
such that pixels with same label
share certain characteristics.

Segmentation (cont.,)
Three important techniques that we will discuss are:
1. Thresholding
2. Region Growing
3. Edge Detection
4. Feature Extraction

Thresholding
In its simplest form, thresholding
is a binary conversion technique
in which each pixel is converted
into a binary level, either black or white.

Region Growing
It is a simple region based segmentation.
Here, the pixels are grouped into regions
called grid elements based on attributes
similarities.

Typical procedure:
1. Select a pixel that meets a criterion for inclusion in a
region. In the simplest case that cloud mean select
white pixel and assign a value of 1.
2. Compare the pixel selectd with all adjacent pixels.
Assign an equivaluent pixels if an attribute match
occurs.
3. Go to equivalent adjacent pixel and repeat the
process untill no equivalent pixels can be added.
Region Growing (cont.,)

Edge Detection
It considers the intensity change that occurs in the pixels
as the boundary or edges of a part.
The outline boundary of an object within an image is
equivalent to identifying the edges of the object that
seperate the object from its background.

Edge Detection (cont.,)
Typical procedure:
1. Scan the image from left to right starting from the
origin until an object with pixel value 1 is obtained
(i.e., starting point).
2. For a pixel within the region, turn left and step for
object pixel, otherwise turn right and step for
background pixel.
3. Repeat the procedure untill the traversed path
returns to starting point.

Feature Extraction
A feature, is a single parameter that permits ease of
comparison and identification.
Some features of objects that can be used in machine
vision systems include are area, diameter, perimeter,
ridge detection, template matching and deformable
paramaterized shapes.

Object recognition – task (within machine vision)
of finding and identifying objects in an image or
video sequence.
Pattern recognition focuses on the recognition
of patterns and regularities in data.
1. Template Matching
2. Weighing Feature

Template Matching
It is used for finding small parts of an image which
match a template image.
It can be used in manufacturing as a part of quality
control, a way to navigate a mobile robot, or as a
way to detect edges in images.
The features of the object that can be used in
template pattern are area, diameter, aspect ratio....

Feature Weighting
Several features like area, length and perimeter
are combined into a single measure by assigning a
weight to each feature according to the relative
importance in the identifying the object.
The score of the object in the image is compared
with the score of the image in the computer
memory to achieve proper identification.

References
 https://en.wikipedia.org/
 http://www.tutorialspoint.com/dip
 https://www.mathworks.com
 https://images.google.com/
 INDUSTRIAL ROBOTICS Technology, Programming and Applications
By MIKELL P. GROOVER, MITCHELL WEISS, POGER N.NAGEL, NICHOLAS G. ODREY
 ROBOTICS AND CONTROL
By R K MITTAL, I J NAGRATH

(DEFINITION)
Machine vision (MV) is the technology and methods used
to provide imaging-based automatic inspection and analysis
for such applications as automatic inspection, process control,
and robot guidance in industry.

(TYPICAL PARTS)
•Lets see the important parts/components of machine vision
system.
•Ligthing system for illumination of the object for so that the
features of the object are seen out and camera can clearly see.
•Lens captures the image and presents to camera in the form
of light.
•The sensor in MV camera converts the light into digital
image.
•Image process involves aquistion of image from camera and
involves functions like enhancement, restoration, analysis.
•The software sends a signal to robotic systems whether to
take an action or not depending upon the object inspection
results.
•Finally the robots will come into action if the product is
defect.

(FUNCTIONS)
Function s of MV systems.
1. Sensing and digizing image data
In above we encounter samping, Quantization and
Encoding which we will see later
2. Image processing and analysis
In above we encounter Data Reduction, Segmentation,
Feature Extraction and object recognition
3. Applications
In above we encounter Inspection, Identification, visual
sensing and navigation.

(IMAGE AQUISTUION)
Lets See the some of the camera types we use in industrial
Machine Vision System.
1. Vidicon System
The lens captures the image in the form of light and forms an
image forms on the glass face plate of the camera.
The face plate consists of two layers , front (Transparent
signal electrode film) and back (photo sensitive material)
The small areas present on the face plate generates a
decreasing electrical resistance in response to increasing
illumination and a charge will be developed leading to the
generation of an image on the face plate

cont., (Image Aquistion)
The scanning is controlled by the deflection coil.
For an aacumulated positive charge the electron beam
deposits enough neutrons to neutralize the charge .
Just read out the other labled names

(Image Aquistion )
2. Charged Couple devices
The CCD is a major piece of technology in digital imaging.
A charge-coupled device (CCD) is a device for the movement
of electrical charge, usually from within the device to an area
where the charge can be manipulated, for example
conversion into a digital value.
An electric charge proportional to the light intensity will be
developed upon projection of an image onto capacitor array.

cont., (Image Aquistion )
A one-dimensional array, used in line-scan cameras, captures
a single slice of the image, whereas a two-dimensional array,
used in video and still cameras, captures a two-dimensional
picture.
Once the array has been exposed the contents in capacitor
transfers it to neighbor as in shift register and the last
capacitor converts the contents into a sequences of voltages.
In a digital device, these voltages are then sampled, digitized,
and usually stored in memory.
Fig:1 Movement of Charged particles.
Fig : 2 Charged particles flow in a single array.
Fig 3: A 2 Mega Pixel CCD in Cameras.

(Analog to Digital Conversion)
1. Sampling:
What should be the first step in getting a digital signal? You need to
discretize the analog signal in time. This process is called the
process of sampling. For the purpose of sampling, the analog
signal values are measured at discrete intervals of time, say at an
interval of 1 second each. Physically this can be done with the help
of a simple circuit that uses a switch. The switch is on for a very
small amount of time at an interval of one second. Thus, the analog
signal passes through the switch during this small amount of time
only, when the switch is on. Hence we approximately get the
amplitude of the analog signal at these discrete periods of time.
Once a signal is sampled, it is called a discrete signal. The next step
towards digitization is Quantization.

Quantization:
Quantization, involved in image processing, is a lossy
compression technique achieved by compressing a range of values
to a single quantum value.
When the number of discrete symbols in a given stream is reduced,
the stream becomes more compressible.
For example, reducing the number of colors required to represent a
digital image makes it possible to reduce its file size.
Specific applications include DCT data quantization
in JPEG and DWT data quantization in JPEG 2000.

cont.,
Fig : 1
Quantization is opposite to sampling. It is done on y axis. When
you are qunaitizing an image , you are actually dividing a signal
into quanta(partitions).
On the x axis of the signal , are the co-ordinate values, and on the
y axis , we have amplitudes. So digitizing the amplitudes is known
as Quantization.

cont.,
FIG : 2
The relation of quantization with gray levels has been further
discussed below.
Relation of Quantization with gray level resolution:
The quantized figure shown above has 5 different levels of
gray. It means that the image formed from this signal ,
would only have 5 different colors. It would be a black and
white image more or less with some colors of gray. Now if
you were to make the quality of the image more better, there
is one thing you can do here. Which is , to increase the levels
, or gray level resolution up. If you increase this level to 256,
it means you have an gray scale image. Which is far better
then simple black and white image.

cont.,
Now 256 , or 5 or what ever level you choose is called gray
level. Remember the formula that we discussed in the previous
tutorial of gray level resolution which is,
We have discussed that gray level can be defined in two ways.
Which were these two.
1.Gray level = number of bits per pixel (BPP).(k in the
equation)
2.Gray level = number of levels per pixel.
Explain The Fig in SLIDE 8 (N = bits)

One of the image processing techniques are Image Histograms
Tonal ( lighting)
From Picture Explaining The Dark sides are represented on
the left side of the histogram graph and the bright sides are on
middle and right sides.
So by observing histogram alone we can adjust the constrat
on the image.

Explain by the fig
Fig 1: Dark shades
Fig 2: Bright
X-axis: Arrival per minute
Y- Axis : Frequency
SLIDE 14
T in en represents filter.

Such that the intensity atevery point (x,y) is obtained by
averaging the intensity values of pixels of I (x,y) in the
predefined neighbourhood of (x,y).
Disadvantage blurs the image and other sharp details.

Image noise is random (not present in the object imaged)
variation of brightness or color information in images, and is
usually an aspect of electronic noise. It can be produced by
the sensor and circuitry of a scanner or digital camera. Image
noise can also originate in film grain and in the
unavoidable shot noise of an ideal photon detector. Image
noise is an undesirable by-product of image capture that adds
spurious and extraneous information.

In principle, image sharpening consists of adding to the
original image a signal that is proportional to a high-pass
filtered version of the original image.

The ability to window is a valuable feature of all digital images.
Windowing is the process of selecting some segment of the total
pixel value range (the wide dynamic range of the receptors) and
then displaying the pixel values within that segment over the full
brightness (shades of gray) range from white to black.
Important point...Contrast will be visible only for the pixel values
that are within the selected window. All pixel values that are either
below or above the window will be all white or all black and
display no contrast.
The person controlling the display can adjust both the center and
the width of the window. The combination of these two
parameters determine the range of pixel values that will be
displayed with contrast in the image.

image segmentation is the process of partitioning a digital
image into multiple segments (sets of pixels, also known as
superpixels). The goal of segmentation is to simplify and/or
change the representation of an image into something that is
more meaningful and easier to analyze. Image segmentation is
typically used to locate objects and boundaries (lines, curves,
etc.) in images. More precisely, image segmentation is the
process of assigning a label to every pixel in an image such that
pixels with the same label share certain characteristics.

Fig: shows a regular image of an object with each pixel having
aspecific grey tone out of possible 255.
When it is not possible to find a single threshold for an entire
image(for an example, if many different objects occupy the
same scene, each having different levels of intensity), one
appoach is to partition of total image into smaller rectangular
areas and dtermine the threshold for each window being
analysed.
Once thresholding is done, then the identification of certain
objects in the image are done

It is a simple region based segmentation.
Here, the pixels are grouped into regions called grid elements
based on attributes similarities.
Defined regions can be examined as to whether they are
independent or can be merged to other regions by means of an
analysis of the difference in their average properties and
spatial connectiveness.

Typical procedure
Select a pixel that meets a criterion for inclusion in a region.
In the simplest case that cloud mean select white pixel and
assign a value of 1.
Compare the pixel selectd with all adjacent pixels. Assign an
equivaluent pixels if an attribute match occurs.
Go to equivalent adjacent pixel and repeat the process untill
no equivalent pixels can be added.
It is used when the images are not distinguashable from each
other by straight thresholding or Edge detection. It occurs
mainly due to poor lighting.

In vision system the feature should not depend on the position
and orientation. i., the camera should not depend on object
being presented in known and fixed position.

Image processing

More Related Content

What's hot (20)

Similar to Image processing (20)

Recently uploaded (20)

Image processing