Scribd
Upload
12 views6 pages

Mod5 Part2

Uploaded by

Apoorva
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
12 views6 pages

Mod5 Part2

Uploaded by

Apoorva
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 6

Image acquisition: This is the first step in the digital image processing.

An image is captured by a
sensor (such as digital camera) and digitized. The image that is acquired is completely unprocessed.
This step involves preprocessing such as scaling.

Image enhancement

1. Image enhancement is the process of manipulating an image to make it more suitable for a
specific application.

2. Various techniques are employed, and these may involve different image processing approaches.

3. The goal is to bring out hidden details or highlight specific features of interest, such as adjusting
brightness, contrast, etc.

Image Restoration

1. Image restoration focuses on improving the appearance of an image using mathematical or


probabilistic models of image degradation.

2. It tends to be more objective than subjective, aiming to reverse or compensate for the effects of
degradation, such as noise or blurring.

Color image processing

1. With the increasing use of digital images, color image processing has become more
important.
2. Color is utilized for extracting features of interest in an image, and this may involve color
modeling and processing in the digital domain.
3. Understanding and manipulating color information is crucial for various applications, es
MODULE 5

Boundary-following algorithm designed to generate an ordered sequence of points along


the boundary of a binary region in a clockwise direction.
Assumptions and Initialization:
1. Working with binary images where object points are labeled as 1, and background points are
labeled as 0.
2. Images are padded with a border of 0s to prevent objects from merging with the image
border.
3. Limiting the discussion to single regions, and extension to multiple disjoint regions is
achieved by processing regions individually.
Chain Code

 Chain codes are used to represent boundaries by connecting sequence of straight-


line segments with specific lengths and directions.

 Freeman chain codes are a type of boundary code that assigns a numerical value to
each segment direction.

 Generating a chain code directly by following a boundary pixel by pixel can result in a
long code and be sensitive to small disturbances or noise along the boundary.

 Resampling the boundary on a larger grid spacing is often used to address these
issues.

 Resampling involves selecting a larger grid spacing and assigning a boundary point to
each node of the larger grid based on its proximity to the original boundary.

 The resampled boundary can then be represented by a 4- or 8-code, indicating the


directions of the segments.

 Chain codes depend on the starting point, but normalization procedures help make
the code independent of the starting point.
 Normalization can also account for rotation and size changes by using the first
difference of the chain code.

 The text provides an example using an image of a circular stroke embedded in small
fragments.

 Smoothing and thresholding processes are applied to simplify the problem.

 Resampling the boundary and obtaining a Freeman chain code results in a more
concise representation of the object's shape.

 Chain codes significantly reduce the data needed to store boundary information.

 They offer a unified way to analyze the shape of an object's boundary.

Shape Number

The shape number of a chain-coded boundary, based on a 4-directional code, is defined as the first
difference of the smallest magnitude.

The order of a shape number refers to the number of digits in its representation.

The shape number is always even for a closed boundary, and its value limits the number of possible
different shapes.

The coded boundary can depend on the orientation of the grid used.

To normalize the grid orientation, one approach is to align the chain-code grid with the sides of the
basic rectangle defined in the previous section.

You might also like