Laboratory Assignment for Image Processing(CS 602, Autumn 2004)

Consider gray-level images for the following experiments. Write MATLAB fuctions if necessary.

The students are advised to do as much assignments as they can do and show to the instructor in the Laboratory period. Number of assignments submitted and the date of submission both will be considered in evaluation of laboratory performance of the students.

1. Perform the following experiment to see the effect of reducing/ increasing spatial resolution of an image. Take an image of size 1024 * 1024. Reduce the spacial resolution by half by replacing each 2 by 2 block of pixel by a single pixel. The gray value of the pixel can be taken as the mean of the four pixels in the block. This will reduce the size of the image and will be difficult to see the effect of reducing the number of samples. To have a better comparison bring the image up to size 1024 * 1024 by row and column pixel replication. Repeat the experiment to reduce the resolution to 256 by 256, 128 by 128, 64 by 64 32 by 32 and 16 by 16. In all cases bring the image to its original size. Sample Image



2. Take any image and reduce the number of gray levels from 256 to 128, 64, 32, ... up to 2. Sample Image



3. Take an image and produce the negative of the image.



4. Write MATLAB function to increase contrast of a low contrast images.



5. Take an 8-bit image and generate eight bit planes of the image. Sample Image



6. Write MATLAB fuction for Histogram equalization



7. Write MATLAB fuctions to enhance an image by histogram specification.


The following problems (8 to 13) are assigned on 05-11-2004


8. Implement FFT for one-dimensional function. Use it to implement two-dimensional FFT.
   
   
9. Write a function to perform spatial domain smoothing(averaging) filter using user-supplied mask size.
   
   
10. Write a function to perform spatial domain smoothing(weighted average) filter of size 3 X 3with user-supplied weights
    
    
11. Implement a unsharp masking and high-boost filter in spatial domain.
    
    
12. Implement the following low-pass filters in frequency domain.
    1. Ideal
    2. Butterworth
    3. Gaussian
    
    
    
13. Implement the following high-pass frequency domain filters.
    1. Ideal
    2. Butterworth
    3. Gaussian