G52IIP Introduction to Image Processing
Deadline: 4:00 PM, 30th November 2011
Implement the following image processing operations using MATLAB.
Image Enhancement using Histogram Equalization (using image1 and image2 as testing images)
Task 1: Convert the image into YCbCr space, perform histogram equalization on the Y image and leave the Cb and Cr signals unchanged, convert the image back to RGB image and output the new image.
Task 2: Divide the input image into m x n blocks (e.g., 32x32), and repeat task 1 for each block and output the processed image.
Task 3: Apply this filter, (0.05, 0.25, 0.4, 0.25, 0.05), to smooth the vertical and horizontal block boundaries of the output image of task 2 (Y image only) and output the processed image (colour image).
Spatial Filtering (using image3 and image4 as testing images)
Task 4: Perform a 1 x 5 (horizontal) median filtering operation on the input image and then a 5 x 1 (vertical) median filtering operation on the output of the horizontal filtering operation (make sure to use appropriate padding to deal with pixels on the edges of the image) .
Task 5: Create two 1 x 7 Gaussian filter masks with a standard deviation of 0.25 pixel and 1 pixel respectively (make sure you normalize the coefficients so that they sum to 1), and apply the filters (first horizontally and then vertically) to the testing images (make sure to use appropriate padding to deal with pixels on the edges of the image) . List the filter masks in the final report.
2D FFT and Frequency Domain Filtering (using image2 and image4 as testing images)
Task 6: Perform 2D FFT on the testing images, display the spectrum images.
Task 7: Perform a low-pass filtering in the frequency domain, using a Gaussian Transfer Function, and display the filtered images. Specify the exact form of the transfer function used and display the transfer function as an image.
Task 8: Perform a high-pass filtering in the frequency domain, using a Gaussian Transfer Function, and display the filtered images. Specify the exact form of the transfer function used and display the transfer function as an image.
What to Submit
1. A hardcopy of your codes (must be commented appropriately) and a hardcopy of a short report (maximum 4 A4 pages including figures) which should include
3. Submit the hardcopy to the School office by the due date and submit the electronic copy through the School's course work submission system.