Image Processing Basics - Digital Signal Processing: Fundamentals and Applications - page 764

c. Perform histogram equalization for the grayscale image in (b) and display the histogram
plots for both the original grayscale image and equalized grayscale image.
d. Perform histogram equalization for the color image in (a) and display the histogram plots
of the Y channel for both the original color image and equalized color image.
14.37. Given the image data “cruise.jpg”, perform the following linear filtering:
a. Convert the image to grayscale and then create an 8-bit noisy image by adding Gaussian
noise using the following code:
noise_image [ imnoise(I,'gaussian');
where I is the intensity image obtained from normalizing the grayscale image.
b. Process the noisy image using a Gaussian filter with the following parameters:
convolution kernel size ¼ 4, SIGMA ¼ 0.8. Compare the filtered image with the noisy
image.
14.38. Given the image data “cruise.jpg”, perform the following filtering process:
a. Convert the image to grayscale and then create an 8-bit noisy image by adding “pepper
and salt” noise using the following code:
noise_image [ imnoise(I,'salt & pepper');
where I is the intensity image obtained from normalizing the grayscale image.
b. Process the noisy image using median filtering with a convolution kernel size of 4 4.
14.39. Given the image data “cruise.jpg”, convert the image to the grayscale and detect the image
boundaries using Laplacian of Gaussian filtering with the following parameters:
a. Kernel size ¼ 4 and SIGMA ¼ 0.9
b. Kernel size ¼ 10 and SIGMA ¼ 10
Compare the results.
14.40. Given the image data “clipim2.gif', perform the following process:
a. Convert the indexed image to grayscale.
b. Adjust the color transformation functions (sine functions) to make the object indicated
by the arrow in the image red and the background color green.
14.41. Given the image data “cruiseorg.tiff”, perform JPEG compression by completing the
following steps:
a. Convert the image to grayscale.
b. Write a MATLAB program for encoding with the following features: (1) divide the
image into 8 8 blocks; (2) transform each block using the discrete-cosine transform;
(3) scale and round DCT coefficients with the standard quality factor. Note that using
lossless compression with the quantized DCT coefficients is omitted here for a simple
simulation.