The document outlines exercises on histogram equalization, Fourier transforms, and filtering for a digital image processing lab session. Histogram equalization aims to spread intensity values over the full range to enhance contrast. Fourier transforms describe images in the frequency domain allowing for filtering operations. A variety of filters including median, average, Gaussian, and unsharp filters are explored to reduce noise, smooth, and sharpen digital images.
The document outlines exercises on histogram equalization, Fourier transforms, and filtering for a digital image processing lab session. Histogram equalization aims to spread intensity values over the full range to enhance contrast. Fourier transforms describe images in the frequency domain allowing for filtering operations. A variety of filters including median, average, Gaussian, and unsharp filters are explored to reduce noise, smooth, and sharpen digital images.
The document outlines exercises on histogram equalization, Fourier transforms, and filtering for a digital image processing lab session. Histogram equalization aims to spread intensity values over the full range to enhance contrast. Fourier transforms describe images in the frequency domain allowing for filtering operations. A variety of filters including median, average, Gaussian, and unsharp filters are explored to reduce noise, smooth, and sharpen digital images.
The document outlines exercises on histogram equalization, Fourier transforms, and filtering for a digital image processing lab session. Histogram equalization aims to spread intensity values over the full range to enhance contrast. Fourier transforms describe images in the frequency domain allowing for filtering operations. A variety of filters including median, average, Gaussian, and unsharp filters are explored to reduce noise, smooth, and sharpen digital images.
Title: Histogram Equalization, Fourier transform and Filtering.
Objective: To understand histogram equalization, Fourier transform and the application of
different filters on the images.
Theory: Histogram equalization is a technique for adjusting image intensities to enhance
contrast. Equalization implies mapping one distribution (the given histogram) to another distribution (a wider and more uniform distribution of intensity values) so the intensity values are spreaded over the whole range. There are two types of histogram equalization techniques: global histogram equalization is a process over the entire area and Local histogram equalization which takes specific area of the image by using windows of different size.
Fourier transformation belongs to a class of digital image processing algorithms that can be utilized to transform a digital image into the frequency domain. After an image is transformed and described as a series of spatial frequencies, a variety of filtering algorithms can then be easily computed and applied, followed by retransformation of the filtered image back to the spatial domain. Filtering in image processing is a process that cleans up appearances and allows for selective highlighting of specific information. Filtering an Image is useful for many applications, including smoothing, sharpening, removing noise, and edge detection. A filter is defined by a kernel, which is a small array applied to each pixel and its neighbors within an image.
Exercises:
1. Histogram equalization
1.1 Consider the image square.tif. Use both global histogram equalization and local histogram equalization (computed over 3 x 3, 5x5 and 7x7 localizing windows) and generate the results. Discuss the effects of window size. See MATLAB function histeq.
1.2 For the images ‘Lake.jpg’, equalize the R, G and B images separately as previously and convert the images back to jpg format. Comments. 2. Fourier Transform
Consider generating a 256*256 image with the middle part a rectangular mask of dimension mn (similar to the one presented on pp. 268 of Gonzalez and Woods, 3rd Edition) and do the followings in Matlab:
2.1 Plot the corresponding (Fourier) amplitude spectrum.
2.2 Translate the rectangle by some amount to the up-right direction and plot the resulting image and the corresponding amplitude spectrum. Was the amplitude different from the amplitude you computed in part (2.1) above? 2.3 Now rotate (instead of translating) the rectangle and plot the resulting image and its corresponding amplitude spectrum. Write down what you observed. Use a centered spectrum and a log scale when you plot the spectrum in cases (2.2) through (2.3) 2.4 In all cases compute and plot the phase spectrum and write down what you observed. (Hint: use atan2 Matlab function).
3. Filtering
3.1 Implement the median filter on the T1 weighted Magnetic Resonance image first randomly pick 4000 pixels and set their values to zero to obtain a pepper noise corrupted image. Then apply the median filter with W = 3, i.e. 3x3 median filter. Use 5*5, 7*7 filters to improve the result. Try simple averaging filter of same size as the median filter and compare the results. Comments 3.2 Taking notably the images ‘cell’ and ‘fibers2’, Try smoothing, look at the Matlab functions imfilter and fspecial, notably average and Gaussian. What happens as the filter size increases? 3.3 Image sharpening: To counter-balance the effects of smoothing, try image sharpening. Use Matlab functions using low-pass versions and Laplacian ones with ‘unsharp’ version of fspecial. Comments.
