AlMustansiriyah University Faculty of Engineering Computer Software Engineering

Al-Mustansiriyah University Faculty of Engineering Computer & Software Engineering Department Digital Image Processing lab. Experiment No. ( 8 ) Part (1): LPF for Noise Removal Aim: To understand the source of the noise and the advantage of cleaning the noisy images using LPF. Theory: Noise is the result of errors in the image pixel values that do not reflect the true intensities of the real scene. There are different kinds of image noise, for example: 1. Salt and pepper 2. Grain noise 3. Gaussian noise 4. Speckle noise 5. Periodic noise The first four types could be decrease or remove using spatial process, while the last one need a transformation to the frequency domain to reduce or remove it. MATLAB functions imnoise (x, ‘noise type’, parameter) imfilter ( x , f ) mask filter 2 ( f, x) fspecial (‘filter type’, parameter) medfilt 2(‘ x ‘, parameter ) x is an image, f is a

Al-Mustansiriyah University Faculty of Engineering Computer & Software Engineering Department Digital Image Processing lab. Requirements, Result and Discussion : 1. Read the colored image “ 8 png. png” and find the grayscale equivalent image “its ok for using built-in functions” a. Add ( 0. 1, 0. 25, 0. 4 ) salt & pepper noise to your grayscale image and display the result b. filter your noisy images with 3× 3 mean filter and 3× 3 median filter. Display your filtered image, which one is best? 2. Repeat step(1 a) and step(1 b) using for-loop statement only 3. write a program to read any true coloured (RGB) image then add 0. 2 salt noise to red plane, 0. 3 pepper to the green plane, and 0. 4 salt & pepper noise to the blue plane. a. display each individual noisy plane separately b. display each filtered plane separately, (use 3× 5 and 5× 7 median filter) and compare your result 25