Lab4 - , and then show the original image and both of the...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
EE440 - Introduction to Digital Imaging Systems Lab. 4 Solve the problems using your own Matlab programs. Unless specified as valid in the problem, do not use advanced image editing commands such as imadjust , medfilt2, histeq, edge, etc. .. . 1) (20 %) 4_1.asc was created by corrupting a clean image with salt-and-pepper noise. More specifically, a uniform noise within the range [-50, 50] was added to each pixel value at a 15% probability. The perturbed intensity values are clipped to [0, 255]. Filter the images using the matlab function 'conv2' and masks of all ones of size 2x2 and 3x3, and print/submit. Which provides a better tradeoff between blurring and noise reduction for this image ? 2) (25 %) Choose two different filters (one low pass filter and one median filter ) to remove the noise of 4_1.asc
Background image of page 1
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: , and then show the original image and both of the results. Comment about them. You can use load to read asc files. Other useful commands are median , conv2 3) (30 %) Load the 3x3 noise-removing blurred image derived in Prob 1). "Sharpen" this blurred image by convolving with the following Laplacian mask. Does this operation reduce the blurring ? What about the original noise? Apply a second sharpening step using this Laplacian mask. Display and submit. Would repeated sharpenings help image interpretation ? 1 1 5 1 1 4) (25 %) Write a program to sharpen 4_2.bmp using a high boost filter. You can choose the scale factor by yourself. Hint: Useful MATLAB commands: rgb2hsv , hsv2rgb , (so that only process the graylevel intensity.), and conv2 to do convolution....
View Full Document

This note was uploaded on 07/08/2011 for the course EE 440 taught by Professor Jenq-nenghwang during the Spring '11 term at University of Washington.

Ask a homework question - tutors are online