Homework 3 Edge Detection and Image Enhancement
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
Homework 3
Edge Detection and Image Enhancement
|
1 - Edge Detection [30pts]
Take color image med as input. Convert image into YCrCb color space. a) Apply Sobel, Prewitt, canny and LoGfilters on the Y channel only and find the edge map by thresholding the results appropriately. You can use function edge in Matlab. Indicate which operator gives the best result. Explain and discuss the results.
b) Add Gaussian noise to the luminance image (you can use imnoise function in Matlab with default parameters), and find the edges as described in (a) and (b) above. Which method is the best in the presence of noise? Explain why. |
|
2 – Contrast enhancement of Color Images [35pts] Take image geisha as input. Convert the images into YCrCb color space. Perform the following image enhancement operations on the Y channel only. a) Automatic gain control b) Histogram equalization c) Adaptive histogram equalization |
|
Review the Y outputs and histograms. Explain the results and differences in the visual quality and the histograms |
|
|
3 – Sharpness enhancement of Color Images [35pts]
Take image butterfly as input. Convert the image into YCrCb color space. Write the code for unsharp masking and test the gain parameter to get noticeable sharpening. Use double type for calculations.
a) Apply unsharp masking to the Y component only, and replace it back in the ycbcr image, then convert it back to RGB
b) Now apply unsharp masking to the Y, Cr, and Cb channels independently. Convert back to the RGB space for displaying the results. Compare the result with using unsharp masking on just the Y channel. Explain and comment on the differences clearly.
Submit:
1. Code for parts 1-3
2. Submit pdf report with observations and comments as required in highlighted text. Watch the size of the pdf file, you may downsize images as needed to keep size reasonable.
Grading:
Code quality: the code must run and produce the outputs presented in the report. NOTICE that although identical code may happen by chance, identical code is suspect and will be scrutinized. Observations and analysis: observations must reflect what was learned in class and any practical aspects of the implementation that lead to typical or atypical outcomes. Please include snapshots of the output images in your pdf file.
2022-09-25