CSC 520
Homework Instructions: All homework must be completed in legible writing. The professor will not attempt to decipher handwriting. If your handwriting is suspect, the professor suggests that the student type up his/her homework.
Homework that is late will be penalized 10% per10. Homework should always be turned in at the beginning of class unless otherwise instructed.
| 0 | 10 | 20 | 20 | 20 | 10 |
| 0 | 40 | 50 | 50 | 40 | 10 |
| 5 | 40 | 100 | 100 | 40 | 10 |
| 5 | 40 | 100 | 50 | 40 | 30 |
| 5 | 40 | 50 | 50 | 50 | 30 |
| 20 | 20 | 40 | 50 | 50 | 30 |
| 20 | 20 | 20 | 40 | 30 | 50 |
1. Compute the histogram for image 1 and plot as a bar graph.
2. From the histogram plot in 1 what can be said about the contrast of this image.
3. Apply the gamma (Power Law) transform to image 1 with c=1 and gamma = 0.40. ( s = c*r^gamma). Show the results in image form as in image 1.
4. Compute the normalized historgram on the image produced in 3. Explain how the histogram differs from the original. Discuss your results relative to contrast enhancement.
5. Apply an automatic contrast stretching (histogram equalization) technique to image 1. Show the results in image form. Compare the contrast stretched histogram plot to that of the original and explain whther the image contrast was improved.
6. Explain when an image negative may be used for enhancement and provide the transform for the creating a negative image.
7. Provide an equation for the inverse log such that it approximates the negative transform. If you cannot, explain.
8. Explain why a strictly monotonically increasing function must be used in the derivation for histogram equalization.
9. Explain the difference between unsharp masking and high boost filtering.
10. Enhance the sharpness of image 1. Expalin your choice of filters. Provide the result in image form.
11. Provide definitions for: a) kernel b) correlation filtering c) convolution filtering