Computational Photography

HW 1
Daniel Jacobs

The project was to take separate red, green, and blue image channels, created by taking black and white images, and combine them to yield full color images. It was assumed that the channels were only translated relative to each other.
The approach I took was to pyramid the image down to about 200x200, then do an search over the window [-10, +10] in both x and y directions. From that estimate, each successive level of the pyramid was used to refine the estimate, over the window [-1, +1]. An issue with this approach was errors at the edge of the image caused the matching to fail. Using a smaller window in the center of the image fixed this issue. A second issue was a few matches failed to match correctly because of lighting, when using a sum square differences matcher. This was fixed by using a normalized cross correlation matcher.
The match on image 87u failed, likely due to the highly repetitive pattern of the hut that fills most of the image. The mismatch is shifted by the size of the repeating pattern, indicating that the matcher failed at one level of the pyramid, and then propagated the error, finding the next best match.


Green offsets
15 39
Red offsets
33 91

Green offsets
38 48
Red offsets
55 107

Green offsets
1 4
Red offsets
-1 9

Green offsets
25 35
Red offsets
38 51

Green offsets
6 43
Red offsets
32 87

Green offsets
7 15
Red offsets
14 50

Green offsets
3 2
Red offsets
5 5

Green offsets
25 57
Red offsets
33 125

Green offsets
1 6
Red offsets
0 12

Green offsets
2 2
Red offsets
3 4

Green offsets
2 16
Red offsets
4 42

Green offsets
0 2
Red offsets
0 6

Green offsets
-1 1
Red offsets
-1 13

Green offsets
1 1
Red offsets
2 4

Green offsets
10 -16
Red offsets
17 11

Green offsets
20 24
Red offsets
33 71

Green offsets
1 5
Red offsets
1 12

Green offsets
2 6
Red offsets
4 14

Green offsets
15 15
Red offsets
30 81

Green offsets
4 51
Red offsets
0 111

Green offsets
17 35
Red offsets
26 82