The goal of this assignment is to take the digitized Prokudin-Gorskii glass plate images and, using image processing techniques, automatically produce a color image with as few visual artifacts as possible. I did the following:
File Name | G Displacement | R Displacement |
---|---|---|
cathedral.jpg | (5, 2) | (12, 3) |
monastery.jpg | (-3, 2) | (3, 2) |
nativity.jpg | (3, 1) | (7, 0) |
settlers.jpg | (7, 0) | (15, -1) |
emir.tif | (45, 19) | (106, 35) |
harvesters.tif | (59, 10) | (121, 7) |
icon.tif | (40, 16) | (89, 22) |
lady.tif | (57, -6) | (103, -13) |
self_portrait.tif | (76, 15) | (171, 23) |
three_generations.tif | (52, 5) | (110, 7) |
train.tif | (40, -3) | (91, 19) |
turkmen.tif | (56, 11) | (109, 15) |
village.tif | (61, -1) | (137, 19) |
Change to compare SSD difference of the canny edge version of the image at every image pyramid level
File Name | G Displacement | R Displacement |
---|---|---|
cathedral.jpg | (5, 2) | (12, 3) |
monastery.jpg | (-3, 2) | (3, 2) |
nativity.jpg | (3, 1) | (7, 0) |
settlers.jpg | (7, 0) | (14, -1) |
emir.tif | (51, 22) | (107, 40) |
harvesters.tif | (55, 10) | (117, 11) |
icon.tif | (38, 16) | (90, 22) |
lady.tif | (57, 9) | (120, 13) |
self_portrait.tif | (81, 15) | (175, 37) |
three_generations.tif | (56, 11) | (106, 7) |
train.tif | (40, 8) | (84, 28) |
turkmen.tif | (55, 17) | (117, 29) |
village.tif | (67, 12) | (138, 22) |
Before
After
Pixelwise trasform each channel. The transformation stretch each channel to it's full range ignoring the pure black/white and intensities appears less than a threshold.
These are the Blue channel histograms of emir.tif before and after the auto process.
Before
After