Question
Homework 09 Use data provided to register images • Use gaussian blur to smooth images • Use optical_flow_tvl1 to register smoothed images to original Get and display registration fit scores • Use scipy.signal.correlate to compare original smoothed image with itself (internal score). Hint: use mean to find the overall score for the image. • Use it again to compare each smoothed image to the original smoothed image (aka before scores), use internal score as expected value to adjust the scores • Once more to compare each registered image to the original smoothed image (aka after scores), use internal score as expected value to adjust the scores • Generate histograms showing before and after scores for the images Hint: Investigate Numpy's pad function. Replace "..." with functional code # Data making
Question image 1