How does a SSD algorithm work?
Posted: 2014-09-24T00:29:22-07:00
How does a Sum of Squared Difference algorithm work when I want to compare two images and find a matching block?
Did you try a Google Search. See http://en.wikipedia.org/wiki/Mean_squared_error and http://en.wikipedia.org/wiki/Template_matching
In IM, there are a number of metrics used to do a subsection search, including RMSE and MSE, which are normalized forms of the SSD. For each shift position of the small image every pixel in the small image is compared to the corresponding pixel in the large image subsection and SSD from all the pixels is computed. For each pixel, the sum of the differences of each color is computed and added together and the squared difference from all the pixels compared is totaled. The best match occurs where the SSD for that shift position is the smallest.