Integral Images


The positions in the integral image represent the sum of all pixels above and to the left of the pixel in the source image. The integral image is calculated in one pass over the image.

The destination buffers need to be padded with at least 1 pixel above and a rapp_alignment to the left of the image start. The border pixels have to be properly cleared.


The table can be used for calculating the integral image sum of any rectangle in the source image as

\[ S = table(C) - table(B) - table(D) + table(A) \]

Figure 1: Calculating the integral sum over a subsection.


Next section: Integral Images

Generated on 1 Jun 2016 for RAPP by  doxygen 1.6.1