This tool transforms an input raster image into an integral image, or summed area table. Integral images are the two-dimensional equivalent to a cumulative distribution function. Each pixel contains the sum of all pixels contained within the enclosing rectangle above and to the left of a pixel. Images with a very large number of grid cells will likely experience numerical overflow errors when converted to an integral image. Integral images are used in a wide variety of computer vision and digital image processing applications, including texture mapping. They allow for the efficient calculation of very large filters and are the basis of several of WhiteboxTools's image filters.
Crow, F. C. (1984, January). Summed-area tables for texture mapping. In ACM SIGGRAPH computer graphics (Vol. 18, No. 3, pp. 207-212). ACM.
def integral_image_transform(self, raster: Raster) -> Raster: ...