Gray scale adjustment algorithms on SIMD architectures

Several gray-scale adjustment algorithms, single-instruction multiple-data (SIMD) implementations, and order of growth performances are discussed. A median filtering algorithm which takes O[(log q)²] time is discussed. A histogram equalization algorithm which causes the integral of this histogram of the output image to be as nearly linear as possible, while following the constraint that any two pixels, if equal in shading in the input image, have the same shading in the output image, is also discussed. The author finds substantial reason to believe that since every algorithm attempted ran quickly (order polynomial in log of the size of the problem at worst), the gray-scale adjustment algorithms´ performance, and their cousins in statistics, graphics, and image processing can benefit from SIMD architectures