Signal Processing and Transforms

This group gathers numerical algorithms that transform or analyze a signal or image, including the fast Fourier transform, the CORDIC algorithm for computing trigonometric functions in hardware, image processing filters such as the Sobel operator, Canny edge detection, Gaussian blur, the median and bilateral filters, optical flow methods such as Lucas-Kanade and Horn-Schunck, the Pan-Tompkins algorithm for detecting heartbeats in an electrocardiogram, and fast approximation methods for magnitude and square root such as the alpha max plus beta min algorithm and fast inverse square root. Their shared work is converting a signal, image or waveform from one representation to another or extracting a feature from it, as distinct from solving a general system of linear equations, which belongs to Linear Algebra and Matrix Decomposition.