Plain-English companion pages for Richard Szeliski — Computer Vision: Algorithms and Applications, 2nd edition. These notes focus on the mid-semester CV1-CV7 arc in BITS Pilani M.Tech AIML: how images are formed, how raw pixels are filtered and transformed, and how stable features such as corners, edges, lines, SIFT descriptors, and HoG representations make matching and recognition possible.
Geometry, pinhole cameras, projection, lenses, photometric image formation, lighting, reflectance, digital sensors, Bayer sampling, and color spaces.
chapter 3Point operators, histograms, convolution, linear and nonlinear filtering, Gaussian smoothing, bilateral filtering, Fourier intuition, pyramids, and warps.
chapter 4Harris corners, structure tensors, adaptive non-max suppression, scale-space, SIFT, descriptor matching, Canny edges, Hough lines, and HoG.