Digital Image Processing Jayaraman Ppt

Color systems model human color perception and device representations: RGB for capture/display, CMYK for printing, and other spaces like HSV/HSI used for intuitive editing and segmentation. Color transformation and correction adjust white balance and color casts; color restoration and enhancement extend grayscale techniques to multi-channel data, respecting inter-channel relationships to avoid artifacts.

: Converting a continuous image into a discrete digital form. Sampling refers to spatial digitization, while quantization refers to amplitude (intensity) digitization. Components digital image processing jayaraman ppt

Direct manipulation of pixels, including intensity transformations (log, power-law), histogram equalization, and spatial filtering (smoothing/sharpening). Color systems model human color perception and device

F(u,v)=∑x=0M−1∑y=0N−1f(x,y)e−j2π(uxM+vyN)cap F open paren u comma v close paren equals sum from x equals 0 to cap M minus 1 of sum from y equals 0 to cap N minus 1 of f of open paren x comma y close paren e raised to the negative j 2 pi open paren the fraction with numerator u x and denominator cap M end-fraction plus the fraction with numerator v y and denominator cap N end-fraction close paren power diffusion models for generative tasks

Deep learning dominates many image-processing tasks, with architectures and training strategies continuously evolving. Self-supervised learning, diffusion models for generative tasks, and transformers for vision are active areas. Edge computing and on-device processing bring resource-aware models for real-time applications, while explainability, robustness, and fairness receive growing attention.