Image Signal Processing

Por: Swayam . en: ,

Overview

This course spans both basics and advances in digital image processing. Starting from image formation in pin-hole and lens based cameras, it goes on to discuss geometric transformations and image homographies, a variety of unitary image transforms, several image enhancement methods, techniques for restoration of degraded images, and 3D shape recovery from images.


INTENDED AUDIENCE :
Any interested learners
PREREQUISITES :Digital Signal Processing. Familiarity with linear algebra and probability theory is desirable.
INDUSTRIES SUPPORT :Google, Amazon, Facebook, Microsoft, KLA-Tencor, Qualcomm, Intel, Analog Devices, Philips, GE, Siemens and many more.

Syllabus

COURSE LAYOUT

Week 1: Introduction to Image Processing, Basics of Imaging, Geometric TransformationsWeek 2: Hierarchy of Transformations, Rotational Representation, Homography ComputationWeek 3: Research Challenges Involving Camera Motion, Basics of Real Aperture Camera, Lens as LSI SystemWeek 4: Blur Kernels, Shape from X, Shape from FocusWeek 5: Shape from Focus, Generalized Shape from Focus, Depth from Defocus (DFD) and Motion BlurWeek 6: Unitary Image Transforms, From 1D to 2D Unitary Transforms, Higher Dimensional Unitary TransformsWeek 7: 2D Unitary Transforms, 2D Discrete Fourier Transform, 2D Discrete Cosine TransformWeek 8: Karhunen-Loeve Transform (KLT), Applications of KLT, Singular Value DecompositionWeek 9: Image Enhancement, Adaptive Thresholding, K-Means Clustering, ISODATA ClusteringWeek 10:Contrast Stretching, Noise Filtering, Non-local Mean Filtering, Impulse Noise Filtering, Noise Filtering in Transform Domain, Illumination CompensationWeek 11:Image Restoration, Ill-posed Problems, Matrix Conditioning, Matrix Numerical Stability, Inverse filter for Image Deblurring, Regularization TheoryWeek 12:Minimum Mean Square Error (MMSE) Estimator, Linear MMSE, Spatial Wiener Filter, Wiener filter in Fourier domain, Image Super-resolution, Super-resolution Examples

Plataforma