Image and Video Analysis
Lectures Summary
Perception
- Lecture 1 - Visual Perception, Perceptual Process, Light.
- Lecture 2 - Visual System, Structure of Eye, Illusions, Visual Cortex Cells.
Image Analysis
- Lecture 1 - Image Representation, Resolution & Dynamic Range, Spatial Frequency, Aliasing, RGB & YUV
- Lecture 2 - Interpolation, Measurement, Thresholding & Binary Images, Connected Components, Run-length Encoding.
- Lecture 3 - Images as Random Variables, Histograms, Measure of Distortion.
- Lecture 4 - Contrast Modification, Histogram Equalisation, Convolution, Gaussian Kernels, Sobel, Laplacian.
- Lecture 5 - Multiresolution Images, REDUCE & EXPAND, Laplacian Pyramids, Discrete Cosine Transform.
- Lecture 6 - Fourier transform, Ideal Low Pass Filtering.
Video Analysis
- Lecture 7 - Background Subtraction, Frame Differencing, Gaussian Model of Background.
- Lecture 8 - Feature Detection and Matching, Good Visual Features, Harris Corner Detector, SIFT, Feature Matching, Image Rectification, SURF.
- Lecture 9 - Motion Estimation, Optical Flow, Lucas-Kanade (LK) Approximation of Local Brightness, Eigenvalues of Hessian.
- Lecture 10 - View Geometry, Camera Calibration, Camera Matrix, Lens Distortion, Planar Homography, Image Mosaic.
Machine Learning
- Lecture 11 - Machine Learning Introduction, Linear Regression, Quadratic Loss, Learning Rate, Perceptrons, Multi-Layer Perceptrons, Feature Engineering, Deep Neural Networks.
- Lecture 12 - Linear Classification, Gradient Descent, Back Propagation.
- Lecture 13 - Data Handling, Data Partitioning, Activation Functions, Loss Functions.
- Lecture 14 - Convolutional Neural Networks, ImageNet Architecture, Transfer Learning.
- Lecture 15 - CNN Programming.
P(B) = 1/2
P(S) = 1/4
P(C) = 1/4
H = [0.5*log2(0.5)] + 2*[0.25*log2(0.25)]
= 0.5*-1 + 2*0.25*-2
= -0.5 +