[TBD]Learning from Computer vision NPTEL course

Image formation

1. It is impression of the real object

2. It is projection of the <position, color intensity>

3. Projection happens to image plane

Image components

1. Background

2. Foreground

Image analysis

1. Histogram

   1. To plot the frequency vs color/brightness value

2. Probability Distribution Function

   1. Normalized histogram

Separation of forground from background

1. Simple approach is thresholding on frequency

   1. Automation

      1. Via Bayesian classification

         1. p(W) -> Computed as area of the frequency graph

      2. Otsu thresholding

         1. 
            

2. Contrast enhancement method

   1. ??

Image processing

1. Segmentation

   1. Partition image to meaningful non-overlapping sets

      1. 4-neighbor and 8-neighbor algorithms can be used

2.