An example of low rank approximation of grayscale image based on SVD

1. a matrix A: Original Grayscale Data
A is a 596 by 596 regular matrix, i.e., A has 596 singular values.
2. A10: the best approximation by matirices of rank 10
10 singular values from the largest
3. A20: the best approximation by matirices of rank 20
20 singular values from the largest
4. A50: the best approximation by matirices of rank 50
50 singular values from the largest
5. A100: the best approximation by matirices of rank 100
100 singular values from the largest
6. A200: the best approximation by matirices of rank 200
200 singular values from the largest