Image compression based on SVD algorithm

SVD(matrix singular value decomposition) through the singular value decomposition of the digital image matrix, the image is transformed into a singular value matrix containing non-zero value, to realize image compression and facilitate the storage and preservation of the image. Based on SVD image compression, this paper proposes a new method for determining the first K eigenvalues. The method of the matrix singular value decomposition to determine the first K eigenvalues is well supported by the python experimental environment. Experiments show that the compression ratio has reached 5.3 at the threshold of 0.5. The compression ratio is effectively increased when the threshold changes in the range of 0.01.