Histogram stretching method, typical image enhancement scheme, enhances image quality by transferring
dynamic range of input image to totally available one. It has low complexity, and keep the histogram
shape of input image. The amount of enhancement, however, is so different according to dynamic range of
input image. Another typical method, histogram equalization method, convert histogram of input image to
one with uniform distribution, but there are some artifacts owing to over transformation. The proposed
algorithm is image enhance method combining strength of these two methods. Input image is decomposed
using SVD(singular value decomposition) in this procedure. After that, we can enhance image via weighted
average of singular values of histogram stretching and equalization.
Simulation results show that the proposed method yields higher subjective image qualities than other
conventional approaches, and also higher entropy that represents variety of pixel values in enhanced images