Research on the Influence of Gamma Correction Method on Local Feature Descriptors

Local feature descriptors are widely used in image processing, but their image matching failures in nonlinear illumination environments are very obvious, especially for daytime and nighttime illumination changes. The Sift algorithm is not effective in matching experiments, So illumination is a very difficult problem for local feature descriptors. In order to solve the problem of image matching under the illumination changing environment, this paper uses a gamma correction method to preprocess the illumination image. This method can improve the matching accuracy of sift descriptors in the illumination changing environment, and the experimental results show that the performance is effective