A Non-Reference Image Quality Objective Evaluation Algorithm Based on Neural Network

In this paper, subjective and objective image quality evaluation algorithms are reviewed, and point out the disadvantages of these algorithms. Aim at infrared image and human visual system, a non-reference image quality objective evaluation algorithm based on neural network is proposed, which is optimized from histogram, information entropy and wavelet transform algorithms. According to NIIRS, the algorithm is optimized after a lot of machine study works have been done. The experiment results are very similar with the results from NIIRS, a fast algorithm for image quality evaluation is implemented.