Image Quality Evaluation Based on SVM and Improved Grid Search Algorithm

In order to improve the influence of distortion type image such as peak signal-to-noise ratio and structural similarity, an image quality evaluation method based on Support Vector Machine (SVM) and Improved Grid Search Algorithm is proposed. First, using Improved Grid Search Algorithm to find the optimal kernel parameters in SVM. Then, input the sample data into the improved SVM for predictive evaluation. Finally, an image quality evaluation model is established. According to the experimental results, the method of evaluating the image quality is efficient and simple to implement.