1results about How to "Accurate quality evaluation" patented technology

The application relates to the technical field of image quality evaluation, in particular to an infrared image quality evaluation method based on multi-scale features, which comprises the following steps: acquiring a large number of infrared images labeled with quality score labels to construct a training sample set; constructing an infrared image quality evaluation model based on multi-scale features; wherein the infrared image quality evaluation model is composed of a main branch network, an auxiliary branch network and an output layer; inputting sample images in the training sample set into the infrared image quality evaluation model, iteratively training the infrared image quality evaluation model by using a mean square error loss function until convergence is achieved, and obtaining a trained infrared image quality evaluation model; inputting an infrared image to be evaluated into the trained infrared image quality evaluation model to obtain a quality score output by the infrared image quality evaluation model for the infrared image to be evaluated. The method effectively improves the accuracy of infrared image quality evaluation and has stronger generalization ability and stability.