Method for evaluating video quality based on artificial neural net

Abstract

A video quality evaluation method based on artificial neural network, which belongs to the field of computer digital video processing. The evaluation algorithm computes damage extent of an image by analyzing spatial characters (blur, entropy, blocking effect, frequency-domain energy analysis, saturation) and time behaviors (frame-to-frame differences). Evaluation effect of the algorithm can be effectively modified by using colour space saturation as one of the parameters for non-reference evaluation algorithm. The system is designed based on artificial neural network, therefore, training process and testing process of the network are included in realization of the algorithm. To the selected training sample (video image sequence), firstly extracts six parameters of the sample, then obtains expected output (subjective evaluation result) in training through subjective evaluation. Characteristic parameters of the training sample and the corresponding subjective evaluation result are inputted to the artificial neural network as training parameters. The experiment shows that evaluation result obtained by the objective evaluation system is highly-consistent with visual sense of human eyes.

Description

technical field [0001] The invention relates to the field of computer digital video processing, and designs a non-reference digital video quality evaluation method based on an artificial neural network. Background technique [0002] Video image is the main form of visual information, and video image processing is one of the important research fields at present. All links in the field of video image applications, such as acquisition, compression, transmission, processing, storage, etc., will inevitably affect the image quality. Because the image is ultimately for users to watch, the correct evaluation of image quality is one of the key technologies in image information engineering. [0003] Evaluation methods of video image quality can be divided into two categories: subjective evaluation and objective evaluation. Subjective evaluation is a test that directly uses the observer's visual perception of the measured image to determine the image quality. The measurement result i...

AI Technical Summary

Problems solved by technology

But there are the following problems: (1) Considering the huge amount of data and high computational complexity of the original video evaluation algorithm; (2) In many applications, undistorted images cannot be obtained at the receiving end;

But there are the following problems: (1) From theoretical analysis, it is very likely that the statistical data of some damaged video images are exactly the same as the corresponding undistorted image, but they are quite different in subjective vision; (2) because the statistical data of the image is easy to Changes with changes in viewing distance, brightness range and other factors, and even get completely different statistical data, and this type of algorithm cannot fully consider this situation; (3) In some application environments, it is impossible to increase the auxiliary channel to transmit without distortion Image feature parameters

However, there is a lack of comprehensive analysis of video image feature parameters.

At present, there are many limitations in the algorithm, which cannot give the overall quality of the image or video, and it will also be subject to various restrictions in the application.

Images