Research And System Implementation Of Blur And Color Cast In Video Quality Inspection

In recent years,with the rapid development of Internet technology,video surveillance systems have become more and more widely used.During the operation of the video surveillance system,some quality problems,such as blur and color cast,will inevitably occur in the video.These problems will greatly affect the effectiveness of monitoring,so the automatic diagnosis of video image quality becomes more and more important.In this context,this article focuses on the problem ofimage blur and color cast in video quality detection,and designs and implements a video quality detection system on the basis of the research.The paper’s main research is as follows:In the aspect of image blur detection,this paper proposes a blur detection algorithm based on decision tree to solve the problem of the difference of detection effect between motion blur image and out-of-focus blur image.The algorithm selects the texture information,DCT(Discrete Cosine Transform)high-frequency coefficients and edge gradients in the image as feature indicators,and converts the detection problem of the blurred image into the classification problem of the image.Experimental results show that in image classification tasks,compared with other comparison methods,the algorithm proposed in this paper has a higher accuracy and recall rate.In the aspect of image color cast detection,in view of the poor detection effect of color cast images in a scene with slight color cast and single color,this paper proposes a color cast detection algorithm based on SVM(Support Vector Machine)and decision tree.The algorithm uses multi-scale technology to process the image,makes full use of the local and global information of the image,and uses a combined classifier to make up for the shortcomings of a single SVM classification model in feature selection.Experimental results show that compared with-other algorithms,the algorithm proposed in this paper has a higher accuracy rate in image classification tasks,and it also shortens the detection time by 19%.Based on these image blur and lateral detection algorithms,the paper has designed and implemented a video quality inspection system.The core modules of the system include login and registration,video upload,video playback,data processing,fuzzy detection,color cast detection,system management,information statistics,and detection result display modules.The client of the system uses the VUE framework,the server uses the SpringBoot framework,and the image anomaly detection algorithm uses OPENCV and SKLEARN open source libraries.The experimental results show that the system is effective for blur detection and colour cast detection of video.