Design Of Video Transcoding Server Software Based On Content

With the continuous progress of the times,video surveillance IP cameras have appeared in all aspects of people's lives in today's society.In this age of science and technology,people are beginning to make more use of surveillance video to obtain useful information.However,with the increase of video surveillance IP cameras,the improvement of the quality of the video captured by cameras also brings a series of problems to people.Among them,how to effectively store big data videos,remove unnecessary information,retain necessary information in videos to reduce storage consumption,and how to reduce bandwidth consumption during video transmission to improve transmission efficiency,are the two central issues to be addressed in this article.This paper designs and implements a content-based video transcoding server software solution based on saliency region extraction algorithm and super-resolution reconstruction algorithm.The main research contents are as follows:(1)Aiming at the traditional Itti visual attention calculation model,edge feature information is introduced to optimize the visual attention calculation model.In the process of introducing edge feature information,the canny edge detection algorithm was improved:the improved bilateral filtering algorithm was used to replace the Gaussian filter to better maintain the edges.Using the sobel operator to calculate the gradient amplitude from four directions instead of the process of calculating from two directions in the original method.The improved OTSU algorithm is used to select double thresholds instead of manually setting thresholds,thereby reducing false detection and missed detection during image segmentation.Finally,saliency region extraction is performed on the original image according to the acquired saliency area.(2)Then,the obtained saliency area is super-resolution reconstructed into a new image of the same size as the original image.This paper combines the optical flow field registration algorithm and the maximized mutual information algorithm to make the inter-frame registration more accurate,and then improves on the basis of the bicubic interpolation method to obtain an adaptive bicubic interpolation method,super-resolution reconstruction combined with the above algorithms can achieve a better reconstruction effect.(3)Finally encode the super-resolution reconstructed video sequence,compared with the original video data,the new video data effectively reduces the size of the video data to be stored while retaining important information,thereby increasing the bandwidth transmission rate.