Research On Adaptive Video Shot Boundary Detection And Key Frame Extraction Method

With the rapid development of information technology,a variety of massive information is filled in people's lives,and video data is also one of them.People not only enjoy the convenience of massive video,but also face the problem of how to manage video data.The follow-up work of video retrieval,dimension,and index,abstract and so on requires analysis of the video content,and the video boundary detection and key frame extraction are the basis and the key to the completion of these follow-up tasks.In this paper,video boundary detection and key frame extraction are taken as the research focus,and the existing problems in shot boundary detection algorithm and key frame extraction algorithm are studied.The simulation platform is designed and fabricated.The main research contents are as follows:1.In view of the situation that the current shot boundary detection algorithm is easy to cause error detection,leakage detection and difficult to detect the feature change between adjacent frames of gradient shot.A novel shot boundary detection algorithm based on RGB color histogram feature and histogram of oriented gradients(HOG)feature is proposed.By calculating the differences between frames of multiple steps,a frame difference mode distance graph is generated,and we analyzed their patterns in the pattern distance diagram to see if they were gradients or abrupt shots.Convolution neural network(CNN)is also used to extract the features of video frames.Finally,the proposed algorithm is compared with single feature algorithm and other literature algorithms.2.Aiming at the fact that the background is dominant in the key frame extraction algorithm,in which the foreground target is too small and it is not easy to extract the features of moving targets in sports video,a key frame extraction algorithm for foreground moving target feature extraction based on background modeling(Vi Be)algorithm is proposed.The foreground target detection of video sequence was firstly carried out using Vi Be algorithm,afterwards the scale-invariant feature transformation(SIFT)features of the foreground moving target were extracted.Based on the similarity calculated from video frameseries,the key frames of video were output according to the key frame discrimination method.3.Finally,based on python,the experimental simulation platform is designed,and the key frame extraction of each shot is carried out on the basis of shot boundary detection,and the experimental simulation system is designed and made.The experimental results show that the shot boundary detection algorithm proposed in this paper can make up for the missed detection and error detection caused by a single color information,and the use of adaptive threshold can effectively reduce the uncertainty and instability of manually determined threshold,and the detection effect of this algorithm is better than that of other algorithms.In this paper,the precision and recall of the key frame extraction algorithm based on vibe algorithm are better than those of other algorithms,and to a certain extent,the missed selection and misselection of moving target features in general moving video and because of the changeable and small proportion of moving objects in moving video are solved.