A method and apparatus for detecting video frames

Abstract

Embodiments of the present application disclose a video frame detection method and apparatus. The method comprises the following steps of: obtaining a target video frame sequence; extracting video feature data of the target video frame sequence using a convolution neural network model, the convolution neural network model being set to be learned according to a plurality of reference history videoframe sequences; carrying out video reconstruction according to the video characteristic data to generate a reconstructed video frame sequence; based on a difference value between the target video frame sequence and the reconstructed video frame sequence, it is determined that there is an abnormal event in the target video frame sequence. With the embodiment of the present application, the detection accuracy and the detection efficiency of the video frame detection can be improved.

Description

technical field [0001] The present application relates to the technical field of video image processing, and in particular to a video frame detection method and device. Background technique [0002] With the progress of urbanization in the world and the popularization of automobiles, problems such as increased traffic congestion, frequent traffic accidents, and deterioration of the traffic environment have become increasingly serious in both developing and developed countries. The intelligent transportation system can comprehensively apply advanced computer processing technology, information technology, data communication transmission technology and electronic automatic control technology to the transportation management system, through the real-time collection, transmission and processing of traffic information, with the help of various scientific and technological means and equipment to coordinate and deal with various traffic conditions. [0003] Abnormal event detection...

AI Technical Summary

Problems solved by technology

In this way, it not only increases the difficulty of obtaining video data containing abnormal traffic events, but also has a greater impact on the training of video classification models due to the unbalanced distribution of positive samples and negative samples;

[0007] (3) There are many classifications of traffic abnormal events, and the differences between different types of traffic abnormal events are relatively large. Therefore, it is difficult to summarize the "commonness" of traffic abnormal events, which is not conducive to the training of video classification models

Images