A monitoring video multi-target classification retrieval method and system based on depth learning

Abstract

The invention discloses a monitoring video multi-target classification retrieval method and system based on depth learning. The method comprises: extracting the object to be retrieved: inputting the image to be retrieved, selecting the object to be retrieved by automatic detection or manual digging, and determining the category of the object; extracting depth features: extracting depth features ofthe object to be retrieved by using depth learning feature extraction model; retrieving object: according to the category of the object to be retrieved, comparing the depth feature of the object to be retrieved with the depth feature of the same kind of object in the retrieval database to obtain the object most similar to it; displaying Retrieval results: displaying the retrieval results according to similarity from high to low. Depth features obtained by depth learning feature extraction module are used as the benchmark to match the retrieved target and the similar target in the retrieval database. Because depth features have the advantages of both strong expression and low dimension, the precision of retrieval results is greatly improved.

Description

technical field [0001] The invention relates to a video target retrieval technology, in particular to a method and system for multi-target classification and retrieval of surveillance video based on deep learning. Background technique [0002] In order to create a safer living environment and meet the increasing security needs of the people, the state vigorously promotes the construction of safe cities, and its core foundation is a huge number of video surveillance equipment deployed in every corner of the city. In addition to the important application scenario of urban public security monitoring, as the most effective way in the security field, video surveillance equipment has been widely used in areas closely related to daily life such as airports, stations, banks, shops, and residential areas. Therefore, in recent years, the number of video surveillance devices has increased dramatically, and a large number of video surveillance devices are generating massive amounts of s...

AI Technical Summary

Problems solved by technology

[0003] Security monitoring video has problems such as complex scene environment, dense targets, and various resolutions. Especially when multi-target classification retrieval is performed, the classification and extraction of targets in surveillance video is more susceptible to factors such as background noise and target occlusion than general target extraction. Shallow features are also difficult to accurately express multiple types of targets with large differences in appearance using unified features. In recent years, the development of deep learning can just solve the above problems

Images