Urban management component intelligent identification method oriented to mobile background

Abstract

A city management component intelligent identification method oriented to a mobile background comprises the following steps: firstly, collecting sample pictures, then marking the sample pictures to form a sample picture library, and training the sample pictures in the sample picture library by adopting an SSD algorithm to construct an analysis model; the mobile video data collected by the vehicle-mounted multi-dimensional sensing terminal are analyzed, the analyzed mobile video data are input into the analysis model, the analysis model processes, analyzes and identifies the analyzed mobile video data and then outputs an identification result, and mobile law enforcement of city management is effectively achieved. Therefore, full time domain and full metropolitan area coverage of management is realized, urban mobile video resources are fully utilized, manpower is effectively liberated, the manpower demand is reduced, the manual operation intensity is reduced, and the fine operation level of urban management is improved; and meanwhile, social contradictions are relieved, and the intelligent level of urban management is improved.

Description

technical field [0001] The invention relates to the technical field of video intelligent recognition, in particular to a method for intelligent recognition of urban management components under a mobile background. Background technique [0002] At present, urban management and law enforcement mainly adopt the manual inspection of grid personnel and the operation mode of manual photography to collect information on violations / violations, which consumes a lot of manpower. Therefore, in order to save manpower and material costs, mainstream manufacturers have realized intelligent collection based on a fixed background (such as Skynet monitoring, etc.), but they are all intelligent identification and evidence collection modes for fixed video backgrounds, and most of them are based on the background difference method. For component recognition, when the background changes, the stability and recognition accuracy will be greatly reduced. At present, many street vendors in cities are...

AI Technical Summary

Problems solved by technology

[0002] At present, urban management and law enforcement mainly adopt the manual inspection of grid personnel and the operation mode of manual photography to collect information on violations / violations, which consumes a lot of manpower

Therefore, in order to save manpower and material costs, mainstream manufacturers have realized intelligent collection based on a fixed background (such as Skynet monitoring, etc.), but they are all intelligent identification and evidence collection modes for fixed video backgrounds, and most of them are based on the background difference method. For component recognition, when the background changes, the stability and recognition accuracy will be greatly reduced

At present, many street vendors in the city are mobile vendors, often hidden in areas that cannot be photographed by fixed cameras. Therefore, grid personnel still need to go to the streets to inspect, take pictures and collect evidence, resulting in low efficiency of urban management and law enforcement, and the degree of intelligence needs to be improved.

Images