Intelligent bee colony clustering method and a vehicle target detection method

Abstract

The invention belongs to the field of vehicle detection and recognition, and particularly relates to an intelligent bee colony clustering method and a vehicle target detection method. The invention relates to an intelligent bee colony clustering method, which mainly comprises the following steps of: 1) establishing a common target of a bee colony, namely searching a plurality of high-quality honeysources; 2) making a work plan; 3) establishing an assessment system; And 4) outputting a clustering center result. According to the intelligent bee colony clustering method, the bee colony algorithmand the clustering model are effectively combined, and a team management mode is integrated, so that a good clustering effect can be achieved, the calculation amount of subsequent similarity measurement is reduced, and the efficiency and the precision are improved. According to the vehicle target detection method, the clustering center is obtained after vehicle characteristics are subjected to efficient clustering analysis, the optimized bee colony algorithm is applied to the clustering analysis by fusing the thought of team management, the working mode that bee members in a team find high-quality honey sources is imitated, and the purpose of rapid clustering is achieved.

Description

technical field [0001] The invention belongs to the field of vehicle detection and identification, and in particular relates to an intelligent bee colony clustering method and a vehicle target detection method. Background technique [0002] As the connection between road traffic and social and economic development becomes closer, how to efficiently and correctly analyze and process traffic image data and detect specific vehicle targets under the background of massive images has profound implications for all aspects of society and economy. Significance. The vehicle target detection technology is to detect similar or identical specific vehicle targets from large-scale images based on the existing vehicle target feature information, and then analyze and speculate the specific vehicle target driving conditions and driving routes. Management, criminal investigation, vehicle monitoring, etc. provide favorable technical support. [0003] At present, the common vehicle target dete...

AI Technical Summary

Problems solved by technology

[0003] At present, the common vehicle target detection method is to extract the features in the image for matching, and perform feature fusion to improve the detection accuracy, but for the continuous growth of the number of features, the accuracy of vehicle target detection needs to be improved

Images