Multi-view vehicle association method and related device

Abstract

The invention discloses a multi-view vehicle association method and a related device, and the method comprises the steps: obtaining at least two to-be-assisted images collected by a main camera deviceand at least one slave camera device, and enabling the main camera device and the at least one slave camera device to collect the same position from different views; extracting a first to-be-matchedvehicle from the to-be-processed images collected by the main camera device, and extracting at least one second to-be-matched vehicle from the to-be-processed images collected by the at least one slave camera device; determining the position similarity between the first to-be-matched vehicle and each second to-be-matched vehicle, and determining the feature similarity between the first to-be-matched vehicle and each second to-be-matched vehicle; according to the position similarity and the feature similarity, determining the final similarity between the first to-be-matched vehicle and each second to-be-matched vehicle; and associating the second to-be-matched vehicle corresponding to the final similarity meeting the preset requirement in all the final similarities with the first to-be-matched vehicle, thereby accurately obtaining the panoramic image of the vehicle.

Description

technical field [0001] The present application relates to the technical field of intelligent transportation, in particular to a multi-view vehicle association method and related devices. Background technique [0002] Image processing technology based on video surveillance has been widely used in the field of intelligent transportation. With the support of advanced hardware equipment, relevant departments can use computer vision and other technologies to process and analyze surveillance video images, which can greatly save labor costs while improving vehicle surveillance. [0003] In intelligent traffic scenarios, commonly used monitoring cameras such as electronic police, checkpoints, and entrances and exits can only obtain a single close-up image of the front or back of the controlled vehicle. In the illegal management and control tasks of some special vehicles, it is often necessary to associate the snapshots of the same vehicle in cameras with different installation angl...

AI Technical Summary

Problems solved by technology

The association of snapshots in the current technology often cannot be accurately matched due to inconsistent recognition of the license plate numbers of the front and rear of the car, unlicensed vehicles, etc., resulting in many "waste films", which will aggravate manual review. burden

Images