A method and device for processing target detection data

Abstract

The invention discloses a method and device for processing target detection data. Specifically, the method includes: when multiple sensors are detected to simultaneously detect the target, sequentially process the detection process of multiple sensors The processing method of the current sensor detection process among multiple sensors includes: correlating the detection information of the current sensor with the track output by the previous sensor to obtain associated information; Updating, track management is performed on the updated track information; the current sensor is the last sensor used for detection among multiple sensors, and the final processing data of the target is output after the detection process of the last sensor is completed. The detection data of multiple sensors for the target is sequentially fused, processed sequentially and in batches, and the computing efficiency is improved.

Description

technical field [0001] The invention relates to the technical field of target detection and tracking, in particular to a method and device for processing target detection data. Background technique [0002] Self-driving or unmanned vehicles are attracting more and more people's attention. Major automakers and technology and Internet companies are actively investing a lot of manpower and material resources in the research and development of unmanned driving systems, hoping to realize the mass production of unmanned vehicles as soon as possible. The advanced automatic driving system generally consists of six subsystems: sensor system, perception system, positioning system, planning and decision-making system, vehicle control system and system management system. The sensor system is the hardware basis for vehicle perception and positioning. Various environmental raw information measured by the system (such as distance, angle and speed of moving vehicles, distance, angle and cat...

AI Technical Summary

Problems solved by technology

[0005] At present, some target tracking and fusion algorithms on the car are designed for a specific number or type of sensors, which is not conducive to the selection of on-board sensors. The fusion of the front monocular camera is realized, and the design of the corresponding tracking and fusion algorithm is aimed at the fusion of these two sensors. If more functions need to be implemented now and the side front sensor is added, the corresponding algorithm needs to be redesigned, which is not conducive to saving Development costs

Images