Scene self-adaptive system for improving vehicle-mounted infrared thermal imaging recognition accuracy

Abstract

The invention discloses a scene self-adaptive system for improving vehicle-mounted infrared thermal imaging recognition accuracy, which belongs to the technical field of vehicle-mounted infrared thermal imaging, and comprises a system starting module, a vehicle speed acquisition module, a parameter mapping module and a target recognition algorithm module, wherein the output of the system startingmodule is electrically connected with the vehicle speed acquisition module; the output end of the vehicle speed acquisition module is electrically connected with a GPS module and an OBD interface, andthe output end of the GPS module and the output end of the OBD interface are both connected with a parameter mapping module. According to the system, in a high-speed environment, the dangerous targetidentification distance and the long-distance dangerous target identification accuracy can be greatly improved, and the driving safety is ensured; in a low-speed environment, the vehicle speed is lower than a threshold value, the vehicle belongs to an urban area or a road section with complex road conditions, a long-distance dangerous target is insufficient to cause driving safety and priority detection accuracy, and the interference of target false identification and false early warning on a driver is greatly reduced.

Description

technical field [0001] The invention belongs to the technical field of vehicle-mounted infrared thermal imaging, and specifically relates to a scene adaptive system for improving recognition accuracy of vehicle-mounted infrared thermal imaging. Background technique [0002] The target detection algorithm module of the vehicle-mounted infrared thermal imaging night vision system can perform target detection and recognition on the data collected by the night vision sensor, and output the target detection results after calculation through the given hardware resources. For target detection and recognition, it is inevitable that the target will be misrecognized and falsely warned, causing driver fatigue; for the recognition accuracy of long-distance targets is not enough, the performance of recognition and warning in high-speed mode can be improved; because the module only provides a single recognition performance, it is impossible to perform differentiated automatic performance ...

AI Technical Summary

Problems solved by technology

[0002] The target detection algorithm module of the vehicle-mounted infrared thermal imaging night vision system can perform target detection and recognition on the data collected by the night vision sensor, and output the target detection results after calculation through the given hardware resources. For target detection and recognition, it is inevitable that the target will be misrecognized and falsely warned, causing driver fatigue; for the recognition accuracy of long-distance targets is not enough, the performance of recognition and warning in high-speed mode can be improved; because the module only provides a single recognition performance, it is impossible to perform different automatic performance adjustments in combination with actual application scenarios

Images