Real-time identification system and method of road surface textures for unmanned vehicle

Abstract

The invention provides a real-time identification system and method of road surface textures for an unmanned vehicle. The system is composed of a laser triangulation subsystem, a three-dimensional structure depth detection subsystem and a road detection center subsystem; and control and data transmission among the three subsystems are completed by an industrial control computer. According to the invention, with the laser triangulation technology, the asphalt pavement texture image is collected and thus road surface textures and data are collected automatically during the driving process of theunmanned vehicle; the three-dimensional structure depth detection subsystem analyzes and processes a collected pavement surface elevation and calculates a construction depth of an asphalt pavement texture of a measured point or a driving section. Therefore, the influence on the traffic flow by the traditional road detection is eliminated; and a defect of large impact of the human factor accordingto the traditional structural depth measurement method is overcome. The intelligent level is improved; the efficiency is enhanced. The real-time identification system and method has the guiding significance in driving strategy of unmanned vehicles.

Description

technical field [0001] The invention belongs to the field of unmanned vehicles, and in particular relates to a road texture recognition system and method for unmanned vehicles. Background technique [0002] Unmanned vehicles are the development direction of future vehicles. Compared with traditional vehicles, they have the advantages of reducing traffic accidents and relieving traffic pressure. At this stage, the research on environment perception of unmanned driving mainly stays in the field of computer AI and automatic control, while the research on road surface texture recognition in the road field is very little, ignoring the fact that unmanned vehicles are still driving on the road. This key point. In addition, the current pavement texture recognition is mainly carried out in the acceptance link of road construction, mainly based on the structural depth as the index. Traditional structural depth detection methods are mostly manual fixed-point detection and sampling de...

AI Technical Summary

Problems solved by technology

Traditional structural depth detection methods are mostly manual fixed-point detection and sampling detection. There are many shortcomings in both on-site measurement and data analysis: manual detection is inefficient, requires experienced professional and technical personnel to conduct detection on foot, and is difficult to apply to pavement structure depth detection on a large scale Medium; the test results are greatly affected by human factors, the accuracy of the test results is directly determined by the professional level of the technicians, and there are certain deviations in the test results between different technicians; the manual test takes the form of sampling, and generally several tests per kilometer of road surface parameters, the road cannot be fully detected

Images