Typical natural driving scene recognition and extraction method for intelligent driving system testing

Abstract

The invention discloses a typical natural driving scene recognition and extraction method for intelligent driving system testing. The method is realized based on a system comprising a data extractionand calculation module and a typical scene type recognition module. The data extraction and calculation module obtains driving scene key parameters from a vehicle driving database through screening, matching and calculation; and the typical scene type recognition module identifies typical scenes according to the provided driving scene key parameters and the driving characteristics of the main vehicle and the relative state of the main vehicle and the target object, and the typical scenes comprise a dangerous scene, a main vehicle lane changing scene, a vehicle following driving scene, an adjacent vehicle switching-in scene, a front vehicle switching-out scene and a line patrol driving scene. According to the method, the typical natural driving scene is comprehensively extracted, offline application to the typical natural driving scene is achieved, the structure is complete, and the applicability is high.

Description

technical field [0001] The invention relates to the application field of testing, analysis and evaluation of an automobile intelligent driving system, in particular to a method for identifying and extracting a typical natural driving scene for intelligent driving system testing. Background technique [0002] In recent years, with the advancement of science and technology, the degree of intelligence of automobiles has become higher and higher, and L2 and L3 intelligent driving systems have been gradually introduced to the market. Before a car enters the market, it must be fully tested and verified to ensure that the intelligent control system can operate safely within its design operating domain. [0003] Natural driving scenarios refer to the real traffic conditions that drivers encounter during actual road driving. The research on typical natural driving scenarios is of great significance to the development of intelligent driving systems, the compilation of test cases and t...

AI Technical Summary

Problems solved by technology

[0005] The present invention aims at the problem of the singleness of natural driving scene extraction in the prior art, and proposes a typical natural driving scene recognition and extraction method for intelligent driving system testing. This method is a comprehensive extraction of typical natural driving scenes, and its purpose is Realize offline standardized storage and application of typical natural driving scenes, and constitute a basic tool for natural driving scene research. The scene has a complete structure and strong applicability

Images