Construction method and device applied to test scene of pilotless automobile

Abstract

The invention discloses a construction method and device applied to the test scene of a pilotless automobile. One specific implementation way of the method comprises the following steps: obtaining a road condition image which contains scene objects, wherein the scene objects comprise a road object, a traffic sign object, a vehicle object and a pedestrian object; on the basis of the road situation image, obtaining the attribute information of the scene object; and on the basis of the attribute information of the scene object, constructing the test scene of the pilotless automobile. The attributes of the real road object, traffic sign object, vehicle object and pedestrian object are used for constructing the test scene of the pilotless automobile so as to reduce the real road situation, and the real road situation is used as the test scene of the pilotless automobile so as to improve the test accuracy of the pilotless automobile.

Description

technical field [0001] The present application relates to the field of artificial intelligence, specifically to the field of unmanned vehicles, and in particular to a method and device for constructing test scenarios applied to unmanned vehicles. Background technique [0002] Vehicle testing is an important means to improve vehicle safety. When testing the vehicle, it is necessary to set up a test scene including roads, vehicles, pedestrians and other elements, and then simulate the test scene to complete the test of the vehicle. At present, when setting a test scene, a common method is to manually set attributes such as position and speed of each element in the scene. [0003] However, when using the above method to set up the test scene, on the one hand, it is necessary to set all possible values ​​of the element for the attributes of each element, resulting in huge overhead. On the other hand, the error caused by the manual setting of the attributes of any element and t...

AI Technical Summary

Problems solved by technology

[0003] However, when using the above method to set up the test scene, on the one hand, it is necessary to set all possible values ​​of the element for the attributes of each element, resulting in huge overhead

On the other hand, the error caused by the manual setting of the attributes of any element and the actual situation will further increase the error of the entire test scene, resulting in distortion of the test scene, thereby reducing the accuracy of the vehicle test

Images