Traffic light identification method and device

Abstract

The invention provides a traffic light identification method and device. The method comprises that positioning information of a vehicle is obtained; position information of a traffic lamp set in a target crossing in map data created in advance is obtained according to the positioning information of the vehicle; coordination transformation is carried out on the position information of the traffic lamp set to determine an ROI (Region of Interest), included by an image captured by a target camera of the vehicle, of the traffic lamp set; and the ROI is analyzed to obtain a lightning state of the traffic lamp set, attribute information of a traffic lamp which is ON at present can be inquired in the map data according to the lightning state of the traffic lamp set, and the attribute informationis used to indicate the passing state of the target crossing. Thus, the traffic lamps can be identified rapidly and accurately.

Description

technical field [0001] The invention relates to the technical field of intelligent transportation, in particular to a traffic light identification method and device. Background technique [0002] With the advancement and development of science and technology, intelligent driving technology has enabled cars to basically have automatic operation and driving capabilities. Intelligent driving technology mainly uses on-board sensors (such as cameras, lidar, etc.) Road, vehicle position, traffic signs and obstacles and other information, control the steering and speed of the vehicle, among which timely and accurate identification of traffic lights is an important prerequisite for the vehicle to be able to drive safely and reliably on the road. [0003] At present, the scheme of intelligent driving technology to identify traffic lights is usually: use the car camera to obtain video images containing traffic lights at intersections, and then detect and identify attributes such as th...

AI Technical Summary

Problems solved by technology

However, on the one hand, when detecting traffic lights, a large area or even the entire video image is usually intercepted, which will result in a huge amount of calculation and low detection efficiency; on the other hand, due to the The resolution is limited, especially when the external environment is relatively poor, for example, in rainy, snowy, foggy weather or at night, which will cause the detected traffic light to appear as a bright spot, and it is difficult to identify its attributes such as color and shape

Images