Guideboard detection method and device based on perspective distortion correction

Abstract

The embodiment of the invention discloses a guideboard detection method and device based on perspective distortion correction. By applying the scheme provided by the embodiment of the invention, perspective correction can be carried out on a guideboard image based on the detection results of the four corners of the guideboard image, so that the accuracy of the guideboard detection result can be improved when guideboard detection is carried out on the corrected guideboard image. Moreover, the detection result can be compared with a guideboard dictionary library to verify whether the detection result is accurate or not. When the guideboard detection result is inaccurate due to failure of corner detection, perspective distortion correction can be performed on the guideboard image based on a pre-trained space converter, so that the accuracy of the guideboard detection result is further improved.

Description

technical field [0001] The present invention relates to the technical field of intelligent driving, in particular to a road sign detection method and device based on perspective distortion correction. Background technique [0002] In the field of autonomous driving, perception algorithms need to detect street signs on the road to control unmanned vehicles to drive safely according to the detection results. For example, the camera installed in the unmanned vehicle can collect the image in front of the vehicle, and then perform character detection on the street sign area in the image to obtain the detection result. [0003] However, in practical applications, the images collected by the camera usually have perspective distortion, which will reduce the detection accuracy of characters in street signs. Therefore, in order to improve the accuracy of street sign detection, there is an urgent need for a street sign detection method. Contents of the invention [0004] The invent...

AI Technical Summary

Problems solved by technology

[0003] However, in practical applications, the images collected by the camera usually have perspective distortion, which will reduce the detection accuracy of characters in street signs.

Images