Information extraction method of traffic signs based on fusion of laser point cloud and image data

Abstract

The invention relates to a method for extracting traffic sign information based on the fusion of laser point cloud and camera image data, comprising the following steps: step 1, preprocessing the original laser point cloud data, and extracting the point cloud of the traffic sign area; step 2, calculating Get the coordinate center point of each point cloud plane extracted in step 1, and then according to the recorded GPS positioning data when the vehicle-mounted CCD camera shoots, obtain all CCD photo sequence sets within the coordinate range of the center point; step 3, to the CCD photo Register with the point cloud of the traffic sign area; step 4, analyze the CCD photo, including color segmentation, shape detection, feature selection, recognition and classification, so as to realize the accurate extraction of traffic sign information. The invention makes full use of the advantages of point cloud data and image data acquired by the mobile measuring vehicle, and can more accurately and quickly determine the position of the traffic sign.

Description

technical field [0001] The method belongs to the technical field of laser remote sensing and image processing, and involves fusion processing of two data sources of laser point cloud and CCD photo, which can effectively avoid the interference of surrounding scenery on detection, and greatly improve the accuracy of street sign information extraction. Background technique [0002] Nowadays, high-precision map production has become a current research hotspot. So how to accurately and efficiently obtain the information of road signs on both sides of the road is particularly important. Various attributes of the current lane can be obtained from traffic signs, such as the number of lanes, lane steering information, speed limit information, etc. However, most of the traditional methods are only for image processing, image analysis and pattern recognition to obtain the navigation information in the street signs, but if the color of the surrounding scenery is consistent with the colo...

AI Technical Summary

Problems solved by technology

When the color segmentation of the image extracts the interesting surface of the traffic sign, it is greatly affected by the light. Jobson et al. use the Retinex algorithm to eliminate the light influence, but the effect is general

In addition, the interference of the color of the surrounding scenery and the change of the color of the street signs will make it difficult to locate the area of ​​interest, which will interfere with the later processing; moreover, this color segmentation method needs to scan the whole image, which cannot reduce the non-street signs. scanning, the calculation time consumption is large, and the operation efficiency is low

Images