A registration method of event triggered camera and three-dimensional radar

Abstract

The present invention relates to the technical field of image processing, point cloud processing and sensor data registration, and more particularly, to a registration method of an event triggered camera and a three-dimensional laser radar. 1, designing a calibration object suitable for registration of an event triggering camera and a three-dimensional laser radar; 2, simultaneously starting an event triggering camera and a three-dimensional laser radar to obtain data of the two sensors at the same time; 3, using image processing method such as edge extraction and specific pattern recognitionto locate a point of a calibration object in that image; 4, locating the same point of the calibration object in the point cloud by using the method of point cloud segmentation based on RANSAC; 5, calculating a transformation matrix on six degrees of freedom in the space according to the results obtained in the step 3 and the step 4; Step 6: the registration result is evaluated by the registrationerror and the edge-based value function proposed by the invention.

Description

technical field [0001] The present invention relates to the technical field of image processing, point cloud processing, and sensor data registration, and more specifically, to a registration method between an event-triggered camera and a three-dimensional laser radar. Background technique [0002] Computer vision algorithms based on ordinary RGB cameras have been gradually improved and have been applied in areas such as autonomous driving and object recognition. Lidar collects 3D spatial information of the surrounding environment through laser beams, and plays an irreplaceable and important role in the sensor solution for autonomous driving. [0003] A vision solution with only a camera will be affected by ambient lighting conditions, and a solution with only a lidar is powerless in recognizing plane environmental information such as road signs and traffic lights. Therefore, the fusion of cameras and lidar has become a key idea to solve the problem. Before fusion, what sh...

AI Technical Summary

Problems solved by technology

[0006] The disadvantage of the above-mentioned prior art is that there is no method that can use static markers suitable for event-triggered camera internal parameter calibration; the above-mentioned prior art is based on the registration developed by ordinary RGB cameras, and lacks a method suitable for Event-triggered markers for automatic registration of cameras and laser radars; in the above-mentioned prior art, the images and point clouds used for registration are multi-frame, and there is a lack of a method for using single-frame data to achieve registration; in the above-mentioned prior art , there is no better quantitative scheme for the evaluation of registration results, and it is impossible to judge the accuracy and error of registration

[0007] The reason for the above-mentioned disadvantages is that in the above-mentioned prior art, the applicable cameras are all common RGB color cameras, which can image static objects, but the event-triggered camera used in the present invention is suitable for static objects with no change in the ordinary surface light intensity. The object cannot be imaged, so the general static checkerboard calibration map cannot be used; in the above-mentioned prior art, the feature point correspondence method used relies on the ordinary camera for active imaging, and these methods cannot be matched with the point cloud for the event-triggered camera feature points; in the above-mentioned prior art, the feature extraction robustness of the markers used in the lidar point cloud and camera image is low, and it is necessary to extract multi-frame data obtained from different positions to improve the accuracy rate; the above-mentioned prior art However, the conversion results after registration are not fully used for quantitative error statistics

Images