Road end camera installation deviation angle calibration method, device and system

Abstract

The invention provides a road end camera installation deviation angle calibration method, device and system. The method is applied to a vehicle-mounted processor in an automatic driving vehicle, and the vehicle-mounted processor is used for carrying out installation deviation angle calibration on a monocular camera installed on a road rod; and an integrated navigation device and a laser radar arearranged in the automatic driving vehicle. The method comprises the following steps of acquiring point cloud data acquired by a laser radar, obtaining a transformation relationship between a world coordinate system provided by integrated navigation equipment and a vehicle coordinate system, acquiring image data acquired by a monocular camera, analyzing and calculating according to the point clouddata, the transformation relationship and the image data to obtain a current estimated pose transformation matrix of the monocular camera, and according to the current estimated pose transformation matrix and the initial estimated pose transformation matrix of the monocular camera, calculating an error pose transformation matrix corresponding to the monocular camera so as to complete installationangle deviation angle calibration of the road-end camera. The self-calibration process of the installation deviation angle of the road end camera can be completed through the vehicle-mounted processor.

Description

technical field [0001] The present application relates to the technical field of camera calibration, in particular to a method, device and system for calibrating installation deviation angles of roadside cameras. Background technique [0002] In L4 and above level of automatic driving, vehicle-road coordination is an essential part. In vehicle-road coordination, the subgrade perception system provides roadside information and environmental information for the decision-making system, and can also assist vehicle positioning when vehicle positioning fails. It is an important support for the automatic driving system. The sensor composition schemes in the roadbed perception system usually include monocular camera scheme, laser plus monocular camera scheme, binocular camera scheme, etc. The monocular camera scheme is widely used because of its low cost and simple structure. [0003] The monocular camera is usually installed on the roadside pole with a boom. It has two degrees of ...

AI Technical Summary

Problems solved by technology

During installation, the pitch and heading will be calibrated to obtain the installation deviation angle. However, after a long period of environmental influence, the installation deviation angle will change, making the data obtained by the camera no longer accurate, and cannot be used for decision-making and positioning systems. Provide accurate reference and even cause wrong decision

[0004] The traditional method is to use a manual hand-held calibration plate to provide a reference for the correction of the installation deviation angle on the pre-determined coordinates. However, there may be more than 100,000 cameras on the roadbed on the expressway, and it is a huge cost to calibrate all the cameras manually.

Images