Calibration method and device for distance measurement device and camera fusion system

Abstract

The invention is suitable for the technical field of distance measurement, and provides an online calibration method for a distance measurement device and a camera fusion system. The method comprises the following steps: controlling a distance measurement device to project spot beams to a target scene and collect the spot beams to obtain a first target image, and synchronously controlling a camera to collect the target scene to obtain a second target image; acquiring three-dimensional coordinate information of a light spot corresponding to the spot light beam in a world coordinate system according to the first target image; obtaining two-dimensional coordinate information of a light spot corresponding to the spot light beam in a pixel coordinate system according to the second target image; determining a plurality of initial point pairs according to a preset projection rule; obtaining the matching confidence of each initial point pair; setting the initial point pair of which the matching confidence meets a preset condition as a target point pair, and calculating an external parameter between the distance measurement device and the camera according to the target point pair. According to the invention, real-time and high-precision online calibration is realized, and the robustness and adaptability of a self-calibration scheme are improved.

Description

technical field [0001] The present application belongs to the technical field of distance measurement, and in particular relates to a calibration method and device for a distance measurement device and a camera fusion system. Background technique [0002] For the current mature intelligent perception solutions, especially L4 and above automatic driving systems, the requirements for perception capabilities are diverse, and it is necessary to use multi-sensor fusion such as distance measurement devices and cameras to achieve accurate, real-time, comprehensive and reliable perception of vehicles and environments. Wherein, the distance measuring device includes a depth camera or LiDAR (Light Detection and Ranging) based on the time-of-flight principle, for acquiring three-dimensional data of the target. In the distance measurement device and camera fusion system, the camera can provide rich visual texture information to make up for the shortcomings in recognition and cognition c...

AI Technical Summary

Problems solved by technology

Usually, the distance measuring device and the camera fusion system will be calibrated before leaving the factory. However, due to the influence of factors such as temperature, mechanical collision, and aging during the actual use of the product, the internal and external parameters of the system will change, resulting in a decrease in the calibration accuracy of the system. , resulting in "misalignment" between multiple sensors, which will have a very obvious impact on the accuracy of subsequent applications such as perception and 3D reconstruction. At this time, online calibration is required to recalibrate the internal and external parameters.

However, the current online calibration algorithm is relatively poor in both accuracy and robustness. In this case, it can only be returned to the factory for offline calibration. The current offline calibration algorithm has high requirements for scenarios and targets, and the actual operation is more complicated. , the labor cost and time cost required for the whole process are very high

Images