A method and system for evaluating relative accuracy confidence of vehicle-end monocular vision measurement

Abstract

The invention relates to a method and system for evaluating the relative accuracy of vehicle-end monocular vision measurement. The method includes: initializing a covariance matrix of a monocular vision system according to prior knowledge, and the covariance matrix includes: feature points between two frames of images Matching pixel error covariance matrix C p ; according to the covariance matrix C p and the pose covariance C(X) of the car body relative to the reference point at the current moment k ) to obtain the position covariance C of the target for position estimation under relative mapping r ; According to the distance between the two targets and the corresponding position covariance C r Calculate the relative position covariance matrix between two targets, and calculate the relative distance confidence under the allowable measurement error according to the relative position covariance matrix; on the one hand, solve the problem of target positioning error calculation caused by various error sources in relative construction , to guide the improvement of the accuracy and reliability of the navigation map; on the other hand, it solves the problem of weight estimation of relative mapping information in the process of multi-sensor data fusion, thereby improving the accuracy of multi-sensor fusion.

Description

technical field [0001] The invention relates to the field of computer vision positioning and mapping, in particular to a method and system for evaluating relative accuracy confidence of vehicle-end monocular vision measurement. Background technique [0002] The vehicle-side monocular vision relative mapping refers to the use of the monocular vision sensor installed on the vehicle-side, combined with the scale information provided by the IMU (Inertial Measurement Unit, inertial measurement unit) and the body CAN during the vehicle's forward process, to restore the vehicle's surrounding environment map in real time. the process of. [0003] Compared with the traditional environment mapping that relies on high-precision GNSS sensors, relative mapping has the following two significant advantages: (1) It does not rely on the absolute position and attitude of the vehicle, which improves the flexibility and reliability of mapping. During the driving process, it is often difficult ...

AI Technical Summary

Problems solved by technology

[0004] However, due to the process error of the IMU sensor and the random noise introduced in the measurement process, there is also a certain error in the relative pose of the vehicle.

The navigation map needs to fuse the relative map with the absolute fixed position information, and the accuracy and reliability of the navigation map will be affected in the case of relative pose errors. evaluation method

Images