Environmental object tracking method and device based on visual-inertial odometry

Abstract

The present application relates to a visual-inertial odometry-based environmental object tracking method, device, computer equipment and storage medium. The method includes: acquiring an environmental object image, and extracting object point features and object line features from the environmental object image; counting the number of point features, and counting the number of line features; determining the point feature according to the number of point features Weights and line feature weights; according to the point feature weights and the line feature weights, weighted adjustments are made to the number of point features and the number of line features to obtain weighted point feature numbers and weighted line feature numbers; Extracting target point features conforming to the number of weighted point features from the object point features, and extracting target line features conforming to the number of weighted line features from the object line features; according to the target point features and the target line feature to track environmental objects. The method can improve the accuracy of motion tracking.

Description

technical field [0001] The present application relates to the technical field of motion tracking, and in particular to an environmental object tracking method, device, computer equipment and storage medium. Background technique [0002] Visual-inertial odometry (visual-inertial odometry, VIO) is an algorithm that fuses visual data and inertial measurement unit (Inertial measurement unit, IMU) data to achieve simultaneous localization and mapping (SLAM). While visual-inertial odometry uses visual data for motion tracking, it fuses inertial measurement unit data to improve the performance of motion tracking. It is a low-cost, high-performance motion tracking algorithm. Therefore, visual-inertial odometry is often used in the fields of automobile automatic driving, virtual augmentation technology, augmented reality technology and robot navigation. [0003] However, the existing visual-inertial odometry provides environmental object tracking methods that only use point features...

AI Technical Summary

Problems solved by technology

[0003] However, the existing visual-inertial odometry provides environmental object tracking methods that only use point features in the visual image when using visual data for motion tracking. The tracking effect is not good; at the same time, when the number of point features is large, the calculation burden of the system will be large, which will also affect the accuracy of motion tracking

[0004] Therefore, the existing environmental object tracking methods have the problem of low tracking accuracy

Images