Positioning tracking error correction method and system for virtual scene

Abstract

The invention provides a positioning tracking error correction method and system for a virtual scene, and the method comprises the steps: employing two tracking cameras to scan a use environment of auser, obtaining a left-eye image and a right-eye image, and constructing a corresponding three-dimensional virtual scene according to a binocular stereo imaging principle; converting the world coordinate systems of the current frames of the two tracking cameras into a world coordinate system of a virtual scene corresponding to the current frames, and obtaining a conversion matrix; and through theconversion matrix and the 6DOF information of the current frame, obtaining the optimal position information of each three-dimensional feature point projected to the current frame under the world coordinate system of the virtual scene through an image matching method, thereby correcting the 6DOF positioning and tracking precision in real time. The method and system are high in positioning and tracking precision and low in power consumption.

Description

technical field [0001] The present invention relates to the field of human-computer interaction technology, and more specifically, to a positioning tracking error correction method and system for virtual scenes. Background technique [0002] The 6DoF (three-axis rotation and three-axis translation information) positioning and tracking accuracy of VR headsets is very important in the immersive experience of VR, and it is also one of the main parameters for evaluating the performance of VR headsets. [0003] At present, major VR head-mounted all-in-one machine manufacturers solve the positioning and tracking of VR head-mounted 6DoF mainly by building one, or two or more environment capture and tracking cameras in the VR head-mounted, and a built-in IMU 9-axis inertial navigation sensor. Using computer vision technology and multi-sensor fusion technology to solve the 6DoF information of VR head-mounted all-in-one machine in the three-dimensional environment space, the currently...

AI Technical Summary

Problems solved by technology

There are also some algorithmic optimizations, such as improving the accuracy of the environmental feature extraction algorithm, and improving the optimization algorithm for solving the camera movement trajectory and the camera’s spatial attitude information. The camera movement trajectory is routinely used through the Kalman filter model algorithm. And attitude information estimation, now using the extended Kalman filter model algorithm to try to optimize some methods such as camera movement trajectory and attitude information estimation accuracy; but in some usage scenarios, the tracking accuracy of the VR head-mounted all-in-one machine with 6DOF is still unsatisfactory , especially in home use scenarios, due to various factors such as home furniture and room lighting, the 6DoF tracking accuracy cannot meet the needs of most home use scenarios

Moreover, if the number of tracking cameras is increased to more than two, it will pose many challenges to the power consumption, heat dissipation, cost and structural design of the overall system.

Images