Multi-focal-length stereoscopic vision simultaneous positioning and mapping implementation method and system

Abstract

The invention provides a multi-focal-length stereoscopic vision simultaneous positioning and mapping implementation method and system, and belongs to the technical field of automatic driving, automobile environment perception and intelligent cabins. The method comprises the following steps: 1, installing a multi-focal-length stereo camera, and calibrating a projection matrix Pl of the multi-focal-length stereo camera and a projection matrix Pr of a right camera by using an improved multi-focal-length camera calibration method; 2, extracting the ORB feature points of the image based on the image pyramid, and increasing the number of extracted ORB feature points for the part, with the same view as the long-focus camera image, in the short-focus camera image; and 3, stereoscopically correcting the ORB feature points by using the multi-focal-length calibration parameters, stereoscopically matching the ORB feature points to obtain the three-dimensional positions of the ORB feature points, and finally solving the pose of the ORB feature points. According to the invention, the positioning precision same as that of a short-focus standard three-dimensional SLAM can be realized, and long-distance scene mapping can be realized.

Description

technical field [0001] The invention belongs to automatic driving and smart car environment map construction and positioning, and specifically designs a multi-focus stereoscopic SLAM and its implementation method. Background technique [0002] Simultaneous localization and mapping (SLAM) has been a research hotspot in the field of computer vision and robotics in the past two decades. It can construct an unknown environment map in real time and locate its own position. Visual SLAM has become a hot field of SLAM research because of its high resolution, rich environmental image information, simplicity and portability, low hardware cost, high positioning accuracy, and no need for environmental prior information. Visual SLAM technology is the basis and key module of various upper-level applications such as robots, augmented reality, and autonomous vehicles. It plays an important role in vehicle positioning and navigation, and is widely used in the fields of automotive automatic d...

AI Technical Summary

Problems solved by technology

In the field of automotive autonomous driving, it is more necessary to use the camera to perceive the surrounding environment. Only using a short focal length camera cannot obtain long-distance scene information, and cannot detect and locate objects in long-distance scenes.

Images