Structured scene vision SLAM method based on improved dot-line features

Abstract

The invention discloses a structured scene vision SLAM method based on improved dot-line features, and the method comprises the steps: firstly carrying out the basic calibration of an RGB-D camera, obtaining the internal reference information, and then carrying out the SLAM initialization of a structured scene through a depth camera; extracting ORB point features and LSD line features in the structured scene, establishing an error model according to spatial points and spatial straight lines corresponding to the point-line features, estimating the pose of the camera by minimizing the model, andgenerating three-dimensional map points of the structured scene; deciding to generate a key frame in the video frame, establishing a bag-of-words model by using the key frame set, and carrying out closed-loop detection on the three-dimensional map points; and after a closed-loop condition is detected, optimizing the camera pose and the structured scene three-dimensional map point through an errormodel, and improving the SLAM effect. According to the invention, the problem that the visual SLAM is not high in closed-loop detection precision and efficiency in a structured scene is solved, and great convenience is provided for visual SLAM work.

Description

technical field [0001] The invention belongs to the technical field of machine vision, and in particular relates to a structured scene vision SLAM method based on improved point-line features. Background technique [0002] With the rapid development of the Internet economy, industries such as logistics, express delivery and mechanical processing and production are developing rapidly. Due to the large workload, single work tasks and high work accuracy requirements, industrial robots are often selected to complete specific tasks. In the image measurement process of robots and the field of machine vision, there is an increasing demand for positioning and mapping in structured scenes. [0003] At the same time, the machine vision technology based on visual synchronous positioning and mapping (visual SLAM) is becoming more and more perfect, and the processing speed, positioning, and mapping accuracy have made great progress compared with traditional methods. However, at present...

AI Technical Summary

Problems solved by technology

[0004] Purpose of the invention: In order to solve the problem of low efficiency of simultaneous positioning and mapping of robots and low precision due to the influence of environmental factors, the present invention provides a structured scene visual SLAM method based on improved point-line features, which can be used in some types of environments Simultaneous positioning and mapping provide great convenience

Images