Trajectory tracking and three-dimensional reconstruction method, system and device

Abstract

The embodiment of the invention discloses a trajectory tracking and three-dimensional reconstruction method, system and device, relates to the technical field of image processing, and solves the problems of poor trajectory tracking stability and low precision in the prior art and low accuracy of a three-dimensional scene reconstruction effect in the prior art. According to the specific scheme, themethod comprises the steps of using a trajectory tracking and three-dimensional reconstruction device for extracting a first view image frame corner point and a second view image frame corner point of a target acquired by a binocular camera; generating a three-dimensional point cloud and a key frame; and tracking the front and rear frames of images by using a multi-plane fitting-based mode to continuously update the three-dimensional point cloud and the key frame set, solving a camera pose, obtaining a dense depth map according to the camera pose and the three-dimensional point cloud, performing confidence judgment on the depth map, and finally obtaining a three-dimensional reconstruction model of the scene. The embodiment of the invention is used in the process of image tracking and three-dimensional scene reconstruction.

Description

technical field [0001] The embodiments of the present application relate to the technical field of image information processing, and in particular to a method, system and device for trajectory tracking and 3D reconstruction based on a binocular camera. Background technique [0002] At present, with the universal application of mobile smart devices, a large number of applications based on virtual reality and augmented reality are also pouring into the market. Various fields such as education and navigation; as well as the development of robots in the current society and extensive research on unmanned vehicles, high-precision real-time positioning and map construction (Simultaneous Localization And Mapping, SLAM) and three-dimensional reconstruction have become the research fields of many researchers. hotspots. [0003] The existing SLAM and 3D reconstruction methods are mainly based on a personal computer (PC) or a monocular camera device. When performing target tracking, th...

AI Technical Summary

Problems solved by technology

However, since the target tracking method is based on a fixed geometric model, its tracking accuracy is low; in addition, the BoW method is used for relocation, because it mainly calculates and matches the BoW of the key frame and the current frame, Therefore, this method is easy to be lost in the area where the feature is not rich, and the relocation effect is not ideal. In addition, using this method needs to load the data dictionary in advance, and the data dictionary is usually relatively large, so the initial loading time is relatively long and takes up memory.

Images