AUV (Autonomous Underwater Vehicle) dynamic integrated positioning method based on visual fusion

Abstract

The invention relates to an AUV (Autonomous Underwater Vehicle) dynamic integrated positioning method based on visual fusion, which comprises the following steps: firstly, in an underwater environment, an integrated AUV acquires a marker image at the rear part of a target AUV by using two monocular cameras in a binocular vision system mounted at the front part of the integrated AUV, and the marker image is divided into a left image and a right image; according to the left image and the right image, thesize of the marker is acquired by resolving a binocular vision algorithm to obtain the size of the marker; in the underwater docking process, the integrated AUV utilizes two monocular cameras in a binocular vision system installed on the front portion of the integrated AUV to obtain images of markers on the rear portion of the target AUV, and according to the sizes of the markers obtained through calculation, two sets of positioning results are obtained through calculation by utilizing a monocular vision positioning algorithm; and weighted fusion is performed on the obtained two groups of positioning results to obtain a final AUV dynamic integrated positioning result. The single-purpose precision is ensured, the problem of high loss rate of binocular vision is solved, and relatively accurate positioning information is provided for dynamic docking control.

Description

technical field [0001] The invention relates to the technical field of AUV dynamic docking, in particular to a visual fusion-based AUV dynamic docking positioning method. Background technique [0002] In the survey of underwater terrain and underwater mechanical operations, AUV is an important carrier platform, and a single AUV individual can realize the functions of detection and mechanical operation underwater. However, its outstanding disadvantage is that in the face of complex underwater environments, a single independent AUV is limited by its volume and length, so it cannot perform complex functions. The payload carried by small-volume AUVs is limited, and large-volume AUVs cannot pass through complex underwater spaces. The emergence of multiple AUVs solves this problem. AUVs with small volumes and different functions are connected in series to meet the diversity of functions and the mobility of operations. [0003] Docking technology is divided into static docking an...

AI Technical Summary

Problems solved by technology

In order to improve the accuracy and error tolerance rate of AUV's relative positioning information obtained through vision, especially the positioning accuracy in the process of dynamic docking, to solve the problem of noise interference in the process of relying on acoustics to obtain relative positioning, and to use monocular vision or binocular vision alone. To solve the problem of large error and small error tolerance rate, the present invention proposes a visual fusion-based AUV dynamic docking positioning method, which uses binocular vision to calculate the size of markers that can offset the system error in the underwater environment, and decomposes the binocular vision For two monocular visions, the two sets of relative positioning information calculated by the PNP algorithm are used for information weighted fusion, and the final relative positioning information is obtained through information fusion, which not only improves the visual coverage, but also improves the The accuracy and error tolerance rate of relative positioning information acquisition not only ensures the monocular accuracy but also solves the problem of high binocular vision loss rate

Images