Search

Research on Underwater Distance Measurement of Binocular Vision by Semantic Separation Picture

Author(s)

Sijia Ren*

Affiliation(s)

Leicester International Institute, Dalian University of Technology, Panjin, Liaoning, China *Corresponding Author.

Abstract

Aiming at the problem of large amount of calculation and slow speed of stereo matching in binocular vision ranging, this paper proposes an underwater distance measurement of binocular vision based on underwater target detection. By optimizing the SGBM algorithm, the disparity search range is limited in the underwater target recognition area, and an adaptive aggregation window is constructed to adapt to different texture changes. This method improves the matching accuracy and effectively reduces the computational burden of unrelated regions. At the same time, a texture similarity dynamic adjustment-matching window is established to flexibly respond to texture changes in different regions, thereby greatly improving the matching speed and accuracy. Experiments show that the ranging speed is increased by 26.8 %. This study can not only improve the efficiency and accuracy of underwater operations, but also lay a solid foundation for subsequent technical research and application promotion.

Keywords

Semantic Segmentation; Binocular Vision; Ranging; Underwater Imaging; Region Matching

References

[1]Ortiz A Simo M Oliver G. A vision system for an underwater cable tracker. Machine Vision and Applications, 2002 13(3):129-140. [2] Li Zhen. Design and research of automatic tracking ROV for underwater pipeline. Dalian University of Technology, 2018, 4. [3] Xu Peng-fei, Hu Zhen, Cui Wei-cheng, et al. Application of Vision-based System in Underwater Pipeline Active Inspection. ShipBuilding of China, 2010, 51(03):142-151. [4] Chen C, Nakajima M. A Study on Underwater Cable Automatic Recognition Using Hough Transformation. Proceedings of IAPR Workshop on Machine Vision Applications, Kawasaki, Japan.1994. [5] Zeng Wen-jing, Xu Yu-ru, Wan Lei, et al. g. Robotics Vision-Based System of Autonomous Underwater Vehicle for an Underwater Pipeline Tracker. Journal of S hanghai Jiaotong University, 2012, 46(02):178-183. [6] Li Shuangshuang. Research on AUV underwater pipeline detection method based on gradient information. Harbin Engineering University, 2016, 3. [7] Zingaretti P, Tascini G, Puliti P, et al. Imaging approach to real-time tracking of submarine pipeline. Proceedings of SPIE - The International Society for Optical Engineering, 1996, 2661:129-137. [8] Tang Xudong. Research on underwater pipeline detection and tracking technology of intelligent underwater vehicle. Harbin Engineering University, 2010, 12. [9] Balasuriya B, Takai M, Lam W, et al. Vision based autonomous underwater vehicle navigation: underwater cable tracking. OCEANSʾ97 MTS/IEEE Conference Proceedings, Halifax Cannada: Marine Technology Society 1997:1418-1424. [10] Cai Sijing, Wang Yanyu. Improved DDeepLabV3+ semantic segmentation network. Journal of Fujian University of Technology. 2024, 22 (01):95-102. [11] Cai Yan-xiang, Xie Hai-cheng, Yu Tian-qi, et al. Ranging model based on convergent binocular stereo vision. Laser & Infrared, 2023, 53 (08):1272-1278.