{"title":"A bio-inspired approach to line segment detection utilizing orientation-selective neurons","authors":"Daipeng Yang , Bo Peng , Xi Wu","doi":"10.1016/j.sigpro.2025.109950","DOIUrl":null,"url":null,"abstract":"<div><div>Line segment detection is essential for tasks like SLAM, camera pose estimation, and 3D reconstruction. Although many excellent line segment detection methods have been proposed, detecting more true positives while rejecting false positives remains challenging. The human visual system can effectively perceive line segments in complex environments through a processing pathway involving multiple visual cortices. Inspired by this, we propose a novel bio-inspired line segment detection method that mimics the perception of line segments in the visual cortex. Our method models orientation-selective neurons in the primary and secondary visual cortices. Based on the preferred orientations of these neurons, we integrate them to mimic the function of orientation and curvature domains in the fourth visual cortex, generating continuous and smooth edge segments. A post-processing step, including least squares line fitting and gap merging, is employed to obtain line segments. We evaluated our method against other state-of-the-art methods on YorkUrban-LineSegment and Wireframe. Results show that our method achieves a higher F-score, improving by 2.9% and 2.1%, respectively, while ensuring both precision and recall. Additionally, in 3D reconstruction, our method produces more complete and accurate scenes with fewer fragments and omissions compared to other methods. Our code is available at <span><span>https://github.com/DaipengYang7/BILSD</span><svg><path></path></svg></span>.</div></div>","PeriodicalId":49523,"journal":{"name":"Signal Processing","volume":"233 ","pages":"Article 109950"},"PeriodicalIF":3.4000,"publicationDate":"2025-02-21","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Signal Processing","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0165168425000647","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"ENGINEERING, ELECTRICAL & ELECTRONIC","Score":null,"Total":0}
引用次数: 0
Abstract
Line segment detection is essential for tasks like SLAM, camera pose estimation, and 3D reconstruction. Although many excellent line segment detection methods have been proposed, detecting more true positives while rejecting false positives remains challenging. The human visual system can effectively perceive line segments in complex environments through a processing pathway involving multiple visual cortices. Inspired by this, we propose a novel bio-inspired line segment detection method that mimics the perception of line segments in the visual cortex. Our method models orientation-selective neurons in the primary and secondary visual cortices. Based on the preferred orientations of these neurons, we integrate them to mimic the function of orientation and curvature domains in the fourth visual cortex, generating continuous and smooth edge segments. A post-processing step, including least squares line fitting and gap merging, is employed to obtain line segments. We evaluated our method against other state-of-the-art methods on YorkUrban-LineSegment and Wireframe. Results show that our method achieves a higher F-score, improving by 2.9% and 2.1%, respectively, while ensuring both precision and recall. Additionally, in 3D reconstruction, our method produces more complete and accurate scenes with fewer fragments and omissions compared to other methods. Our code is available at https://github.com/DaipengYang7/BILSD.
期刊介绍:
Signal Processing incorporates all aspects of the theory and practice of signal processing. It features original research work, tutorial and review articles, and accounts of practical developments. It is intended for a rapid dissemination of knowledge and experience to engineers and scientists working in the research, development or practical application of signal processing.
Subject areas covered by the journal include: Signal Theory; Stochastic Processes; Detection and Estimation; Spectral Analysis; Filtering; Signal Processing Systems; Software Developments; Image Processing; Pattern Recognition; Optical Signal Processing; Digital Signal Processing; Multi-dimensional Signal Processing; Communication Signal Processing; Biomedical Signal Processing; Geophysical and Astrophysical Signal Processing; Earth Resources Signal Processing; Acoustic and Vibration Signal Processing; Data Processing; Remote Sensing; Signal Processing Technology; Radar Signal Processing; Sonar Signal Processing; Industrial Applications; New Applications.