Shuvo Kumar Paul , Ovi Paul , Monica Nicolescu , Mircea Nicolescu
{"title":"ROS-Pose: Simplified object detection and planar pose estimation for rapid robotics application development","authors":"Shuvo Kumar Paul , Ovi Paul , Monica Nicolescu , Mircea Nicolescu","doi":"10.1016/j.simpa.2024.100624","DOIUrl":null,"url":null,"abstract":"<div><p>We introduce an open-source package, ROS-Pose, that allows simultaneous object detection and planar pose estimation through the utilization of image feature detectors and descriptors. Specifically, when presented with an image capturing any of the object’s planes, the software is designed to achieve real-time estimation of the object’s location and pose. The software further offers a visualization mechanism for object detection and the corresponding estimated pose by superimposing directional vectors on the objects. Most importantly, this software requires only a single image per object, making it conducive to fast prototyping and development of robotic applications.</p></div>","PeriodicalId":29771,"journal":{"name":"Software Impacts","volume":null,"pages":null},"PeriodicalIF":1.3000,"publicationDate":"2024-02-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2665963824000125/pdfft?md5=396445d2b07d48aad7e7527ad79ba4b3&pid=1-s2.0-S2665963824000125-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Software Impacts","FirstCategoryId":"1085","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2665963824000125","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, SOFTWARE ENGINEERING","Score":null,"Total":0}
引用次数: 0
Abstract
We introduce an open-source package, ROS-Pose, that allows simultaneous object detection and planar pose estimation through the utilization of image feature detectors and descriptors. Specifically, when presented with an image capturing any of the object’s planes, the software is designed to achieve real-time estimation of the object’s location and pose. The software further offers a visualization mechanism for object detection and the corresponding estimated pose by superimposing directional vectors on the objects. Most importantly, this software requires only a single image per object, making it conducive to fast prototyping and development of robotic applications.