{"title":"Model based part segmentation of range data-hyperquadrics and dividing planes","authors":"Senthil Kumar, Dmitry Goldgof","doi":"10.1109/PBMCV.1995.514663","DOIUrl":null,"url":null,"abstract":"In this paper, we present a novel techn,ique for the volum.etric decomposition of ran,ge data into parts. In this model based technique, we use hyperquadrics as part models. A hyperquadric is represented by an implicit equation that is composed of an arbitrary number of terms. A given hyperquadric model with a certain number of terms can be split into two separate models (or parts) by the introduction of an additional term. This additional term can be viewed as a dividing plane that divides th,e parent model. We use such a splitting scheme to subdivide a given object recursively into its constitute parts. We provide a two-stage segmentation process in which the first stage uses dividing planes to produce an, approximate segmentation, of the data and the second stage uses hyperquadric models to refine the above segmentation, A final merging stage corrects oversegmentation to obtain good part decomposition. Experimental results with real data are presented.","PeriodicalId":343932,"journal":{"name":"Proceedings of the Workshop on Physics-Based Modeling in Computer Vision","volume":"7 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"1995-06-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the Workshop on Physics-Based Modeling in Computer Vision","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/PBMCV.1995.514663","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 6
Abstract
In this paper, we present a novel techn,ique for the volum.etric decomposition of ran,ge data into parts. In this model based technique, we use hyperquadrics as part models. A hyperquadric is represented by an implicit equation that is composed of an arbitrary number of terms. A given hyperquadric model with a certain number of terms can be split into two separate models (or parts) by the introduction of an additional term. This additional term can be viewed as a dividing plane that divides th,e parent model. We use such a splitting scheme to subdivide a given object recursively into its constitute parts. We provide a two-stage segmentation process in which the first stage uses dividing planes to produce an, approximate segmentation, of the data and the second stage uses hyperquadric models to refine the above segmentation, A final merging stage corrects oversegmentation to obtain good part decomposition. Experimental results with real data are presented.