Xiaoliang Zou, Guihua Zhao, Jonathan Li, Yuanxi Yang, Yong Fang
{"title":"移动序列图像处理的多视图匹配算法","authors":"Xiaoliang Zou, Guihua Zhao, Jonathan Li, Yuanxi Yang, Yong Fang","doi":"10.1061/(ASCE)SU.1943-5428.0000235","DOIUrl":null,"url":null,"abstract":"The paper presents a multiview matching algorithm for processing sequence images acquired by a mobile mapping system (MMS). The workflow of the multiview matching algorithm is designed, and the algorithm is based on motion analysis of sequence images in computer vision. To achieve a high multiview matching accuracy, camera lens distortion in sequence images is first corrected, and images can then be resampled. Image points on sequence images are extracted using the Harris operator. The homologous image points are then matched based on correlation coefficients and used to make a robust estimation for a fundamental matrix F between the two adjacent images using the random sample consensus (RANSAC) algorithm. The fundamental matrix F is calculated under the condition of epipolar line constraints. Finally, the trifocal tensor T of the three-view images is calculated to achieve highly accurate triplet image points. These triplet image points are then provided as the initial value for bundle adjustment. The algorithm was tested using a set of sequence images. The results demonstrate that the designed workflow is available and the algorithm is promising in terms of both accuracy and feasibility. DOI: 10.1061/(ASCE) SU.1943-5428.0000235. This work is made available under the terms of the Creative Commons Attribution 4.0 International license, http:// creativecommons.org/licenses/by/4.0/. Author keywords: Sequence images; Multiview matching; Harris operator; Correlation coefficient; Random sample consensus (RANSAC); Trifocal tensor; Computer vision (CV); Mobile mapping system (MMS).","PeriodicalId":210864,"journal":{"name":"Journal of Surveying Engineering-asce","volume":"13 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2017-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Multiview Matching Algorithm for Processing Mobile Sequence Images\",\"authors\":\"Xiaoliang Zou, Guihua Zhao, Jonathan Li, Yuanxi Yang, Yong Fang\",\"doi\":\"10.1061/(ASCE)SU.1943-5428.0000235\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"The paper presents a multiview matching algorithm for processing sequence images acquired by a mobile mapping system (MMS). The workflow of the multiview matching algorithm is designed, and the algorithm is based on motion analysis of sequence images in computer vision. To achieve a high multiview matching accuracy, camera lens distortion in sequence images is first corrected, and images can then be resampled. Image points on sequence images are extracted using the Harris operator. The homologous image points are then matched based on correlation coefficients and used to make a robust estimation for a fundamental matrix F between the two adjacent images using the random sample consensus (RANSAC) algorithm. The fundamental matrix F is calculated under the condition of epipolar line constraints. Finally, the trifocal tensor T of the three-view images is calculated to achieve highly accurate triplet image points. These triplet image points are then provided as the initial value for bundle adjustment. The algorithm was tested using a set of sequence images. The results demonstrate that the designed workflow is available and the algorithm is promising in terms of both accuracy and feasibility. DOI: 10.1061/(ASCE) SU.1943-5428.0000235. This work is made available under the terms of the Creative Commons Attribution 4.0 International license, http:// creativecommons.org/licenses/by/4.0/. Author keywords: Sequence images; Multiview matching; Harris operator; Correlation coefficient; Random sample consensus (RANSAC); Trifocal tensor; Computer vision (CV); Mobile mapping system (MMS).\",\"PeriodicalId\":210864,\"journal\":{\"name\":\"Journal of Surveying Engineering-asce\",\"volume\":\"13 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2017-11-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Surveying Engineering-asce\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1061/(ASCE)SU.1943-5428.0000235\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Surveying Engineering-asce","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1061/(ASCE)SU.1943-5428.0000235","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Multiview Matching Algorithm for Processing Mobile Sequence Images
The paper presents a multiview matching algorithm for processing sequence images acquired by a mobile mapping system (MMS). The workflow of the multiview matching algorithm is designed, and the algorithm is based on motion analysis of sequence images in computer vision. To achieve a high multiview matching accuracy, camera lens distortion in sequence images is first corrected, and images can then be resampled. Image points on sequence images are extracted using the Harris operator. The homologous image points are then matched based on correlation coefficients and used to make a robust estimation for a fundamental matrix F between the two adjacent images using the random sample consensus (RANSAC) algorithm. The fundamental matrix F is calculated under the condition of epipolar line constraints. Finally, the trifocal tensor T of the three-view images is calculated to achieve highly accurate triplet image points. These triplet image points are then provided as the initial value for bundle adjustment. The algorithm was tested using a set of sequence images. The results demonstrate that the designed workflow is available and the algorithm is promising in terms of both accuracy and feasibility. DOI: 10.1061/(ASCE) SU.1943-5428.0000235. This work is made available under the terms of the Creative Commons Attribution 4.0 International license, http:// creativecommons.org/licenses/by/4.0/. Author keywords: Sequence images; Multiview matching; Harris operator; Correlation coefficient; Random sample consensus (RANSAC); Trifocal tensor; Computer vision (CV); Mobile mapping system (MMS).