{"title":"融合关键帧聚焦叠加图像的医疗服务机器人全场景重构","authors":"Yanzhu Hu, Yingjian Wang, Song Wang","doi":"10.1145/3429889.3430077","DOIUrl":null,"url":null,"abstract":"Facing the precise service requirements of medical service robots in irregular environments, to solve the problem of scene depth estimation and medical service scene reconstruction of monocular robots, an algorithm framework for whole scene three-dimensional (3D) point cloud reconstruction based on scene focus stack images fusion with monocular SLAM sparse key frames is proposed. The algorithm first collects the focus stack image under the initial viewing angle through the medical service robot mobile terminal of the camera. On the server side, multi-frame focused images are iteratively reconstructed by light field imaging technology to reconstruct the scene depth image. Further generate a high-precision 3D point cloud reconstruction image of one single scene. Finally combine the sparse key frame information of the monocular SLAM on the mobile terminal and the camera movement pose information. The 3D point cloud reconstruction within the scope of the whole medical service scene is realized through multi-perspective point cloud matching. Provide scene point cloud reconstruction information for the target recognition and navigation of medical service robots.","PeriodicalId":315899,"journal":{"name":"Proceedings of the 1st International Symposium on Artificial Intelligence in Medical Sciences","volume":"387 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-09-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"1","resultStr":"{\"title\":\"Whole scene reconstruction of medical service robots fused with key frame focus stack images\",\"authors\":\"Yanzhu Hu, Yingjian Wang, Song Wang\",\"doi\":\"10.1145/3429889.3430077\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Facing the precise service requirements of medical service robots in irregular environments, to solve the problem of scene depth estimation and medical service scene reconstruction of monocular robots, an algorithm framework for whole scene three-dimensional (3D) point cloud reconstruction based on scene focus stack images fusion with monocular SLAM sparse key frames is proposed. The algorithm first collects the focus stack image under the initial viewing angle through the medical service robot mobile terminal of the camera. On the server side, multi-frame focused images are iteratively reconstructed by light field imaging technology to reconstruct the scene depth image. Further generate a high-precision 3D point cloud reconstruction image of one single scene. Finally combine the sparse key frame information of the monocular SLAM on the mobile terminal and the camera movement pose information. The 3D point cloud reconstruction within the scope of the whole medical service scene is realized through multi-perspective point cloud matching. Provide scene point cloud reconstruction information for the target recognition and navigation of medical service robots.\",\"PeriodicalId\":315899,\"journal\":{\"name\":\"Proceedings of the 1st International Symposium on Artificial Intelligence in Medical Sciences\",\"volume\":\"387 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-09-11\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"1\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 1st International Symposium on Artificial Intelligence in Medical Sciences\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3429889.3430077\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 1st International Symposium on Artificial Intelligence in Medical Sciences","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3429889.3430077","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Whole scene reconstruction of medical service robots fused with key frame focus stack images
Facing the precise service requirements of medical service robots in irregular environments, to solve the problem of scene depth estimation and medical service scene reconstruction of monocular robots, an algorithm framework for whole scene three-dimensional (3D) point cloud reconstruction based on scene focus stack images fusion with monocular SLAM sparse key frames is proposed. The algorithm first collects the focus stack image under the initial viewing angle through the medical service robot mobile terminal of the camera. On the server side, multi-frame focused images are iteratively reconstructed by light field imaging technology to reconstruct the scene depth image. Further generate a high-precision 3D point cloud reconstruction image of one single scene. Finally combine the sparse key frame information of the monocular SLAM on the mobile terminal and the camera movement pose information. The 3D point cloud reconstruction within the scope of the whole medical service scene is realized through multi-perspective point cloud matching. Provide scene point cloud reconstruction information for the target recognition and navigation of medical service robots.