{"title":"设计和开发一个自动系统,用于创建图像数据集,以允许服务机器人识别和抓取对象","authors":"J. A. Fabro, M. Vaz, A. Oliveira","doi":"10.21528/cbic2019-136","DOIUrl":null,"url":null,"abstract":"—This work presents the design and construction of an operational prototype that allows the creation of datasets that allows service robots to identify and grasp household objects. The prototype was designed and built with several movable and removable parts. Unlike similar systems described in the literature, this structure can be disassembled and transported. Due to the lack of specific datasets designed to allow detection, recognition and grasping tasks for service robots, it was decided to build a structure for data collection in competitive environments, such as RoboCup@Home. With the prototype ready, a first image capture test was performed and 251 images were obtained for a single object. After the necessary adjustments, the prototype is expected to produce automatically 3600 images from various points of view for each object of interest without human interference.","PeriodicalId":160474,"journal":{"name":"Anais do 14. Congresso Brasileiro de Inteligência Computacional","volume":"19 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Design and Development of an Automated System for creation of Image Datasets intended to Allow Object Identification and Grasping by Service Robots\",\"authors\":\"J. A. Fabro, M. Vaz, A. Oliveira\",\"doi\":\"10.21528/cbic2019-136\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"—This work presents the design and construction of an operational prototype that allows the creation of datasets that allows service robots to identify and grasp household objects. The prototype was designed and built with several movable and removable parts. Unlike similar systems described in the literature, this structure can be disassembled and transported. Due to the lack of specific datasets designed to allow detection, recognition and grasping tasks for service robots, it was decided to build a structure for data collection in competitive environments, such as RoboCup@Home. With the prototype ready, a first image capture test was performed and 251 images were obtained for a single object. After the necessary adjustments, the prototype is expected to produce automatically 3600 images from various points of view for each object of interest without human interference.\",\"PeriodicalId\":160474,\"journal\":{\"name\":\"Anais do 14. Congresso Brasileiro de Inteligência Computacional\",\"volume\":\"19 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-01-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Anais do 14. Congresso Brasileiro de Inteligência Computacional\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.21528/cbic2019-136\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Anais do 14. Congresso Brasileiro de Inteligência Computacional","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.21528/cbic2019-136","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Design and Development of an Automated System for creation of Image Datasets intended to Allow Object Identification and Grasping by Service Robots
—This work presents the design and construction of an operational prototype that allows the creation of datasets that allows service robots to identify and grasp household objects. The prototype was designed and built with several movable and removable parts. Unlike similar systems described in the literature, this structure can be disassembled and transported. Due to the lack of specific datasets designed to allow detection, recognition and grasping tasks for service robots, it was decided to build a structure for data collection in competitive environments, such as RoboCup@Home. With the prototype ready, a first image capture test was performed and 251 images were obtained for a single object. After the necessary adjustments, the prototype is expected to produce automatically 3600 images from various points of view for each object of interest without human interference.