Real-Time Active Acquisition and Representation of 3-Dimensional Objects

3 维物体的实时主动采集和表示

• 批准号: 9022468
• 负责人: Terrance Boult
• 金额: $ 3.59万
• 依托单位: Columbia University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1991
• 资助国家: 美国
• 起止时间: 1991-04-01 至 1992-03-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9022468&HistoricalAwards=false
• 关键词: Real; Time; Active; Acquisition; Representation

This award provides funds to purchase a compact, lightweight rangefinder for use on a robot arm. The research supported by the arm with the rangefinder includes the development of algorithms to recover superquadrics from depth data, evaluation of the quality of various "shape-from-" algorithms, investigation of the use of real- time sensor feedback to control assembly and parts mating activities using a hand mounted scanner, and the development of algorithms using the sensor data for intelligent grasping tasks. The use of multiple information sources is characteristic of much of the current research in robotics. A rangefinder, capable of real time data acquisition, mounted on a robot arm provides the capability, through moving the arm, of obtaining accurate three dimensional surface data. Two distinct types of research are supported under this award. The first is the development of algorithms that will use the rangefinder and the robot arm motion to obtain three dimensional information, in real time, about the objects being examined. The second type of research is to incorporate the additional information into algorithms for grasping and assembling objects.

该奖项提供资金购买一个紧凑，轻便 测距仪上使用的机器人手臂。研究支持的 与测距仪的手臂包括算法的发展， 从深度数据恢复超二次曲面， 各种“形从”算法，调查使用的真实的- 时间传感器反馈，以控制组件和部件配合 使用手持扫描仪的活动，以及 算法使用传感器数据进行智能抓取任务。 使用多种信息来源的特点是 当前机器人研究的一部分。 测距仪，能够 安装在机械臂上的真实的实时数据采集提供了 能力，通过移动手臂，获得准确的三个 三维表面数据。 两种不同类型的研究是 在这个奖项的支持下。 第一是发展 使用测距仪和机械臂运动的算法 为了在真实的时间内获得关于 正在检查的对象。 第二类研究是 将额外的信息纳入抓取算法 和组装物品。