【作者】 刘硕； 吴兆跃； 袁家政； 刘宏哲；

【Author】 LIU Shuo;WU Zhao-yue;YUAN Jia-zheng;LIU Hong-zhe;Beijing Key Laboratory of Information Service Engineering,Beijing Union University;

【机构】 北京联合大学北京市信息服务工程重点实验室；

【摘要】 近年来,随着智能化进程的加快,服务机器人技术得到了飞速发展。物品识别与定位技术作为智能服务机器人的必备功能,也逐渐成为人们的研究热点。而精准的物品识别和定位能力则是机器人执行指定任务的先决条件。传统的物品识别与定位方法大多是基于二维RGB图像,根据物品在颜色图像中呈现的诸多特征来进行相应的识别和定位。虽然这些方法取得了不错的效果,但由于物品具有三维特性,无论是形状识别还是6自由度位姿估计都需要充足的空间信息。因此针对基于二维图像和基于点云的定位与识别方法进行总结,并对发展前景进行了展望。更多还原

【Abstract】 In recent years,with the acceleration of the intelligent process.service robot technology has been developed rapidly.As a smart service robot essential function,object recognition and location technology have gradually become a research hotspot And accurate object identification and location capabilities are prerequisites for the robot to perform the specified task.Conventional object recognition and location methods are mostly based on two-dimensional RGB images,according to the items in the color image presented in a number of features to identify and locate the corresponding.Although these methods have achieved good results,as objects have the natural characteristics of the threedimensional items,whether shape recognition or 6 degrees of freedom pose estimation both need sufficient space information.In this paper,we summarized the location and recognition methods based on two-dimensional and point-based cloud and development prospects.更多还原