文献类型：期刊文献

英文题名：Pose estimation for workpieces in complex stacking industrial scene based on RGB images

作者：Zhang, Yajun[1];Yi, Jianjun[1];Chen, Yuanhao[1];Dai, Zhiyong[2];Han, Fei[3,4];Cao, Shuqing[3,4]

机构：[1]East China Univ Sci & Technol, Sch Mech & Power Engn, Shanghai, Peoples R China;[2]DeepBlue Acad Sci, Shanghai, Peoples R China;[3]Shanghai Aerosp Control Technol Inst, Shanghai, Peoples R China;[4]Shanghai Key Lab Aerosp Intelligent Control Techn, Shanghai, Peoples R China

年份：2022

卷号：52

期号：8

起止页码：8757

外文期刊名：APPLIED INTELLIGENCE

收录：;EI(收录号：20214511127843);WOS:【SCI-EXPANDED(收录号:WOS:000714510500001)】；

基金：This paper was supported by the Natural Science Fund of China (NSFC) under Grant No.51575186, the Major Program of National Natural Science Foundation of China under Grant No. 61690214, Shanghai Science and Technology Action Plan under Grant No.18DZ1204000, 18510745500, 18510750100, 18510730600, Shanghai Aerospace Science and Technology Innovation Fund (SAST) under Grant No. 2019-080, 2019-116 and Shanghai Sailing Program under Grant No. 20YF1417300.

语种：英文

外文关键词：Pose estimation; Stacking scene; Symmetric workpieces; Spatial expression

摘要：Recently, pose estimation algorithms have been widely discussed. Although been investigated by many prior works, pose estimation for heavily stacked workpieces, like bins, is still a challenge for industrial applications. Moreover, the scene of stacked workpieces is much more arduous than the general scene for robots to carry out routine tasks, such as object detection and picking. This paper aims to address the problem of pose estimation for stacked symmetrical workpieces in the presence of partial occlusions and cluttered backgrounds. To tackle those problems, we propose a novel pose estimation method for industrial stacking scenes based on RGB images. Specifically, due to the common symmetry characteristics of industrial workpieces, we firstly present a new standardized spatial representation method, which could auto-encode the 2D-3D correspondences of symmetrical workpieces. Besides, we introduce a novel GAN-based deep neural network model to reconstruct the representation of stacked workpieces. Based on that, the pose of the target workpieces is predicted based on the reconstructed expression and an improved RANSAC-PnP algorithm. Finally, comprehensive experiments demonstrate that the proposed method outperforms state-of-the-art methods, especially in complex stacking scenes.