局部特征表征的6D位姿估计算法

6D pose estimation algorithm for local feature representation

为解决有纹理模型在遮挡条件下6D位姿估计精确度不高的问题,提出了一种局部特征表征的端到端6D位姿估计算法。首先为了得到准确的定位信息,提出了一个空间—坐标注意力机制(spatial and coordinate attention),通过在YOLOv5网络中加入空间—坐标注意力机制和加权双向特征金字塔网络(bidirectional feature pyramid network),YOLOv5-CBE算法的精确度(precision)、召回率(recall)、平均精度均值(mAP@0.5)分别提升了3.6%、2.8%、2.5%,局部特征中心点坐标误差最高提升了25%;然后用YOLOv5-CBE算法检测局部特征关键点,结合3D Harris关键点通过奇异值分解法(singular value decomposition)计算模型的6D位姿,最高遮挡70%的情况下仍然可以保证二维重投影精度(2D reprojection accuracy)和ADD度量精度(ADD accuracy)在95%以上,具有较强的鲁棒性。

In order to solve the problem of low accuracy of 6D pose estimation for textured models under occlusion,this paper proposed an end-to-end 6D pose estimation algorithm based on local feature representation.Firstly,this paper proposed a spatial and coordinate attention mechanism to obtain accurate localization information.A YOLOv5-CBE detection network formed by adding the attention mechanism to the backbone network and introducing a weighted bidirectional feature pyramid Network in the detection layer.The precision,recall and mAP@0.5 of YOLOV5-CBE algorithm rise by 3.6%,2.8%and 2.5%respectively,and the coordinate error of local feature center point decreases by 25%at most.Secondly,the YOLOv5-CBE network detected the local feature key points and calculated 6D pose of the model with 3D Harris key points by singular value decomposition,and the algorithm can guarantee 2D reprojection accuracy and ADD accuracy above 95%with 70%occlusion,which has a strong robustness.