对托盘的检测是仓储机器人进行货物搬运的关键步骤.针对当前检测方法光照鲁棒性不强、受托盘与传感器之间相对位姿的约束等问题，提出一种基于点云平面轮廓匹配的检测方法.该方法使用ToF(Time-of-Flight)相机采集点云；点云经预处理后，使用以法线为约束的区域生长算法进行平面分割，并沿其主法线方向投影生成栅格图，解决受相对位姿约束的问题；最后在对栅格图进行轮廓提取后，利用融合Hu不变矩和尺度比例特征的轮廓特征进行目标与模板的匹配，实现对托盘的检测.实验结果表明，该方法在光照不同、托盘位姿不定等场景下均具有高识别率和强鲁棒性.
Pallet detection is the key step of cargo handling for warehouse robots. A pallet detection method based on point clouds plane contour matching is proposed to solve the current detection method problems that are not robust to illumination and relative position between the pallet and sensor. Point clouds generated by time-of-flight (ToF) camera are filtered and segmented to different planes using region-growing method constrained by surface normal. Then point clouds are projected to the grid image along the plane’s principle normal direction. Fusion contour feature of Hu moment invariants and scale feature extracted from grid image contour is applied for similarity matching between the target and template pallet contour. The experimental results show that the method has high recognition rate and strong robustness under the circumstance of complex illumination, uncertain distance and relative pose between the pallet and sensor.
［1］CUI G Z, LU L S, HE Z D, et al. A robust autonomous mobile forklift pallet recognition［C］//International Asia Conference on Informatics in Control, Automation and Robotics. Wuhan, China: IEEE, 2010: 286-290.
［2］CHEN G, PENG R, WANG Z, et al. Pallet red localization method for vision guided forklift［C］//2012 8th International Conference on Wireless Communications, Networking and Mobile Computing, Shanghai, China: IEEE, 2012: 1-4.
［3］SEELINGER M, YODER J D. Automatic pallet engagment by a vision guided forklift［C］//IEEE International Conference on Robotics and Automation. Orlando, USA: IEEE, 2006: 4068-4073.
［4］SYU J L, LI H T, CHIANG J S, et al. A computer vision assisted system for autonomous forklift vehicles in real factory environment［J］. Multimedia Tools and Applications, 2017, 76(18): 18387-18407.
［5］VARGA R, NEDEVSCHI S. Vision-based autonomous load handling for automated guided vehicles［C］//IEEE International Conference on Intelligent Computer Communication and Processing. Santa Clara, USA: IEEE, 2014: 239-244.
［6］VARGA R, COSTEA A, NEDEVSCHI S. Improved autonomous load handling with stereo cameras［C］//IEEE International Conference on Intelligent Computer Communication and Processing. Cluj-Napoca, Romania: IEEE, 2015: 251-256.
［7］VARGA R, NEDEVSCHI S. Robust pallet detection for automated logistics operations［C］//International Conference on Computer Vision Theory and Applications. Rome, Italy: SciTePress, 2016: 470-477.
［8］BAGLIVO L. Pallet pose estimation with LIDAR and vision for autonomous forklifts［J］. IFAC Proceedings Volumes, 2009, 42(4): 612-617.
［9］BAGLIVO L, BIASI N, BIRAL F, et al. Autonomous pallet localization and picking for industrial forklifts: A robust range and look method［J］. Mea-surement Science & Technology, 2011, 22(8): 1312-1327.
［10］HU M. Visual pattern recognition by moment invariants［J］. Information Theory Ire Transactions on, 1962, 8(2): 179-187.
［11］李宝顺, 岑红燕, 包亚萍, 等. 基于平面提取的点云数据分割算法［J］. 计算机应用与软件, 2014, 31(7): 145-148.
LI Baoshun, CEN Hongyan, BAO Yaping, et al. Segmentation algorithm for point cloud data based on plane extraction［J］. Computer Application and Software, 2014, 31(7): 145-148.
