基于深度学习框架的装配机器人零件实时检测方法

doi:10.3969/j.issn.1000-1093.2020.10.022

摘要/Abstract

摘要： 针对工业生产线上装配机器人在粘连、堆叠、光照变化及环境因素干扰等复杂条件下零件检测率低、鲁棒性差等问题，提出一种基于改进YOLOv3深度学习框架的零件实时检测方法。在基础特征提取网络Darknet-53的每个残差网络后嵌入CFE模块，构建融合CFE模块和Darknet-53的深度特征提取网络CFE-Darknet53，建立YOLOv3深度学习框架下基于CFE-Darknet53的零件实时检测模型，提升检测网络在复杂环境下特征提取能力；设计一种改进K-means算法来预测边界框，通过对零件数据集进行聚类分析，选取最优的锚框个数和尺寸，进一步提高检测准确性。实验结果表明：在复杂条件下，改进算法对相似度很高的多类零件检测准确率能达到91.6%以上，相比YOLOv3算法提升了近10%以上；检测时间为43 ms，在视频传输帧率（24帧/s）下实现了零件实时准确检测。

关键词: 机器人, 深度学习, 实时检测, 特征提取, 聚类分析

Abstract: For the problems about low accuracy and poor robustness under complex conditions such as adhesion, stacking, illumination change and environmental factors, a part real-time detection method based on improved YOLOv3 deep learning framework is proposed. The CFE module is embedded in each residual network of Darknet-53 basic feature extraction network. A deep feature extraction network CFE-Darknet53 that combines CFE module and Darknet-53 is constructed. A real-time part detection model based on CFE-Darknet53 under the YOLOv3 deep learning framework is established, which improves the feature extraction ability of the detection network in complex environment. An improved K-means algorithm is designed to predict the bounding box. The data sets of parts are clustered to select the optimal number and size of anchor frames, which further improves the detection accuracy. The experimental results show that the detection accuracy of the improved YOLOv3 algorithm for multi class parts with high similarity can reach 91.6% under complex conditions. Compared with the YOLOv3 algorithm, the improved YOLOv3 algorithm improves the detection accuracy by more than 10%. The detection time is 43 ms, so the improved method achieves real-time detection of parts at the video transmission frame rate of 24 frames/s.

Key words: robot, deeplearning, real-timedetection, featureextraction, clusteranalysis

中图分类号:

TP242.2

余永维，彭西，杜柳青，陈天皓. 基于深度学习框架的装配机器人零件实时检测方法[J]. 兵工学报, 2020, 41(10): 2122-2130.

YU Yongwei， PENG Xi， DU Liuqing， CHEN Tianhao. Real-time Detection of Parts by Assembly Robot Based on Deep Learning Framework[J]. Acta Armamentarii, 2020, 41(10): 2122-2130.

参考文献

［1］ HARRIS C, STEPHENS M. A combined corner and edge detector[C]∥Proceedings of the 4th Alvey Vision Conference. Manchester, UK: AVC,1988:147-151.
[2］ BAY H, TUYTELAARS T, VAN GOOL L. SURF:speeded up robust features[C]∥Proceedings of the 9th European Conference on Computer Vision.Berlin, Germany:Springer, 2006:404-417.
[3］杨龙, 苏娟, 李响. 基于生成式对抗网络的合成孔径雷达舰船数据增广在改进单次多盒检测器中的应用[J].兵工学报, 2019, 40(12):2489-2496.
YANG L, SU J, LI X. Application of SAR ship data augmentation based on generative adversarial network in improved SSD[J]. Acta Armamentarii, 2019, 40(12):2489-2496. (in Chinese)
[4］梁杰, 李磊, 任君, 等. 基于深度学习的红外图像遮挡干扰检测方法[J].兵工学报,2019,40(7):1401-1410.
LIANG J, LI L, REN J, et al. Infrared image occlusion interference detection method based on deep learning[J]. Acta Armamentarii, 2019, 40(7):1401-1410. (in Chinese)
[5］张玉燕, 李永保, 温银堂, 等. 基于Faster-R-卷积神经网络的金属点阵结构缺陷识别方法[J]. 兵工学报, 2019, 40(11):2329-2335.
ZHANG Y Y, LI Y B, WEN Y T, et al. Internal defect detection of metal three-dimensional multi-layer lattice structure based on Faster R-CNN[J]. Acta Armamentarii, 2019, 40(11):2329-2335. (in Chinese)
[6］呙鹏程, 吴礼洋. 融合卷积特征与判别字典学习的低截获概率雷达信号识别[J].兵工学报, 2019, 40(9):1881-1889.
GUO P C, WU L Y. LPI radar signal recognition with convolution feature and discrimination dictionary learning [J]. Acta Armamentarii, 2019, 40(9):1881-1889. (in Chinese)
[7］冷鹏飞, 徐朝阳. 一种深度强化学习的雷达辐射源个体识别方法[J]. 兵工学报, 2018, 39(2):2421-2426.
LENG P F, XU C Y. Specific emitter identification based on deep reinforcement learning[J]. Acta Armamentarii, 2018, 39(2):2421-2426. (in Chinese)
[8］ HE K M, GKIOXARI G, DOLLAR P, et al. Mask R-CNN[C]∥Proceedings of 2017 IEEE International Conference on Computer Vision. Venice, Italy:IEEE,2017:2980-2988.
[9］ DAI J F, LI Y, HE K M, et al. R-FCN:object detection via region based fully convolutional networks[C]∥Proceedings of the 30th International Conference on Neural Information Processing Systems. Barcelona, Spain: IEEE, 2016:1-8.
[10］ REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]∥Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, NV, US:IEEE, 2016.779-788.
[11］ LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot multibox detector[C]∥Proceedings of the 14th European Conference on Computer Vision. Amsterdam, Netherlands: Springer, 2016:21-37.
[12］ REN S Q, HE K M, GIRSHICK R, et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2017, 39(6):1137-1149.
[13］ REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]∥Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition. Washington, DC, US: IEEE, 2017: 6517-6525.
[14］戴伟聪, 金龙旭, 李国宁, 等. 遥感图像中飞机的改进YOLOv3实时检测算法[J]. 光电工程, 2018, 45(12):84-92.
DAI W C, JIN L X, LI G N, et al. Real-time airplane detection algorithm in remote-sensing images based on improved YOLOv3[J]. Opto-Electronic Engineering, 2018, 45(12):84-92. (in Chinese)
[15］王殿伟, 何衍辉, 李大湘, 等. 改进的YOLOv3红外视频图像行人检测算法[J]. 西安邮电大学学报, 2018, 23(4):48-52, 67.
WANG D W, HE Y H, LI D X, et al. An improved infrared video image pedestrian detection algorithm[J]. Journal of Xian University of Posts and Telecommunications, 2018, 23(4):48-52, 67. (in Chinese)
[16］张富凯, 杨峰, 李策. 基于改进YOLOv3的快速车辆检测方法[J]. 计算机工程与应用, 2019，55(2):12-20.
ZHANG F K, YANG F, LI C. Fastvehicle detection method based on improved YOLOv3[J]. Computer Engineering and Applications, 2019,55(2):12-20. (in Chinese)
[17］高宗, 李少波, 陈济楠, 等. 基于YOLO网络的行人检测方法[J].计算机工程, 2018,44(5):215-219.
GAO Z, LI S B, CHEN J N, et al. Pedestrian detection method based on YOLO network[J]. Computer Engineering, 2018, 44(5): 215-219. (in Chinese)
[18］ ZHAO Q J, SHENG T, WANG Y T, et al. CFENet: an accurate and efficient single-shot object detector for autonomous driving[C]∥Proceedings of the 2018 IEEE Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT, US:IEEE,2018:1-6.

第41卷第10期2020 年10月
兵工学报ACTA ARMAMENTARII
Vol.41No.10Oct. 2020

[1]	李良福，陈卫东，高强，许开銮，刘轩，何曦，钱钧. 基于深度学习的光电系统智能目标识别[J]. 兵工学报, 2022, 43(S1): 162-168.
[2]	丁力，谷加辉，周金宇，康绍鹏，陈逸飞，李子依. 基于包络法的冗余机器人运动可靠性[J]. 兵工学报, 2022, 43(6): 1426-1434.
[3]	李锦青，刘泽飞，满振龙. 基于生成对抗网络的密钥生成方法及其在微光图像加密中的应用[J]. 兵工学报, 2022, 43(2): 337-344.
[4]	李庆中，李晓丹，于福杰，陈原. 介电弹性体式蛙型仿生软体机器人设计[J]. 兵工学报, 2022, 43(1): 140-147.
[5]	闫银坡，于福杰，陈原. 开架式水下机器人水动力系数计算与动力学建模[J]. 兵工学报, 2021, 42(9): 1972-1986.
[6]	李恒, 叶祖坤, 查文彬, 王禹林. 基于多传感器信息决策级融合的刀具磨损在线监测[J]. 兵工学报, 2021, 42(9): 2024-2031.
[7]	普运伟，刘涛涛，郭江，吴海潇. 基于卷积神经网络和模糊函数主脊坐标变换的雷达辐射源信号识别[J]. 兵工学报, 2021, 42(8): 1680-1689.
[8]	朱敏，许爱强，许晴，李睿峰. 基于改进多层核超限学习机的模拟电路故障诊断[J]. 兵工学报, 2021, 42(2): 356-369.
[9]	刘亚丽，宋遒志，赵明升，周能兵，刘悦. 基于力位混合控制的踝关节外骨骼机器人四段式助力技术[J]. 兵工学报, 2021, 42(12): 2722-2730.
[10]	张凤, 张超, 杨华民, 王发斌. 作战辅助系统中基于深度学习的投影图像颜色补偿方法[J]. 兵工学报, 2021, 42(11): 2418-2423.
[11]	薛灵芝, 曾向阳, 杨爽. 基于生成对抗网络的水声目标识别算法[J]. 兵工学报, 2021, 42(11): 2444-2452.
[12]	严浩, 王洪波, 陈鹏, 张雷雷, 李云贵. 具有广义肩关节的上肢康复机器人优化设计[J]. 兵工学报, 2021, 42(11): 2491-2502.
[13]	王成，吴岩，杨廷飞. 利用改进单分类支持向量机提升舰船尾流目标的检测准确率[J]. 兵工学报, 2020, 41(9): 1887-1893.
[14]	郑荣，辛传龙，汤钟，宋涛. 无人水面艇自主部署自主水下机器人平台技术综述[J]. 兵工学报, 2020, 41(8): 1675-1687.
[15]	桑董辉，陈原，高军. 由2自由度并联髋关节构成的脊柱式四足步行机器人步态规划和稳定性分析[J]. 兵工学报, 2020, 41(6): 1188-1200.