[1]魏文戈,谭晓阳.密集堆叠下的高相似度木块横截面检测[J].智能系统学报,2019,14(04):642-649.[doi:10.11992/tis.201806001]
 WEI Wenge,TAN Xiaoyang.Highly similar wood blocks detection under dense stacking[J].CAAI Transactions on Intelligent Systems,2019,14(04):642-649.[doi:10.11992/tis.201806001]
点击复制

密集堆叠下的高相似度木块横截面检测(/HTML)
分享到:

《智能系统学报》[ISSN:1673-4785/CN:23-1538/TP]

卷:
第14卷
期数:
2019年04期
页码:
642-649
栏目:
出版日期:
2019-07-02

文章信息/Info

Title:
Highly similar wood blocks detection under dense stacking
作者:
魏文戈12 谭晓阳12
1. 南京航空航天大学 计算机科学与技术学院, 江苏 南京 211106;
2. 软件新技术与产业化协同创新中心, 江苏 南京 211106
Author(s):
WEI Wen’ge12 TAN Xiaoyang12
1. College of Computer Science and Technology, Nanjing University of Aeronautics and Astronautics, Nanjing 211106, China;
2. Collaborative Innovation Center of Novel Software Technology and Industrialization, Nanjing 211106, China
关键词:
密集堆叠高相似度木块横截面检测木块生产交易损失函数鲁棒性
Keywords:
dense stackinghighly similarcross section of wooddetectionproduction of wood blockloss functionrobustness
分类号:
TP181
DOI:
10.11992/tis.201806001
摘要:
快速有效地检测和获取木块横截面信息,是提升木块生产交易效率的关键。由于木块往往被密集堆叠、木块横截面相似度高且边界不明显,给检测木块横截面信息带来了较大的挑战。针对密集堆叠下的高相似度木块横截面检测困难,本文提出了简单高效的Wood R-CNN网络模型,通过改进模型的损失函数和非极大值抑制算法来提升检测精度,简化网络结构和改进特征金字塔网络来保证检测速度。实验证明:该模型可在密集堆叠情况下精确地检测高相似度木块横截面,检测速度较快且鲁棒性良好,可实际运用于木块生产和交易中。
Abstract:
Quick and effective detection of wood blocks cross section and acquisition of the required information are the key to improving the efficiency of wood block production and transactions. However, since wood blocks are often densely stacked, their cross sections have high similarity, while their boundaries are not obvious; this poses a great challenge in detecting the cross sections and acquiring the required information. Considering the difficulty in detecting the cross section of highly similar wood blocks under dense stacking, a simple and efficient wood R-CNN network model is proposed in this paper. The detection accuracy is increased by improving the model loss function and the non-maximum suppression algorithm, and detection speed is ensured by simplifying the network structure and modifying the feature pyramid network. A series of experiments prove that the algorithm model can precisely detect the cross sections of blocks with high similarity under dense stacking. Moreover, it can guarantee fast detection speed and good robustness and can be actually used in the production and transaction of wood blocks.

参考文献/References:

[1] GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Columbus, USA, 2014:580-587.
[2] GIRSHICK R. Fast R-CNN[C]//Proceedings of the IEEE International Conference on Computer Vision. Santiago, Chile, 2015:1440-1448.
[3] REN Shaoqing, HE Kaiming, GIRSHICK R, et al. Faster R-CNN:Towards real-time object detection with region proposal networks[C]//Proceedings of the 28th International Conference on Neural Information Processing Systems. Montreal, Canada, 2015:91-99.
[4] HE Kaiming, GKIOXARI G, DOLLÁR P, et al. Mask R-CNN[C]//2017 IEEE International Conference on Computer Vision. Venice, Italy, 2017:2980-2988.
[5] HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Spatial pyramid pooling in deep convolutional networks for visual recognition[C]//European Conference on Computer Vision. Zurich, Switzerland, 2014:346-361.
[6] LIU Wei, ANGUELOV D, ERHAN D, et al. SSD:Single shot multibox detector[C]//European Conference on Computer Vision. Amsterdam, The Netherlands, 2016:21-37.
[7] REDMON J, DIVVALA S, GIRSHICK R, et al. You only look once:Unified, real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, USA, 2016:779-788.
[8] REDMON J, FARHADI A. YOLO9000:better, faster, stronger[J]. arXiv:1612.08242, 2016.
[9] UIJLINGS J R R, VAN DE SANDE K E A, GEVERS T, et al. Selective search for object recognition[J]. International journal of computer vision, 2013, 104(2):154-171.
[10] 于文勇, 王立海, 杨慧敏, 等. 超声波木材缺陷检测若干问题的探讨[J]. 森林工程, 2006, 22(6):7-9 YU Wenyong, WANG Lihai, YANG Huimin, et al. The study of wood defects detection based on ultrasonic techniques[J]. Forest engineering, 2006, 22(6):7-9
[11] 杨春梅, 胡万义, 白帆, 等. 木材缺陷检测理论及方法的发展[J]. 林业机械与木工设备, 2004, 32(3):8-10 YANG Chunmei, HU Wanyi, BAI Fan, et al. The development of theory and method in assaying wood defection[J]. Forestry machinery and woodworking equipment, 2004, 32(3):8-10
[12] 戴天虹, 李琳, 解朦. 基于四元数矩阵奇异值分解的木材缺陷检测分析[J]. 森林工程, 2014, 30(1):52-55, 59 DAI Tianhong, LI Lin, XIE Meng. Wood defect detection and analysis based on quaternion matrix singular value decomposition[J]. Forest engineering, 2014, 30(1):52-55, 59
[13] 梁善庆, 蔡智勇, 王喜平, 等. 北美木材无损检测技术的研究与应用[J]. 木材工业, 2008, 22(3):5-8 LIANG Shanqing, CAI Zhiyong, WANG Xiping, et al. Developments and applications of nondestructive tests for wood in North America[J]. China wood industry, 2008, 22(3):5-8
[14] 戚大伟, 牟洪波. 基于Hu不变矩和BP神经网络的木材缺陷检测[J]. 东南大学学报(自然科学版), 2013, 43(S1):63-66 QI Dawei, Mu Hongbo. Detection of wood defects types based on Hu invariant moments and BP neural network[J]. Journal of Southeast University (natural science edition), 2013, 43(S1):63-66
[15] 戚大伟, 牟洪波. 人工神经网络在木材缺陷检测中的应用[J]. 森林工程, 2006, 22(1):21-23 QI Dawei, MU Hongbo. Application of artificial neural networks in the testing of wood defects[J]. Forest engineering, 2006, 22(1):21-23
[16] 牟洪波. 基于人工神经网络的木材缺陷检测研究[D]. 哈尔滨:东北林业大学, 2006. MU Hongbo. Study on wood defects testing based on artificial neural network[D]. Harbin:Northeast Forestry University, 2006.
[17] 夏春江, 王培良, 张媛. 基于深度学习的木材含水率预测[J]. 杭州电子科技大学学报(自然科学版), 2015, 35(1):31-35 XIA Chunjiang, WANG Peiliang, ZHANG Yuan. Prediction of moisture content of wood based on deep learning[J]. Journal of Hangzhou Dianzi University (natural sciences), 2015, 35(1):31-35
[18] 崔金刚. 图像处理技术在木材检测方面的应用[D]. 哈尔滨:东北林业大学, 2007. CUI Jingang. The digital image processing using on wood detests testing[D]. Harbin:Northeast Forestry University, 2007.
[19] 戚大伟, 乔景禄, 张杰. 图象处理技术在木材年轮检测中的应用[J]. 东北林业大学学报, 1994, 22(2):106-111 QI Dawei, QIAO Jinglu, ZHANG Jie. Image processing technique on wooden annual rings x-ray picture[J]. Journal of Northeast Forestry University, 1994, 22(2):106-111
[20] 徐华东, 王立海, 倪松远. 模态分析技术在木材性质检测中应用的研究进展[J]. 森林工程, 2007, 23(6):15-17 XU Huadong, WANG Lihai, NI Songyuan. Literature review on application of modal analysis technology in the wood testing[J]. Forest engineering, 2007, 23(6):15-17
[21] 杨学春, 王立海. 应力波技术在木材性质检测中的研究进展[J]. 森林工程, 2002, 18(6):11-12 YANG Xuechun, WANG Lihai. Research progress of testing timber characteristics by using stress wave technique[J]. Forest engineering, 2002, 18(6):11-12
[22] 冯海林, 李光辉, 方益明, 等. 应力波传播模型及其在木材检测中的应用[J]. 系统仿真学报, 2010, 22(6):1490-1493 FENG Hailin, LI Guanghui, FANG Yiming, et al. Stress wave propagation modeling and application in wood testing[J]. Journal of system simulation, 2010, 22(6):1490-1493
[23] 韩书霞, 于雷, 孔超, 等. 基于计算机断层扫描技术的木材密度检测研究[J]. 森林工程, 2007, 23(1):19-23 HAN Shuxia, YU Lei, KONG Chao, et al. Research on wood density testing based on computer tomography technology[J]. Forest engineering, 2007, 23(1):19-23
[24] 王立海, 孙天用. 木材腐朽检测及防治的研究进展[J]. 黑龙江大学工程学报, 2011, 2(2):63-69 WANG Lihai, SUN Tianyong. Research on progressing of testing and prevention of wood decay[J]. Journal of Engineering of Heilongjiang University, 2011, 2(2):63-69
[25] HE Kaiming, ZHANG Xiangyu, REN Shaoqing, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. Las Vegas, USA, 2016:770-778.
[26] LIN T Y, DOLLÁR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA, 2017:936-944.
[27] NEUBECK A, VAN GOOL L. Efficient non-maximum suppression[C]//18th International Conference on Pattern Recognition. Hong Kong, China, 2006:850-855.
[28] XIE Saining, GIRSHICK R, DOLLÁR P, et al. Aggregated residual transformations for deep neural networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition. Honolulu, USA, 2017:5987-5995.

备注/Memo

备注/Memo:
收稿日期:2018-06-01。
基金项目:国家自然科学基金项目(61672280,61373060,61732006);江苏省333高层次人才培养工程(BRA2017377);青蓝工程.
作者简介:魏文戈,男,1993年生,硕士研究生,主要研究方向为图像识别和深度学习;谭晓阳,男,1971年生,教授,主要研究方向为模式识别、计算机视觉和机器学习。江苏省高校"青蓝工程"中青年学术带头人,中国计算机学会会员,中国计算机学会计算机视觉专业组委员,中国计算机学会人工智能与模式识别专委会委员,江苏省计算机协会人工智能专委会委员。主持多项科研项目,曾获国际电气与电子工程师协会信号处理协会最佳论文奖、教育部自然科学奖二等奖、国家自然科学奖二等奖、教育部自然科学奖一等奖、国际学术期刊《Pattern Recognition》 2006-2010 年度高引用论文奖等。发表学术论文50余篇,被引用近5000次。
通讯作者:谭晓阳.E-mail:x.tan@nuaa.edu.cn
更新日期/Last Update: 2019-08-25