基于卷积神经网络模型的木材宏、微观辨识方法

doi:10.11707/j.1001-7488.20210615

Abstract

Abstract:

Objective: In order to effectively improve the accuracy and speed of wood identification, a new method of wood identification based on PWoodIDNet model, a convolutional neural network model, was proposed with expects to provide advanced identification method and instruments for customs, import and export quarantine inspection, furniture enterprises and other legal departments and enterprises, and to promote the scientific and technological progress of China's wood import and export quarantine inspection industry and wood processing and manufacturing enterprises. Method: Firstly, 16 kinds of wood samples were selected, and 50 high-resolution microscopic CT images and industrial camera images were obtained from each sample with a total of 1 600 images. Then, a total of 4 800 target regions with wood rays, parenchyma, axial tracheids, pits and textures were intercepted. The image set was expanded to 19 200 images through image enhancement algorithms, such as horizontal flipping, vertical flipping, mirroring and brightness transformation. The PWoodIDNet model for the identification of tree species based on convolutional neural network was constructed. The stochastic gradient descent (SGDM) method with momentum added was used to optimize the model, and GPU was used to optimize the parallel operation library. The classification accuracy of wood macro- and micro-structure data sets was compared. Result: Compared with the existing GoogLeNet identification method, the accuracy and speed of PWoodIDNet model were increased by 1.49% and 59.69%, respectively. Compared with the existing AlexNet identification method, the accuracy and speed of PWoodIDNet model were increased by 3.76% and 2.63%, respectively. Conclusion: PWoodIDNet model could break through the difficulties of existing identification methods, such as narrow range of wood species, low accuracy and low identification speed, effectively identify wood, and could achieve the best identification effects in less training time. It would be expected to provide a new method and thought for the identification of wood in China.

Key words: wood identification, convolutional neural network, micro-structure, macro-structure

CLC Number:

Ziyu Zhao,Xiaoxia Yang,Hui Guo,Zhedong Ge,Yucheng Zhou. Recognition Method of Wood Macro- and Micro-Structure Based on Convolution Neural Network[J]. Scientia Silvae Sinicae, 2021, 57(6): 134-143.

Figures/Tables 11

Fig.1

Fig.2

Table 1

Fig.3

Fig.4

Table 2

Fig.5

Fig.6

Table 3

Fig.7

Table 4

References 0

	陈龙现, 葛浙东, 罗瑞, 等. 基于CNN的木材内部CT图像缺陷辨识. 林业科学, 2018, 54 (11): 130- 136.
	Chen L X , Ge Z D , Luo R , et al. Identification of CT image defects in wood based on convolution neural network. Scientia Silvae Sinicae, 2018, 54 (11): 130- 136.
	费本华, 赵勇, 覃道春, 等. 应用CT技术研究木材断口形态特征. 林业科学, 2007, 43 (4): 137- 140.
	Fei B H , Zhao Y , Qin D C , et al. Applying computerized tomography(CT)to study the feature of wood fracture. Scientia Silvae Sinicae, 2007, 43 (4): 137- 140.
	李楠. 2018. 一种基于卷积神经网络的轻量级木材图像识别模型研究. 杭州: 浙江农林大学硕士学位论文.
	Li N. 2018. A lightweight wood image recognition model based on convolutional neural networks. Hangzhou: MS thesis of Zhejiang Agriculture and Forestry University. [in Chinese]
	刘英, 周晓林, 胡忠康, 等. 基于优化卷积神经网络的木材缺陷检测. 林业工程学报, 2019, 4 (1): 115- 120.
	Liu Y , Zhou X L , Hu Z K , et al. Wood defect recognition based on optimized convolution neural network algorithm. Journal of Forestry Engineering, 2019, 4 (1): 115- 120.
	刘子豪, 祁亨年, 张广群, 等. 基于横切面微观构造图像的木材识别方法. 林业科学, 2013, 49 (11): 116- 121.
	Liu Z H , Qi H N , Zhang G Q , et al. Wood identification method based on microstructure images in cross-section. Scientia Silvae Sinicae, 2013, 49 (11): 116- 121.
	彭冠云, 江泽慧, 覃道春, 等. 基于CT技术的人造板剖面密度检测. 木材工业, 2009, 23 (5): 19- 21. doi: 10.3969/j.issn.1001-8654.2009.05.006
	Peng G Y , Jiang Z H , Qin D C , et al. Density profile determination of wood-based panels using X-ray computer tomography. China Wood Industry, 2009, 23 (5): 19- 21. doi: 10.3969/j.issn.1001-8654.2009.05.006
	项宇杰, 陈月芬, 卢卫国, 等. 基于深度卷积神经网络的木材表面缺陷检测系统设计. 系统仿真技术, 2019, 15 (4): 253- 257. doi: 10.3969/j.issn.1673-1964.2019.04.008
	Xiang Y J , Chen Y F , Lu W G , et al. Wood surface defect detection system design based on deep convolutional neutral network. System Simulation, 2019, 15 (4): 253- 257. doi: 10.3969/j.issn.1673-1964.2019.04.008
	朱佳, 汪杭军. 基于Graph Cuts的木材扫描电镜图像特征提取方法. 林业科学, 2014, 50 (4): 111- 117.
	Zhu J , Wang H J . Feature extraction method of wood SEM micrographs based on Graph Cuts. Scientia Silvae Sinicae, 2014, 50 (4): 111- 117.
	Barmpoutis P , Dimitropoulos K , Barboutis I , et al. Wood species recognition through multidimensional texture analysis. Computers & Electronics in Agriculture, 2018, 144, 241- 248.
	Chen L , Papandreou G , Kokkinos I , et al. DeepLab: semantic image segmentation with deep convolutional nets, atrous convolution, and fully connected CRFs. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2017, 40 (4): 834- 848.
	Dimauro G , Ciprandi G , Deperte F , et al. Nasal cytology with deep learning techniques. International Journal of Medical Informatics, 2019, 122, 13- 19. doi: 10.1016/j.ijmedinf.2018.11.010
	He T , Liu Y , Xu C , et al. A fully convolutional neural network for wood defect location and identification. IEEE Access, 2019, 7, 453- 462.
	Hinton G , Tieleman T . Lecture 6. 5-rmsprop: divide the gradient by a running average of its recent magnitude. COURSERA: Neural Networks for Machine Learning, 2012, 4, 26- 31.
	Kingma D P , Ba J L . Adam: a method for stochastic optimization. Computer Science, 2014, 1- 15.
	Kiranyaz S , Ince T , Gabbouj M , et al. Real-time patient-specific ECG classification by 1-D convolutional neural networks. IEEE Transactions on Biomedical Engineering, 2016, 63 (3): 664- 675. doi: 10.1109/TBME.2015.2468589
	Krizhevsky A , Sutskever I , Hinton G E . ImageNet classification with deep convolutional neural networks. International Conference on Neural Information Processing Systems, 2012, 25 (2): 1097- 1105.
	Lecun Y , Boser B , Denker J , et al. Backpropagation applied to handwritten zip code recognition. Neural Computation, 1989, 1 (4): 541- 551. doi: 10.1162/neco.1989.1.4.541
	Rumelhart E D , Hinton E G , Williams J R . Learning representations by back-propagating errors. Nature, 1986, 323, 533- 536. doi: 10.1038/323533a0
	Szegedy C, Liu W, Jia Y, et al. 2015. Going deeper with convolutions. Proceedings of the IEEE conference on computer vision and pattern recognition, June 8-10, Boston, 1-9.
	Traoreb B B , Kamsu-Foguem B , Tangara F . Deep convolution neural network for image recognition. Ecological Informatics, 2018, 48, 257- 268. doi: 10.1016/j.ecoinf.2018.10.002
	Yang S Y , Lee H G , Park Y , et al. Wood species classification utilizing ensembles of convolutional neural networks established by near-infrared spectra and images acquired from Korean softwood lumber. Journal of the Korean Wood Science and Technology, 2019, 47 (4): 385- 392.

木材种类 Wood species	尼泊尔紫檀 Gluta sp.	纳托山榄 Palaquium sp.	黄杨木 Buxus sp.	轻木 Ochroma sp.	南美红鸡翅 Andira sp.
微观构造 Micro-structure 宏观构造 Macro-structure
木材种类 Wood species	东非黑黄檀 Dalbergia melanoxylom	红栎 Quercus sp.	普拉藤黄 Platonida insignls	枫木 Acer sp.	非洲檀香 Spirostachys africana
微观构造 Micro-structure 宏观构造 Macro-structure

优化算法 Optimization algorithm	平均分类准确率 Average classification accuracy(%)	平均损失值 Average loss value	训练步数 Training steps
SGDM	99.17	0.025 4	4 200
Adam	97.45	0.119 2	4 200
RMSprop	96.31	0.211 6	4 200

辨识方法 Identification methods	平均分类准确率 Average classification accuracy(%)	平均损失值 Average loss value	模型训练时间 Model training time/s	训练步数 Training steps
PWoodIDNet-微观Micro	99.17	0.025 4	2 243	4 200
PWoodIDNet-宏观Macro	99.11	0.025 8	2 281	4 200
GoogLeNet-微观Micro	98.81	0.036 2	5 696	4 200
GoogLeNet-宏观Macro	96.49	0.113 7	5 743	4 200
AlexNet-微观Micro	96.46	0.112 3	2 308	4 200
AlexNet-宏观Macro	94.30	0.180 9	2 338	4 200

辨识方法 Identification methods	网路层数 Network layers	网络连接数 Number of network connections	平均分类准确率 Average classification accuracy(%)	平均损失值 Average loss value	模型训练时间 Model training time/s	训练步数 Training steps
PWoodIDNet	72	83	99.17	0.025 4	2 243	4 200
GoogLeNet Inception v1	144	170	98.81	0.036 2	5 696	4 200
GoogLeNet Inception-ResNet v2	825	922	96.16	0.289 3	37 843	4 200
GoogLeNet Inception v3	315	349	97.69	0.132 7	19 770	4 200

[1]	Lin Zhao,Xiaoli Zhang,Yanshuang Wu,Bin Zhang. Subtropical Forest Tree Species Classification Based on 3D-CNN for Airborne Hyperspectral Data [J]. Scientia Silvae Sinicae, 2020, 56(11): 97-107.
[2]	Guangqun Zhang,Yingjie Li,Hangjun Wang,Houkui Zhou. Forest Image Classification Based on Fine-Tuning CaffeNet [J]. Scientia Silvae Sinicae, 2020, 56(10): 121-128.
[3]	Zhiwei Lin,Qilu Ding,Jinfu Liu. Bird Species Identification Based on Deep Convolutional Network with Fusing Global and Local Features [J]. Scientia Silvae Sinicae, 2020, 56(1): 133-144.
[4]	Li Yanmei, Chen Qibo, Wang Shaojun, Sun Yingdu, Yang Haozhou, Yang Siying. Effects of Leaf Surface Micro-morphology Structure on Leaf Dust-Retaining Ability of Main Greening Tree Species in Kunming City [J]. Scientia Silvae Sinicae, 2018, 54(5): 18-29.
[5]	Liu Zihao, Wang Hangjun. Wood Identification Based on Feature Fusion of PCA and FisherTrees [J]. Scientia Silvae Sinicae, 2013, 49(6): 122-128.
[6]	Liu Zihao, Qi Hengnian, Zhang Guangqun, Wang Hangjun. Wood Identification Method Based on Microstructure Images in Cross-Section [J]. Scientia Silvae Sinicae, 2013, 49(11): 116-121.
[7]	Zhao Zhongsheng;Cui Yongzhi;Yu Haipeng;Liu Yixing;Shen Jun. Micro-Structure Examination of Strengthened Wood Sheets by Compression Perpendicular to Grain [J]. Scientia Silvae Sinicae, 2010, 46(11): 124-130.
[8]	Zhang Lifei;Jiang Xiaomei;Zhou Yin. A STUDY ON CRYSTALS IN WOOD SPECIES OF CHINESE ANGIOSPERM [J]. , 1993, 29(5): 424-428.
[9]	Research team of computer-aided identification of wood. COMPUTER-AIDED IDENTIFICATION FOR CHINESE HARD-WOODS [J]. , 1989, 25(3): 236-242.

Recognition Method of Wood Macro- and Micro-Structure Based on Convolution Neural Network

RichHTML

PDF (PC)

Knowledge

Abstract

Cite this article

share this article

Figures/Tables 11

References 0

Related Articles 9

Recommended Articles

Metrics

Comments