共查询到20条相似文献，搜索用时 265 毫秒
1.
基于MEA-BP神经网络对木材内部缺陷诊断的研究 总被引：1，自引：0，他引：1
为了提高木材内部缺陷的自动识别率,采用电阻层析成像(ERT)的方法获取电导率波动信号,通过小波包变换对采集的数据进行3层小波包分析,对八维特征向量进行提取,利用思维进化算法(MEA)优化权值和阈值,孔洞、节子、腐朽试样各45组数据,进行BP神经网络训练,每种缺陷20组作为测试集,识别木材内部缺陷。结果表明:MEA-BP神经网络对木材孔洞、节子和腐朽的识别率分别为96.92%、95.38%和92.31%,该模型解决了复杂组合的优化问题,提高了搜索效率,并且达到最佳的预测效果。 相似文献
2.
3.
基于小波神经网络的木材内部缺陷类型识别的研究 总被引：7，自引：0，他引：7
利用榆木标准试件,在实验室内用超声波检测仪器对试件进行缺陷分类检测,检测信号作为原始信息.各类试件的原始信号用小波包分解,计算缺陷试件与完好试件在小波包第5层各结点的信号能量变化值.试验发现:木材缺陷引起能量的变化值主要由木材缺陷的大小或严重程度来决定,亦即木材的缺陷程度越严重,能量的变化幅度就越大;对小波包5层分解后各信号结点的能量变化值进行分析,发现在32个结点中,(5,0)结点在各类缺陷试件中能量值变化最大;使用经小波压缩后的信号作为神经网络的输入,形成应用频带能量变化值和应用(5,0)结点小波包系数的2个不同输入特征的人工神经网络.对比分析2个网络识别木材缺陷类型的能力,(5,0)结点小波包系数作为特征训练得到的神经网络检测精度更高. 相似文献
4.
基于Faster R-CNN模型的木材表面节子缺陷检测 总被引：1，自引：0，他引：1
《木材工业》2020,(2)
研究基于Faster R-CNN模型的木材表面节子缺陷检测方法,解决现有方法存在的特征选取困难、不能适应木材以及节子的多样性变化、检测精度不高等问题,而且能够给出节子缺陷的位置以及图像的大小。小样本集测试结果表明,该方法能够取得较高的检测精度,准确率为94.0%;对其中4个典型样本进行分析,位置检测最大误差仅为5个像素,大小检测最大误差仅为7个像素;相比OSTU方法具有更好的检测精度。 相似文献
5.
木材缺陷检测理论及方法的发展 总被引：4，自引：0，他引：4
通过对国内外板材节子视频检测理论现状和发展的研究，对计算机视频技术在木材缺陷检测上的应用与发展进行了综述，并对视频技术在木材缺陷检测上的应用前景进行展望。 相似文献
6.
基于Faster R-CNN的实木板材缺陷检测识别系统 总被引：1，自引：0，他引：1
我国木材资源有限,为了提高木材的利用率,采用机器视觉来实现木材缺陷快速而稳定的检测,不仅可以克服人工检测的低效率和木材缺陷识别的低准确率,而且对提高木材加工企业的智能化水平具有重要意义。为了高效、快速、准确地进行无损检测,采用深度学习方法,建立了一种基于快速深度神经网络的实木板材缺陷识别模型。首先采用Resnet V2结构对采集到的实木板材缺陷图像进行特征提取,然后应用该模型对节子、孔洞等实木板材缺陷进行训练学习,最后构建了Faster R-CNN检测框架,并使用tensorflow开发平台对节子、孔洞等实木板材缺陷进行预测输出。具体选取了2 000块杉木样本,通过旋转对原始的实木板材图像进行数据扩充,扩充后图像的80%作为训练集,20%作为验证集来进行仿真。仿真结果表明,该模型对实木板材节子缺陷检测正确率为98%,对实木板材孔洞缺陷检测正确率为95%,验证了将深度学习算法应用于实木板材缺陷检测中的有效性。 相似文献
7.
8.
为了探讨振动法测量木材动态弹性模量时应力波波速的差异性,本实验选取含有不同节子及内部孔洞的国产油松、马尾松原木为试验材料,采用纵向敲击法测量原木试材的动态弹性模量。建立原木缺陷的数学模型,模拟计算和分析表明,木材节子及孔洞对应力波波速及弹性模量均有影响,而且降低了木材品质。 相似文献
9.
综述了数字图像处理技术在木材缺陷检测、木材结构和木材美学三个方面的研究进展,分析了其中存在的不足。结果表明,数字图像处理技术可实现木材虫眼、节子和腐朽等缺陷的快速准确识别;木材结构的图像分析对于气候变化、木材长势及力学性能研究均有一定的参考价值;利用数字图像处理技术可以提取出木材独一无二的纹理图案。展望了数字图像处理技术在木材科学中的应用前景,以及其在木材缺陷自动检测、物种识别和美学研究等方面的商业价值。 相似文献
10.
【目的】针对木材蛀干害虫羽化孔洞缺陷检测问题,通过对声发射信号的时频分析,研究木材蛀干害虫羽化孔洞缺陷的AE信号特征。【方法】首先,对无孔洞和3种不同尺寸的钻孔缺陷的木材试件,参照ASTM-E976标准采用铅芯折断方式产生AE源,通过采样频率为500 kHz的2通道木材声发射信号采集系统获取原始AE信号。然后,对原始AE信号进行降噪滤波,再对滤波后的信号进行小波分解并重构AE波形,分析孔径对AE信号频率分布的影响。最后,采用信号相关性分析法和时差定位法,计算AE信号的传播速率,进而分析羽化孔径对传播速率的影响。【结果】AE信号通过无孔洞、5 mm孔洞、8 mm孔洞、15 mm孔洞缺陷,AE信号的主频率分别为38、43、51、117 kHz,AE信号传播的平均速率分别为1 380.7、1 067.3、848.6、437.1 m/s。【结论】AE信号在不同孔径缺陷的木材试件传播时,随着孔洞缺陷的增大,AE信号的幅值发生明显衰减,AE信号的主频率增大,AE信号传播的平均速率减小。为了验证试验结果和结论,找了具有天然蛀干害虫羽化孔洞缺陷的木材试件,孔洞直径约为20 mm,使用相同的方法采集和处理AE信号,AE信号通过天然蛀干害虫羽化孔洞缺陷的主频率为121 kHz,AE信号传播的平均速率为324 m/s。 相似文献
11.
《Scandinavian Journal of Forest Research》2012,27(3):311-315
Both foresters and sawmillers are interested in the knot structure of trees; in particular, position and number of knots, knot diameter, knot length and dead knot border. For research purposes, it is possible today to carry out non‐destructive measurements using computer tomography (CT) and image analysis. The aim of this study was to measure knot parameters on Norway spruce (Picea abies (L.) Karst.) using a non‐destructive method developed for Scots pine (Pinus sylvestris L.), and to compare the results of this method with the results of two different destructive methods. In order to do this, two Norway spruce stems were scanned by CT. Then five logs from one stem were cut into flitches 20 mm thick and the defects on the sawn surfaces were scanned manually. The other stem was cut just above every whorl and then each knot was split through its centre and the knot parameters were measured manually. The study showed that the CT method compares well with the destructive methods. It is a reasonably fast, non‐destructive method which measures position and diameter of knots and detects larger knots with acceptable accuracy. The study also showed that a large number of smaller knots were not found by the CT method and that the CT method measured knot length and dead knot border with low accuracy. This means that the CT method has to be adjusted to Norway spruce in order to improve its ability to measure knot length and dead knot border and to detect smaller knots. 相似文献
12.
《Scandinavian Journal of Forest Research》2012,27(1):50-56
In an attempt to develop a better understanding of the effect of knottiness on the yield of logs for sawn timber products, the present study examines the position and geometrical characteristics of 429 knots found in seven Norway spruce (Picea abies (L.) Karst) middle logs from southern Finland. The raw data used in this study were in the form of coordinates of points located on the knot surface recorded in a cylindrical (R, L, T) reference frame. The data were smoothed using a non linear three‐dimensional model. A curvature analysis of the knot pith was used to differentiate the curved and straight portions of the knots. This analysis made possible calculation of knot angles, symmetry, ovality and volume. The major finding of this study is the uneven distribution of knots around the log circumference. Nearly 50% of the total knot volume was concentrated in one‐third of the log facing the south‐east. Knots were also more prominent in higher portions of the logs studied. These observations support the argument that the uneven distribution of knots offers potential for improving the grade yield of timber sawn from trees harvested in high‐latitude countries. 相似文献
13.
C. Foley 《Wood Science and Technology》2001,35(5):453-465
Knowledge of the three-dimensional orthogonal directions of wood material at any position within a tree is necessary for
the understanding of strength reducing effects of knots and essential for the continuation of research in areas which relate
small clear wood specimen behavior to the behavior of full size structural timber. A complete three-dimensional paradigm describing
the geometry of knots and related fiber distortion, initially derived to predict the strength-reducing behavior of knots in
structural timber of Norway Spruce with the finite element method, is presented in this article. Besides strength prediction
analyses, it is believed that the paradigm may be useful in other areas of research on structural timber that are effected
by fiber orientation, such as drying and form change of structural timber. The paradigm generates fiber orientation in any
position within a log or lumber from assumed fiber patterns in planes parallel to the longitudinal direction of the original
tree. Fiber patterns in the radial and tangential directions are derived from physical restraints related to fiber production
within the annual increase surfaces of the tree and from theories of knot formation. The adaptability of the paradigm allows
practically any softwood knot to be modeled with an accuracy that is limited only by input-data. The knot-axis may be non-linear,
and the knot cross-section oval with its vertical and horizontal axis increasing from the pith of the stem at chosen rates.
Spiral grain may also be included in the paradigm and vary with the annual growth layers. Investigations presented in this
article showed that generated fiber orientations for Picea abies complied well with measured fiber distortions, and that the general trends of fiber orientation, explained by the applied
knot formation theory, is reflected in the measured specimens.
Received 12 May 1999 相似文献
14.
Summary Superposition of the longitudinal stress distributions associated with neighboring knots was used to determine the interaction distances between different knot sizes. The interaction between two knots was considered significant when the superposition of the stress distribution caused an increase in the longitudinal stress level of ten percent or more. Out of 4,374 knots measured in 44 Douglas-fir poles only 55 knots showed significant interactions with adjacent knots. However, in more than half of the poles at least one interacting knot pairs existed. Therefore, it has been concluded that stress interaction between knots could play a major role in controlling the failure of poles in bending. 相似文献
15.
Jianjun Zhu Tatsuo Nakano Morihiko Tokumoto Takashi Takeda 《Journal of Wood Science》2000,46(4):284-288
To examine the effectiveness of long rotation forestry and the potential of complete utilization of Japanese larch (Larix kaempferi Carriere), we designed a tensile test using the lumber from six 87-year-old sample trees. Results showed that strength properties of lumber varied greatly in the radial direction within trees, but all sample trees showed a similar trend. There was little difference in dynamic Young's modulus but a large difference in tensile strength (TS) between the lumber and small clear specimens from undestroyed parts of the lumber. These differences decreased with an increase in ring number and became constant after 30 years. The presence and distribution of knots markedly affected the TS; and among the knot indices, the knot number (Kn) and knot area ratio of a maximum single knot (Km) proved to be effective for explaining the effect of knots. The distribution of Kn and Km in the radial direction agreed with the variation of TS in the radial direction. By investigating the variation patterns of lumber and small clear specimens in the radial direction, it was found that the strength properties of both required a long time, about 30 years, to reach a relatively constant state.Part of this report was presented at the 49th annual meeting of the Japan Wood Research Society, Tokyo, April 1999 相似文献
16.
M. Samson 《Wood Science and Technology》1993,27(6):429-437
A geometrical model was derived to describe knots in logs and on the surface of lumber beams sawn from those logs. Each knot is defined by 7 parameters related to the shape and position in the stem. A computer simulation program was written to study knot shapes on a variety of observation planes.An initial investigation on the shape of knots in Scots pine logs and lumber beams suggests that the model is sufficiently accurate to describe knottiness in this species. Potential applications of this model include automated lumber grading, computerized log reconstitution and yield optimization studies.This research was carried out when the author was a Visiting Scientist at the Technical Research Centre of Finland (VTT), Forest Products Laboratory, P.O. Box 207, SF-02151 Espoo, Finland. Technical assistance of U. Saarelainen, A. Usenius and C. O. Sommardahl from VTT is gratefully acknowledged. This work was partially supported by Forestry Canada and the Natural Sciences and Engineering Research Council of Canada 相似文献
17.
Javad Torkaman Mojgan Vaziri Dick Sandberg Soleiman Mohammadi Limaei 《Wood material science & engineering》2018,13(2):117-120
The classification of roundwood is inextricably linked to the measurement of a particular single wood defect. The appearance, location, and number of defects are important in the quality evaluation of logs and sawn timber, and the most important defects are knots. The purpose of this study was to investigate the relationship between the appearance of branch scars and features of the related knot inside oriental beech logs, and to model the relationship between well-defined branch-scar and knot parameters. One hundred and fifty knots in 15 stems of oriental beech trees were studied. Image analysis software was used to measure the branch-scar and knot features. The results showed a significant positive correlation between the branch-scar parameter “moustache length” and the knot length. The ratio of branch-seal length to width was found to be a good estimator of the stem diameter at the time of knot occlusion and the amount of clear wood between the knot occlusion and the bark. The relationship obtained for the oriental beech stem radius at time of knot occlusion confirms relationship reported for European beech (Fagus sylvatica L.). 相似文献
18.
19.
Claudia B. Cáceres Svetka Kuljich Ahmed Koubaa 《Wood material science & engineering》2018,13(1):28-35
Fifteen stems of jack pine (Pinus banksiana Lamb.) of 3 commercial thinned plots (control, moderate, and intensive thinning) and 15 stems of black spruce (Picea mariana (Mill.) B.S.P), both coming from the Abitibi-Témiscamingue region – Canada, were cross-cut into three 2.4?m length sections: bottom, middle, and top logs. Logs were processed with a chipper-canter at three cutting widths (12.7, 19.1, and 25.4?mm), producing chips and a three-faced cant. The middle section of the cant was used to evaluate surface quality across the grain on each face. Roughness and waviness parameters and depth of torn grain were recorded. Knot characteristics were assessed in the three cant faces. Poorer surface quality was found in the lower part compared to the upper part of the cant for both species. At larger cutting widths, jack pine logs coming from a natural stand showed lower surface quality compared to logs from thinned stands. Black spruce waviness increased with the cutting width and stem height. These results were attributed to the increase of forces and vibration when cutting at larger cutting widths, which was worsened by the presence of bigger, more numerous knots at the control plot and in the top logs. Black spruce had deeper torn grain compared to jack pine. Their differences in knot characteristics resulted in a maximum torn grain depth favored by the presence of more knots rather than by bigger knot size. Other knot characteristics, such as the position of the knots in the cant face, the insertion angle of the branch and the distribution of the knots around the log, should be studied to better understand the relationship between torn grain formation and knottiness. 相似文献
20.
A new method has been developed for detecting localized defects such as edge knots using a bending deflection curve. The coordinates
of a bottom edge (edgeline) of an unloaded piece of lumber are extracted from a digital image, and a bending deflection curve
is obtained from the displacement of the edgeline of the lumber using a digital image correlation (DIC) technique. Depending
on the knots within the beam, the bending deflection curve is shifted from the curve of a defect-free beam. The measured bending
deflection curve is regressed to a theoretical curve by elementary beam theory. A finite element method (FEM) model of the
beams including defects as simplified knot structure has been performed. Comparison between the bending experiment and FEM
analysis shows that cross-sectional reductions cause characteristic variations in the bending deflection curves depending
on the position of encased knots, and local grain distortions cause variations in the curves depending on the direction of
spike knots. Using the residual variance between the measured deflection curve and a polynomial regression curve, it is possible
to detect knots at which failures initiate.
Part of this article was presented at the 57th Annual Meeting of the Japan Wood Research Society, Hiroshima, Japan, August
2007 相似文献