基于BP和RBF神经网络的木材缺陷检测研究
|
摘要
木材无损检测是一门综合性的、非破坏性的木材缺陷的检测方法。在不损坏木材表面和结构的前提下,准确地检测出木材的缺陷。能够确保合理选材、科学用材,提高木材的使用率和经济价值,有效节约木材资源。本课题以北方典型的两种树种：落叶松和桦木(它们分别属于针叶材和阔叶材)为研究对象,对它们中的节子、腐朽、虫害三种典型木材缺陷类型进行了具体的研究。
研制了木材X射线的无损检测系统和木材缺陷图像的采集系统的硬件平台。采用X射线作为检测手段,对木材进行无损检测。通过检测透过被检物体后的射线差异,来判断被检测木材是否有缺陷存在。不仅可以检测木材的表面缺陷,还可以检测木材的内部缺陷。在木材的另一端利用图像增强器进行接收,再经过微光摄像机送入A/D转换器,将木材X射线模拟图像转换成一数字图像存入计算机中,完成木材的缺陷图像的采集工作。
应用数字图像处理技术对采集到的木材缺陷图像进行预处理。对三种木材缺陷图像分别进行了灰度变换增强、改进的加权均值滤波处理、中值滤波处理。处理后的图像对比度明显增强,没有出现传统均值滤波加宽的现象,最大程度地保留了图像的缺陷细节,易于后续的图像特征提取。运用常用的几个边缘检测算子：Roberts算子、Sobel算子、Prewitt算子、Log算子、Canny算子对木材缺陷图像进行边缘检测,检测结果表明Sobel算子检测效果较好,速度较快。为了提高边缘提取的效果,对木材缺陷的原始图像、灰度变换增强后的图像、加权均值滤波后的图像、中值滤波后的图像分别进行二值化处理,对加权均值滤波后的图像的二值化处理效果最佳,对该二值图像缺陷区域进行填充处理。应用Sobel算子对填充后的二值图像进行边缘提取,提取出清晰的木材缺陷边缘。
提出了一种通过对图像实现木材缺陷图像分割的新方法。首先对效果最佳的木材缺陷的二值图像进行图像翻转,将其数组转化为uint8类型,然后将翻转后得到的图像与中值滤波后的木材缺陷图像进行,木材缺陷图像的结果使缺陷区域完全从背景中分离,完成了图像分割,同时并没有改变缺陷区域的灰度、结构等特征。
木材缺陷在特征的选择上具有四个特点：可区别性,对不同类别的对象,特征要具有明显的差异性；可靠性,对同类别的对象,特征具有相近性；独立性,所用的特征之间互不干涉,互不相关；数量少,数量多虽然能更好的区别不同的对象,但是计算量增大,运算时间增长,同样会使识别准确率下降。根据节子、腐朽、虫害的缺陷特征差异,主要提取木材缺陷图像的灰度特征和形状特征。首先对边缘提取后的木材缺陷图像进行扫描,记录缺陷区域的边缘的坐标,确定了木材缺陷的位置和尺寸,再根据确定的位置和尺寸分别对三种木材缺陷的二值图像和分割后的图像进行特征提取,提取了木材缺陷长宽比、圆形度、灰度均值、灰度方差4个特征量。结合节子、腐朽、虫害各自的结构特点,引入了矩及几何中心矩,通过对几何矩的非线性组合,得到一组对于图像平移、旋转和比例不变的矩—Hu不变矩。根据Hu不变矩的物理意义,对原有的Hu不变矩进行抽象,由原来的7个特征抽象为10个特征,满足了图像的平移、缩放和旋转不变性。提取出了10个木材缺陷的结构特征值。这样对三种典型缺陷各提取了14个木材缺陷的特征值,对这14个特征进行归一化处理,归一化后的特征值将作为神经网络输入的特征向量。
设计了BP神经网络模型,确定BP网络的层数及各层神经元数目,隐层和输出层选择S型函数作为激励(传递)函数,通过对BP各种算法的比较,得出Levenberg-Marquardt学习算法是最优的算法,用已知样本对BP神经网络进行学习,将训练成熟的网络对未知样本进行仿真,BP神经网络对木材缺陷的识别准确率达到90%。
构建了RBF神经网络模型,RBF神经网络对数据具有较好的逼近能力。RBF神经网络为三层网络,确定了各层的节点数,隐层传递函数为高斯函数,输出层传递函数为线性函数,经过反复试验确定了径向基分布常数spread,用已知样本对RBF神经网络进行训练,训练完成后,对未知样本进行仿真试验,对木材缺陷的识别准确率达到92%。
组建了BP-RBF混合神经网络模型。BP神经网络具有较好的数据压缩能力,RBF神经网络对数据具有较好的逼近效果,将BP神经网络和RBF神经网络串联。BP-RBF混合神经网络中的BP神经网络部分的输出作为RBF神经网络的输入,采用分布训练的方法,先对BP神经网络部分进行训练,然后再对RBF网络进行训练。由于BP-RBF混合神经网络中的BP网络部分与前面构建的BP网络结构相同,已训练成熟,不需要再进行训练,但混合神经网络中的RBF网络部分与前面的RBF神经网络有所不同,对RBF网络部分进行训练,重新确定了径向基分布常数spread,待RBF网络部分训练达到精度要求后,整个BP-RBF混合神经网络训练完成,对未知样本进行识别。对样本的识别准确率达到96%。三种神经网络的木材缺陷识别率进行比较,BP-RBF混合神经网络识别准确率最高,而且实际输出值更接近于目标输出值。
实验结果表明,应用BP和RBF神经网络方法可以成功地对木材的三种典型缺陷进行无损检测,木材缺陷的识别率都在90%以上,组建的BP-RBF混合神经网络模型对木材缺陷识别的精度高,更加有效。此方法为实现木材缺陷的自动化检测提供了重要的理论依据。
Wood nondestructive testing is a synthetic, nondestructive detecting method. It can detect the wood defects accurately without destroying the surface and structures of the wood. So it can make sure to select wood properly, use wood effectively, which will enhance the occupating coefficient and economic value of the wood and save the wood resource effectively. This thesis studies the two typical kinds of north woods:larch and birch, which belong to softwood and broad-leaved wood. The three typical wood defects:knot, rot and grub-hole have been studied in detail.
The X-ray wood nondestructive testing system, hardware platform image acquisition for wood defect testing is invented. The wood nondestructive testing has been done using X-ray method. The defects can be found by detecting the X-ray difference between the rays fore and after transmitted the wood. This method can not only detect the surface defects of wood, but also the internal defects. The signals are received by the image enhancement equipment on the other end of the wood, and then are transmitted into A/D converter, which convert the X-ray analog pattern of wood into digital images and are stored into computer. This is the all process of image acquisition for wood defect.
Digital image processing is applied for preprocessing of wood defects images. Gray transform enhancement, improved weighted mean filtering and median filtering are applied for the wood defects images of these three kinds of trees discussed above. The results show that the contrasts of these images enhance apparently, and there is no the widened mean filtering traditionally, the image defect details are reserved maximatily, and it is easy for the image feature extraction. The several commonly used edge detection operators:Roberts operator, Sobel operator, Prewitt operator, Log operator and Canny operator, are used for edge detection of the wood defect images. The detect result shows that Sobel operator is better and faster. In order to improve the effect of edge detection, binarization processing has been applied for the original images, the gray transform enhancement images, the weighted mean filtering images and the median filtering images of wood defects respectively, which found that the weighted mean filtering images is the best. Then the defect areas of binary images of wood defects are filled up. For extracting the clear edge of wood defects, the filled binary images are edge extracted with Sobel operator.
A new method for segmenting the wood defects images by image adduct has been presented. First, invert the best binary image of wood defect, and convert its array into uint8 type, and then adduct the image obtained after median filtering with the defect image, which is turned over. The result of wood defect image plus is that defects regions take apart completely with their background, which means the image segmentation is completed. At the same time the characteristics of the defect regions, such as shade of gray, structure, do not changed.
The choice of wood defects has four characteristics:distinctiveness, which means there should be an obvious difference for different objects; reliability, which means features should be similar for the same types of objects; independence, which means the characteristics used should be non-interacting, unrelated; small quantity, although big quantity can make it easier to distinguish different objects, but the amount of computation and the computation time are increased, which will decrease the identify accuracy. According to feature differences of knot, rot, grub-hole, the gray and shape characteristics of wood defects are extract mainly. First, scan the wood defects images after the edge extraction, record the coordinates of the edge of defect regions, determine the location and size of wood defects, then feature extraction is applied for binary images and segmented images of three kinds of wood defects. Four characteristic quantities have been extracted for wood defects:aspect ratio, circular degrees, the mean gray, gray variance. Combined the structural characteristics of knot, rot and grub-hole, geometric moment is introduced. By non-linear combination of geometric moment, a set of moment:Hu invariant moment, which is invariant for image translation, rotation and proportion is derived. According to the physical meaning of Hu invariant moment, the original Hu invariant moments are abstracted, its seven features change into ten features, but still invariant for image translation, convergent-divergent and rotation. Ten structural characteristics of wood defects are extracted. So the fourteen characteristic values of wood defects are extracted for these three kinds of typical defects, and are normalized. The normalized eigenvalues will serve as the input feature vectors of neural network.
The BP neural network model has been designed, determined the number of layers of BP network and neuron numbers for each layer. The hidden layer and output layer select S-function as an incentive (transfer) function. Through the comparison of various algorithms of BP network, it is found that Levenberg-Marquardt learning algorithm is the optimal algorithm. Learn the BP neural network with known samples, simulate the unknown samples with this mature network, which find that the accuracy rate of this BP neural network for wood defects recognition is as high as 90%.
RBF neural network model has been constructed, RBF neural network has good approximation for the data. RBF neural network is three-layer network, identified the nodes of each layer. Hidden layer transfer function belongs to Gaussian function, the output layer transfer function is a linear function, after repeated testing, radial based distribute function "spread" is determined. The RBF neural network is trained with known samples, and then simulation is applied on unknown samples. It is found that the accuracy rate of this network for wood defects recognition is as high as 92%.
Set up a new BP-RBF mixed neural network model. BP neural network has a better data compression capability, RBF neural network has a good approximation to the data results, will be BP neural network and RBF neural network in series. BP-RBF mixed novel neural network part of the output of BP neural network RBF neural network as input, using the distribution of training methods, the first part of the BP neural network for training, and then to the RBF network training. As the BP-RBF mixed neural network and BP network portion of the BP network structure built in front of the same, had trained mature, do not need to carry out training, but a new neural network RBF network portion of the front of the RBF neural network with different pairs of RBF network portion of the training, re-established the distribution of constant RBF spread, to be part of the training RBF networks to achieve precision, the entire BP-RBF mixed neural network training is completed, to identify unknown samples. The sample recognition accuracy rate is 96%. Three kinds of neural networks for wood defect recognition rate comparison, BP-RBF mixed neural network to recognize the highest accuracy, but the actual output closer to the target output value.
Experimental results show that application of BP and RBF neural network can successful nondestructive test for three kinds of typical defects of wood. The recognition rate of wood defect is higher above 90%. The formed BP-RBF mixed neural network model for wood defect is higher in accuracy and more effective. This method provides an important theoretical basis for automatic detection of defects in wood.
研制了木材X射线的无损检测系统和木材缺陷图像的采集系统的硬件平台。采用X射线作为检测手段,对木材进行无损检测。通过检测透过被检物体后的射线差异,来判断被检测木材是否有缺陷存在。不仅可以检测木材的表面缺陷,还可以检测木材的内部缺陷。在木材的另一端利用图像增强器进行接收,再经过微光摄像机送入A/D转换器,将木材X射线模拟图像转换成一数字图像存入计算机中,完成木材的缺陷图像的采集工作。
应用数字图像处理技术对采集到的木材缺陷图像进行预处理。对三种木材缺陷图像分别进行了灰度变换增强、改进的加权均值滤波处理、中值滤波处理。处理后的图像对比度明显增强,没有出现传统均值滤波加宽的现象,最大程度地保留了图像的缺陷细节,易于后续的图像特征提取。运用常用的几个边缘检测算子：Roberts算子、Sobel算子、Prewitt算子、Log算子、Canny算子对木材缺陷图像进行边缘检测,检测结果表明Sobel算子检测效果较好,速度较快。为了提高边缘提取的效果,对木材缺陷的原始图像、灰度变换增强后的图像、加权均值滤波后的图像、中值滤波后的图像分别进行二值化处理,对加权均值滤波后的图像的二值化处理效果最佳,对该二值图像缺陷区域进行填充处理。应用Sobel算子对填充后的二值图像进行边缘提取,提取出清晰的木材缺陷边缘。
提出了一种通过对图像实现木材缺陷图像分割的新方法。首先对效果最佳的木材缺陷的二值图像进行图像翻转,将其数组转化为uint8类型,然后将翻转后得到的图像与中值滤波后的木材缺陷图像进行,木材缺陷图像的结果使缺陷区域完全从背景中分离,完成了图像分割,同时并没有改变缺陷区域的灰度、结构等特征。
木材缺陷在特征的选择上具有四个特点：可区别性,对不同类别的对象,特征要具有明显的差异性；可靠性,对同类别的对象,特征具有相近性；独立性,所用的特征之间互不干涉,互不相关；数量少,数量多虽然能更好的区别不同的对象,但是计算量增大,运算时间增长,同样会使识别准确率下降。根据节子、腐朽、虫害的缺陷特征差异,主要提取木材缺陷图像的灰度特征和形状特征。首先对边缘提取后的木材缺陷图像进行扫描,记录缺陷区域的边缘的坐标,确定了木材缺陷的位置和尺寸,再根据确定的位置和尺寸分别对三种木材缺陷的二值图像和分割后的图像进行特征提取,提取了木材缺陷长宽比、圆形度、灰度均值、灰度方差4个特征量。结合节子、腐朽、虫害各自的结构特点,引入了矩及几何中心矩,通过对几何矩的非线性组合,得到一组对于图像平移、旋转和比例不变的矩—Hu不变矩。根据Hu不变矩的物理意义,对原有的Hu不变矩进行抽象,由原来的7个特征抽象为10个特征,满足了图像的平移、缩放和旋转不变性。提取出了10个木材缺陷的结构特征值。这样对三种典型缺陷各提取了14个木材缺陷的特征值,对这14个特征进行归一化处理,归一化后的特征值将作为神经网络输入的特征向量。
设计了BP神经网络模型,确定BP网络的层数及各层神经元数目,隐层和输出层选择S型函数作为激励(传递)函数,通过对BP各种算法的比较,得出Levenberg-Marquardt学习算法是最优的算法,用已知样本对BP神经网络进行学习,将训练成熟的网络对未知样本进行仿真,BP神经网络对木材缺陷的识别准确率达到90%。
构建了RBF神经网络模型,RBF神经网络对数据具有较好的逼近能力。RBF神经网络为三层网络,确定了各层的节点数,隐层传递函数为高斯函数,输出层传递函数为线性函数,经过反复试验确定了径向基分布常数spread,用已知样本对RBF神经网络进行训练,训练完成后,对未知样本进行仿真试验,对木材缺陷的识别准确率达到92%。
组建了BP-RBF混合神经网络模型。BP神经网络具有较好的数据压缩能力,RBF神经网络对数据具有较好的逼近效果,将BP神经网络和RBF神经网络串联。BP-RBF混合神经网络中的BP神经网络部分的输出作为RBF神经网络的输入,采用分布训练的方法,先对BP神经网络部分进行训练,然后再对RBF网络进行训练。由于BP-RBF混合神经网络中的BP网络部分与前面构建的BP网络结构相同,已训练成熟,不需要再进行训练,但混合神经网络中的RBF网络部分与前面的RBF神经网络有所不同,对RBF网络部分进行训练,重新确定了径向基分布常数spread,待RBF网络部分训练达到精度要求后,整个BP-RBF混合神经网络训练完成,对未知样本进行识别。对样本的识别准确率达到96%。三种神经网络的木材缺陷识别率进行比较,BP-RBF混合神经网络识别准确率最高,而且实际输出值更接近于目标输出值。
实验结果表明,应用BP和RBF神经网络方法可以成功地对木材的三种典型缺陷进行无损检测,木材缺陷的识别率都在90%以上,组建的BP-RBF混合神经网络模型对木材缺陷识别的精度高,更加有效。此方法为实现木材缺陷的自动化检测提供了重要的理论依据。
Wood nondestructive testing is a synthetic, nondestructive detecting method. It can detect the wood defects accurately without destroying the surface and structures of the wood. So it can make sure to select wood properly, use wood effectively, which will enhance the occupating coefficient and economic value of the wood and save the wood resource effectively. This thesis studies the two typical kinds of north woods:larch and birch, which belong to softwood and broad-leaved wood. The three typical wood defects:knot, rot and grub-hole have been studied in detail.
The X-ray wood nondestructive testing system, hardware platform image acquisition for wood defect testing is invented. The wood nondestructive testing has been done using X-ray method. The defects can be found by detecting the X-ray difference between the rays fore and after transmitted the wood. This method can not only detect the surface defects of wood, but also the internal defects. The signals are received by the image enhancement equipment on the other end of the wood, and then are transmitted into A/D converter, which convert the X-ray analog pattern of wood into digital images and are stored into computer. This is the all process of image acquisition for wood defect.
Digital image processing is applied for preprocessing of wood defects images. Gray transform enhancement, improved weighted mean filtering and median filtering are applied for the wood defects images of these three kinds of trees discussed above. The results show that the contrasts of these images enhance apparently, and there is no the widened mean filtering traditionally, the image defect details are reserved maximatily, and it is easy for the image feature extraction. The several commonly used edge detection operators:Roberts operator, Sobel operator, Prewitt operator, Log operator and Canny operator, are used for edge detection of the wood defect images. The detect result shows that Sobel operator is better and faster. In order to improve the effect of edge detection, binarization processing has been applied for the original images, the gray transform enhancement images, the weighted mean filtering images and the median filtering images of wood defects respectively, which found that the weighted mean filtering images is the best. Then the defect areas of binary images of wood defects are filled up. For extracting the clear edge of wood defects, the filled binary images are edge extracted with Sobel operator.
A new method for segmenting the wood defects images by image adduct has been presented. First, invert the best binary image of wood defect, and convert its array into uint8 type, and then adduct the image obtained after median filtering with the defect image, which is turned over. The result of wood defect image plus is that defects regions take apart completely with their background, which means the image segmentation is completed. At the same time the characteristics of the defect regions, such as shade of gray, structure, do not changed.
The choice of wood defects has four characteristics:distinctiveness, which means there should be an obvious difference for different objects; reliability, which means features should be similar for the same types of objects; independence, which means the characteristics used should be non-interacting, unrelated; small quantity, although big quantity can make it easier to distinguish different objects, but the amount of computation and the computation time are increased, which will decrease the identify accuracy. According to feature differences of knot, rot, grub-hole, the gray and shape characteristics of wood defects are extract mainly. First, scan the wood defects images after the edge extraction, record the coordinates of the edge of defect regions, determine the location and size of wood defects, then feature extraction is applied for binary images and segmented images of three kinds of wood defects. Four characteristic quantities have been extracted for wood defects:aspect ratio, circular degrees, the mean gray, gray variance. Combined the structural characteristics of knot, rot and grub-hole, geometric moment is introduced. By non-linear combination of geometric moment, a set of moment:Hu invariant moment, which is invariant for image translation, rotation and proportion is derived. According to the physical meaning of Hu invariant moment, the original Hu invariant moments are abstracted, its seven features change into ten features, but still invariant for image translation, convergent-divergent and rotation. Ten structural characteristics of wood defects are extracted. So the fourteen characteristic values of wood defects are extracted for these three kinds of typical defects, and are normalized. The normalized eigenvalues will serve as the input feature vectors of neural network.
The BP neural network model has been designed, determined the number of layers of BP network and neuron numbers for each layer. The hidden layer and output layer select S-function as an incentive (transfer) function. Through the comparison of various algorithms of BP network, it is found that Levenberg-Marquardt learning algorithm is the optimal algorithm. Learn the BP neural network with known samples, simulate the unknown samples with this mature network, which find that the accuracy rate of this BP neural network for wood defects recognition is as high as 90%.
RBF neural network model has been constructed, RBF neural network has good approximation for the data. RBF neural network is three-layer network, identified the nodes of each layer. Hidden layer transfer function belongs to Gaussian function, the output layer transfer function is a linear function, after repeated testing, radial based distribute function "spread" is determined. The RBF neural network is trained with known samples, and then simulation is applied on unknown samples. It is found that the accuracy rate of this network for wood defects recognition is as high as 92%.
Set up a new BP-RBF mixed neural network model. BP neural network has a better data compression capability, RBF neural network has a good approximation to the data results, will be BP neural network and RBF neural network in series. BP-RBF mixed novel neural network part of the output of BP neural network RBF neural network as input, using the distribution of training methods, the first part of the BP neural network for training, and then to the RBF network training. As the BP-RBF mixed neural network and BP network portion of the BP network structure built in front of the same, had trained mature, do not need to carry out training, but a new neural network RBF network portion of the front of the RBF neural network with different pairs of RBF network portion of the training, re-established the distribution of constant RBF spread, to be part of the training RBF networks to achieve precision, the entire BP-RBF mixed neural network training is completed, to identify unknown samples. The sample recognition accuracy rate is 96%. Three kinds of neural networks for wood defect recognition rate comparison, BP-RBF mixed neural network to recognize the highest accuracy, but the actual output closer to the target output value.
Experimental results show that application of BP and RBF neural network can successful nondestructive test for three kinds of typical defects of wood. The recognition rate of wood defect is higher above 90%. The formed BP-RBF mixed neural network model for wood defect is higher in accuracy and more effective. This method provides an important theoretical basis for automatic detection of defects in wood.
引文
[1]国家林业局森林资源管理司.第六次全国森林资源清查及森林资源状况[J].绿色中国,2006,1：10-12
[2]王立海,杨学春,徐凯宏.木材无损检测技术的研究现状与进展[J].森林工程,2001,17(6)：1-3
[3]戚大伟.基于分数布朗随机场与分形参数的木材漏节图像处理[J].林业科学,2004,40(4)：145-147
[4]李坚.木材科学新篇[M].哈尔滨：东北林业大学出版社,1993
[5]S.Ya.Yarema. Nondestructive Testing and Technical Diagnostics[J]. Materials Science, 2003,37(2):361-362
[6]林金国,徐永吉,陈晞,等.木质材料无损检测方法及应用[J].木工机床,2002(2)：30-37
[7]Roy, F.Pellerin, Robert, J.Ross. Nondestructive Evaluation of Wood[M]. WI:Forest Products Society,2002
[8]刘镇波,刘一星,于海鹏,等.木材弹性模量自动检测的研究进展[J].林业科技,2004,29(1)：45-48
[9]胡英成,顾继友,王逢瑚.木材及人造板物理力学性能无损检测技术研究的发展与展望[J].世界林业研究,2002,15(4)：39-46
[10]孟令联,赵钟声,刘一星.木材无损检测技术及其应用与进展[J].林业机械与木工设备,2001,29(9)：4-6
[11]Robert, J.Ross, Brian, K.Brashaw, et al. F.Pellerin.Nondestructive Evaluation of Wood[J]. Forest Products Journal,1998,48(1):14-19
[12]王立海,杨学春,徐凯宏.木材缺陷无损检测技术研究现状[J].林业科技,2002,27(5)：35-38
[13]段新芳,李玉栋,王平.无损检测技术在木材保护中的应用[J].木材工业,2002,16(5)：14-16
[14]Satoru Tsuchikawa, Toshimasa Takahashi, Tsutsumi. Nondestructive Measurement of Wood Properties by Using Near-infrared Laser Radiation[J]. Forest Products Journal,2000, 1:81-86
[15]朱益民.木材无损检测的现状与展望[J].云南林业科技,1996,12(1)：82-84
[16]成俊卿.木材学[M].北京：中国林业出版社,1985
[17]张旭辉,马宏伟.超声无损检测技术的现状和发展趋势[J].机械制造,2002,5(2)：24-26
[18]吴建生,周优军,金龙.神经网络及其研究进展[J].广西师范学院学报,2005, 22(1)：92-97
[19]王学武,谭得健.神经网络的应用与发展趋势[J].计算机工程与应用,2003,39(3)：98-113
[20]Datta S, Banerjee M K. Mapping the input-output relationship in HSLA steels through expert neural network[J]. Materials Science and Engineering:A,2006,420(2):254-264
[21]Y.C. Liang, D.P. Feng, H.P. Lee, S.P. Lim, K.H. Lee. Successive approximation training algorithm for feedforward neural networks[J]. Neure computing,2002,42:311-322
[22]胡方明,简琴,张秀君.基于BP神经网络的车型分类器[J].西安电子科技大学学报,2005,32(3)：439-442
[23]王科俊,姚向辉.人脸图像检测与识别方法综述[J].自动化技术与应用,2004,23(12)：5-8
[24]关山,王伟杰,赵学增.基于BP神经网络的车牌字符自动识别方法的研究[J].东北电力学院学报,2004,24(4)：56-60
[25]Sergios Theodoridis, Konstantinos Koutroumbas. Pattern Recognition and Neural Networks[J]. Lecture Notes in Computer Science,2001,20(9):67-69
[26]飞思科技产品研发中心.神经网络理论与MATLAB7实现[M].北京：电子工业出版社,2005：1-2
[27]王家文,曹宇.MATLAB6.5图形图像处理[M].北京：国防工业出版社,2004：166-167
[28]飞思科技产品研发中心.MATLAB7基础与提高[M].北京：电子工业出版社,2005：1-3
[29]闻新,周录,王丹力.MATLAB神经网络应用设计[M].北京：科学出版社,2000：88-95
[30]闻新,周露,李翔.MATLAB神经网络仿真与应用[M].北京：科学出版社,2003：53-62
[3l]许东,吴铮.基于MATLAB6.X的系统分析与设计神经网络[M].西安：西安电子科技大学出版社,2002：165-172
[32]曾祥照.X射线实时成像检测技术现场测评方法[J].无损探伤,2000,13(5)：27-32
[33]戚大伟.木材无损检测图像处理系统的研究[J].林业科学,2001,37(6)：92-96
[34]Daniil Ryabko. On Computability of Pattern Recognition Problems[J]. Lecture Notes in Computer Science,2005,37(34):148
[35]Hardcastle M J, Worrall D M. ROSAT X-ray observations of 3CRR radio sources[J]. Mon Not Roy Astron Soc,1999,309(7):969-990
[36]Brinkmann W, Siebert J, Boller T. The X-ray AGN content of the Molonglo 408 MHz Survey:Bulk properties of previously optically identified sources[J]. Astron Astrophys, 1994,281(9):355-374
[37]Rogerio Ferreira de Paiva, John Lynch, Elisabeth Rosenberg, et al. A beam hardening correction for X-ray microtomography[J]. NDT & International,1998,31(1):17-22
[38]Wu Q W, Yuan F, Cao X W. Origin of the X-ray emission in the nuclei of FRIs[J]. J Phys-Conference Ser,2008,131(2):20-60
[39]C. David, B. Kaulich, R. Barrett, M. Salome, and J. Susini. High-resolution lenses for sub-100nm x-ray fluorescence microscopy [J]. APPLIED PHYSICS LETTERS,2000,77(23): 3851-3853
[40]B. Lai, W. B. Yun, D. Legnini, et al. Hard x-ray phase zone plate fabricated by lithographic techniques[J]. Appl.Phys.Lett.,1992,61(16):1877-1879
[41]Hardcastle M J, Evans D A, Croston J H. The X-ray nuclei of intermediate-redshift radio sources[J]. Mon Not Roy Astron Soc,2006,370(11):1893-1904
[42]Hanke R, Bobel F. Determination of material flaw size by intensity evaluation of polychromatic X-ray t ransmission[J]. NDT & E International,1992,25(2):87
[43]Leiliveld C J, Maas J G, Van Eijk C W E, et al. On the significance of scattered radiation in ndust rial X-ray computerized tomographic imaging. IEEE Transaction on Nuclear Science,1994,41(1):290-294
[44]阮炜,郑明山.X射线检测钢丝绳成像系统设计[J].自动化技术与应用,2005,24(12)：67-83
[45]Siebert J, Brinkmann W, Morganti R, et al. The soft X-ray properties of a complete ample of radio sources[J]. Mon Not Roy Astron Soc,1996,279:1331-1344
[46]Hardcastle M J, Harris D E, Worrall D M, et al. The origins of X-ray emission from the hot spots of FR Ⅱ radio sources[J]. Astrophys J,2004,612:729-748
[47]Fabbiano G, Miller L, Trinchieri G, et al. An X-ray survey of a complete sample of 3CR radio galaxies[J]. Astrophys J,1984,277:115-131
[48]王华.X射线数字成像系统及其应用[J].计测技术,2005,25(6)：28-32
[49]Canosa C M, Worrall D M, Hardcastle M J, et al. X-ray observations of low-power radio galaxies from the B2 catalogue[J]. Mon Not Roy Astron Soc,1999,310:30-38
[50]朱红.数字图像处理[M].北京：科学出版社,2006：10-19
[51]Rafael C.Gonzalez, Richard E.Woods. Digital image processing [M]. BEIJING:Publishing House of Electronics Industry,2003:496
[52]龚声蓉,刘纯平,王强.数字图像处理与分析[M].北京：清华大学出版社,2006：25-30
[53]吴传孙,邹扬德,周定康.二值化算法在车牌识别中的应用[J],计算机与现代化,2003,18(6)：13-16
[54]姚敏.数字图像处理[M].北京：机械工业出版社,2006：1-3
[55]王家文,李仰军.MATLAB7.0图形图像处理[M].国防工业出版社,2006：1-2
[56]E.T. Lee, P. K. Kim, H. J. Kim. Automatic Recognition of a Car License Plate Using[J]. Color Image Processing.2001,6(1):50-58
[57]何东健,耿楠,张义宽.数字图像处理[M].西安：西安电子科技大学出版社,2003：3-5
[58]Luminita A.Vese, Tony F.Chan. A Multiphase Level Set Framework for Image Segmentation Using the Mumford and Shah Model[J]. International Journal of Computer Vision,2002,50(3):271-293
[59]J.W.Funck, Y.Zhong, D.A.Butler, C.C.Brunner, J.B.Forrer. Image segmentation algorithms applied to wood defect detection[J]. Computers and Electronics in Agriculture,2003,41: 157-179
[60]Pew-Thian Yap, Raveendran Paramesran, Seng-Huat Ong. Image Analysis by Krawtchouk Moments. IEEE Trans. on Image Processing,2003,12(11):1367-1377
[61]杨高波,杜青松.MATLAB图像/视频处理应用及实例[M].北京：电子工业出版社,2010：92-105
[62]Berend Jan van der Zwaag, Kees Slump, Lambert Spaanenburg. On the Analysis of Neural Networks for Image Processing[J]. Lecture Notes in Computer Science,2003,27(5): 950-957
[63]Hong Yan. Human Face Image Processing Techniques[J]. International Journal of Image and Graphics,2001,1(1):197-215
[64]Rafael C. Gonzalez, Richard E. Woods, Steven L. Eddins著.阮秋琦等译.数字图像处理(MATLAB版)[M].北京：电子工业出版社,2005：55-61
[65]王爱玲,叶明生,邓秋香.MATLABR2007图像处理技术与应用[M].北京：电子工业出版社,2008：166-170
[66]J.Canny. A computational approach to edge detection. IEEE Trans.Patern Analysis and Machine Intelligence,1996,8(6)679-698
[67]J.W.Funck, Y.Zhong, D.A.Butler, C.C.Brunner, J.B.Forrer. Image segmentation algorithms applied to wood defect detection[J]. Computers and Electronics in Agriculture,2003, 35(41):157-179
[68]T.Tanaka, T.Joke, T.Oka. Cell nucleus segmentation of skin tumor using image processing[C]. Engineering in Medicine and Biology Society 2001 Proceedings of the 23rd Annual International Conference of the IEEE,2001:2716-2719
[69]Rishi R R, Chaudhuri P, Thresholding in edge detection:a statistical approach. IEEE Tran on Image Processing,2004,13(7):927-936
[70]K. H. Liang, T. Tjahjadi, Y H. Yang. Roof edge detection using regularized cubic B-spline fitting[J]. Pattern Recognition,1997,30 (5):719-728
[71]D.Demigny. On Optimal Linear Filtering for Edge Detection. IEEE Trans. Image Processing,2002,11(7):728-737
[72]A.Nedzved, S.Ablameyko, I.Pitas. Morphological segmentation of histology cell image[C]. IEEE pattern recognition 2000 proceedings 15th international conference on,2000,5(1): 500-503
[73]陈强,朱立新.结合Canny算了的图像二值化[J].计算机辅助设计与图形学学报,2005,17(6)：1302-1306
[74]高朝晖,黄卫.基于彩色图像车牌分割研究.公路交通科技[J],2004,21(8)：114-117
[75]Wells W M, Grimson W E L., Kikinis R., Jolesz F A. Adaptive Segmentation of MRI Data. IEEE Trans on Medical Imaging,1996,15 (4):429-442
[76]Wermser D, Haussmann G, Liedtke C E. Segmentation of blood smears by hierarchical thresholding[J].Computer Vision Graphics and Image Processing,1984,25(2):151-168
[77]付峰,应义斌.生物图像阈值分割方法研究[J].浙江大学学报,2003：29(1),108-112
[78]Corneloup G, Moysan J. BSCAN. Image Segmentation by Thresholding Using Cooccurence Matrix Analysis[J]. Pattern Recognition.1996,29(2):281-296
[79]Liou S P, Jain R C. An Approach to Three Dimensional Image segmentation[J]. Computer Vision Graphics and Image Processing,1991,46 (3):237-252
[80]李坚.木材科学[M].哈尔滨：东北林业大学出版社,1994：283-285
[81]赵建民.木材概论[M].北京：高等教育出版社,2002：90-100
[82]刘继敏,史忠植.一种基于形状的图像信息检索方法[J].软件学报,2001,11(1)：110-115
[83]刘进,张天序.图像不变矩的轮廓链快速算法[J].华中科技大学学报(自然科学版),2003,31(1)：67-69
[84]童卓,李霆.基于遗传算法的图像不变矩匹配[J].计算机工程与科学,2002,24(3)：14-17
[85]杨静,丘江,王岩飞.线性不变矩及其在图像识别中的应用算法研究[J].光子学报,2003,32(3)：336-339
[86]王波涛,孙景鳌,蔡安妮.相对矩及在几何形状识别中的应用[J].中国图象图形学报,2001,6A(3)：296-300
[87]YS.Kim, W.Y.Kim. Content-based trademark retrieval system using a visually salient feature[J]. Image and Vision Computing.1998,16:931-939
[88]Richard J. Prokop and Anthony P.Reeves. A Survey of Moment-Based techniques for Unoccluded Object Representation and Recognition [J]. CVGIP,1992,54(5):438-460
[89]杜亚娟,张洪才,潘泉.基于矩特征的三维飞机目标识别[J].数据采集与处理, 2000,15(3)：390-394
[90]初秀民,王荣本,储江伟,等.基于不变矩特征的沥青路面破损图像识别[J].吉林大学学报,2003,33(1)：1-7
[91]Flusser J, Suk T. Pattern recognition by affine moment invariants[J].Pattern Recognition,1993,26(1):167-174
[92]HU M K.Visual Pattern Recognition by Moment In variants. IEEE Trans on Information Theory,1962,22(8):179-187
[93]HU M K.Improved Moment Invariants for Shape Discrim Ination[J]. Pattern Recognition, 1993,26(5):683-686
[94]孙君顶,赵珊.图像低层特征提取与检索技术[M].北京：电子工业出版社,2009：95-99
[95]Chen C C. Improved momentinvariant for shape discrimination[J]. Pattern Recognition, 1993,26(5):683-686
[96]蒋宗礼.人工神经网络导论[M].北京：高等教育出版社,2001：16-18
[97]魏海坤.神经网络结构设计的理论与方法[M].北京：国防工业出版社,2005：11-15
[98]Malki, S., Spaanenburg, L., Ray, N. Neural vision sensors for surface defect detection. IEEE International Joint Conferenc,2004,4(25):3155-3160
[99]Hou L, Wu W, Zhu B, Li F. A segmentation method for merged characters using self-organizing map neural networks. Journal of Information and Computational Science,2006, 3(2):219-226
[100]Shishir Bashyal. Classification of Psychiatric Disorders Using Artificial Neural Network[J]. Lecture Notes in Computer Science,2005,34(9):386-394
[101]Alireza Givehchi, Gisbert Schneider. Impact of Descriptor Vector Scaling on the Classification of Drugs and Nondrugs with Artificial Neural Networks[J]. Journal of Molecular Modeling,2004,10(3):204-211
[102]Jun Kong, Wei wu. Online gradient methods with a punishing term for neural networks[J]. Northeastern Mathematical Journal,2001,17(3):371-378
[103]Hirohito Shintani, Hirofumi Nagashino, Masatake Akutagawa, et al. A Neural Network Model for Pattern Recognition[J]. Lecture Notes in Computer Science,2003,27(7): 814-821
[104]Yasuo Saito, Toshiharu Hatanaka, Katsuji Uosaki, et al. Neural Network Application to Eggplant Classification[J]. Lecture Notes in Computer Science,2003,27(7):933-940
[105]N. Ampazis, S.T. Perantonis. Two highly efficient second order algorithms for training feedforward networks. IEEE Trans. Neural Networks,2002,13:1064-1284
[106]吴成东,刘文涵,傅小菲,等.基于粗网络神经网络的车牌字符识别方法[J].沈 阳建筑大学学报(自然科学版),2007,23(4)：693-697
[107]Kwang Kyu Seo, Sung Hwan Min, Hun Woo Yoo. Artificial Neural Network Based Life Cycle Assessment Model for Product Concepts Using Product Classification Method[J]. Lecture Notes in Computer Science,2005,34(8):458
[108]董剑龙,徐美芳.神经网络在缺陷识别中的应用[J].山西电子技术,2005,15(6)：24-26
[109]Haralambos Sarimveis, Philip Doganis, Alex Alexandridis. A classification technique based on radial basis function neural networks[J].Advances in Engineering Software,2006, 37(4):218-221
[110]C.Harpham,C.W.Dawson. The effect of different basis functions on a radial basis function network for time series prediction:A comparative study [J]. Neurocomputing, 2006,69(18):2161-2170
[111]Meng Joo Er,Weilong Chen,Shiqian Wu. High-Speed Face Recognition Based on Discrete Cosine Transform and RBF Neural Networks. IEEE Transaction on Neural networks,2005,16(3):679-691
[112]植浚文,戴青云,黄雪贤.一种基于径向基神经网络的车牌字符识别方法[J].计算机应用,2005,25(12)：311-313
[113]王春,刘波,周新志.采用BP神经网络的车牌字符识别方法研究[J].中国测试技术,2005,31(1)：26-28
[114]杨宏伟,钟新辉,胡建强.基于粗糙集和神经网络的润滑油中磨损磨粒的识别[J].润滑与密封,2007,32(1)：162-164
[115]林培群,徐建闽,傅惠.多分支BP网络模型及其在车型分类中的应用[J].人工智能,2005,21(12)：182-184
[116]王科俊,李国斌.几种变学习率的快速BP算法比较研究[J].哈尔滨工程大学学报,1997,18：31-35
[117]Guang-Bin Huang, P.Saratchandran, Narasimhan Sundararajan. An Efficient Sequential Learning Algorithm for Growing and Pruning RBF(GAP-RBF) Networks. IEEE Trans.on Cybernetics.2004,34(6):2284-2292
[118]植浚文,戴青云,黄雪贤.一种基于径向基神经网络的车牌字符识别方法[J].计算机应用,2005,25(12)：311-313
[119]Ali Ghodsi, Dale Schuurmans. Automatic basis selection techniques for RBF Networks[J]. Neural Networks,2003, (16):809-816
[120]Douglas Turnbull,Charles Elkan. Fast Recognition of Musical Genres Using RBF Networks. IEEE Transactions on Knowledge and Data Engineering,2005,17(4):580-584
[121]Alex Alex Haralambos Sarim, Georage Bafas. A new algorithm for online structure and parameter adaptation of RBF networks[J]. Neural Networks,2003,21(16):1003-1017
[122]王晓东,秦超英.基于径向基神经网络的目标识别研究[J].西南民族大学学报,2006,12(3)：195-197
[123]田凯,郑丽颖,王科俊.一种用于模式识别的新型神经网络模型[J].哈尔滨工程大学学报,2002,23(6)：82-84
[2]王立海,杨学春,徐凯宏.木材无损检测技术的研究现状与进展[J].森林工程,2001,17(6)：1-3
[3]戚大伟.基于分数布朗随机场与分形参数的木材漏节图像处理[J].林业科学,2004,40(4)：145-147
[4]李坚.木材科学新篇[M].哈尔滨：东北林业大学出版社,1993
[5]S.Ya.Yarema. Nondestructive Testing and Technical Diagnostics[J]. Materials Science, 2003,37(2):361-362
[6]林金国,徐永吉,陈晞,等.木质材料无损检测方法及应用[J].木工机床,2002(2)：30-37
[7]Roy, F.Pellerin, Robert, J.Ross. Nondestructive Evaluation of Wood[M]. WI:Forest Products Society,2002
[8]刘镇波,刘一星,于海鹏,等.木材弹性模量自动检测的研究进展[J].林业科技,2004,29(1)：45-48
[9]胡英成,顾继友,王逢瑚.木材及人造板物理力学性能无损检测技术研究的发展与展望[J].世界林业研究,2002,15(4)：39-46
[10]孟令联,赵钟声,刘一星.木材无损检测技术及其应用与进展[J].林业机械与木工设备,2001,29(9)：4-6
[11]Robert, J.Ross, Brian, K.Brashaw, et al. F.Pellerin.Nondestructive Evaluation of Wood[J]. Forest Products Journal,1998,48(1):14-19
[12]王立海,杨学春,徐凯宏.木材缺陷无损检测技术研究现状[J].林业科技,2002,27(5)：35-38
[13]段新芳,李玉栋,王平.无损检测技术在木材保护中的应用[J].木材工业,2002,16(5)：14-16
[14]Satoru Tsuchikawa, Toshimasa Takahashi, Tsutsumi. Nondestructive Measurement of Wood Properties by Using Near-infrared Laser Radiation[J]. Forest Products Journal,2000, 1:81-86
[15]朱益民.木材无损检测的现状与展望[J].云南林业科技,1996,12(1)：82-84
[16]成俊卿.木材学[M].北京：中国林业出版社,1985
[17]张旭辉,马宏伟.超声无损检测技术的现状和发展趋势[J].机械制造,2002,5(2)：24-26
[18]吴建生,周优军,金龙.神经网络及其研究进展[J].广西师范学院学报,2005, 22(1)：92-97
[19]王学武,谭得健.神经网络的应用与发展趋势[J].计算机工程与应用,2003,39(3)：98-113
[20]Datta S, Banerjee M K. Mapping the input-output relationship in HSLA steels through expert neural network[J]. Materials Science and Engineering:A,2006,420(2):254-264
[21]Y.C. Liang, D.P. Feng, H.P. Lee, S.P. Lim, K.H. Lee. Successive approximation training algorithm for feedforward neural networks[J]. Neure computing,2002,42:311-322
[22]胡方明,简琴,张秀君.基于BP神经网络的车型分类器[J].西安电子科技大学学报,2005,32(3)：439-442
[23]王科俊,姚向辉.人脸图像检测与识别方法综述[J].自动化技术与应用,2004,23(12)：5-8
[24]关山,王伟杰,赵学增.基于BP神经网络的车牌字符自动识别方法的研究[J].东北电力学院学报,2004,24(4)：56-60
[25]Sergios Theodoridis, Konstantinos Koutroumbas. Pattern Recognition and Neural Networks[J]. Lecture Notes in Computer Science,2001,20(9):67-69
[26]飞思科技产品研发中心.神经网络理论与MATLAB7实现[M].北京：电子工业出版社,2005：1-2
[27]王家文,曹宇.MATLAB6.5图形图像处理[M].北京：国防工业出版社,2004：166-167
[28]飞思科技产品研发中心.MATLAB7基础与提高[M].北京：电子工业出版社,2005：1-3
[29]闻新,周录,王丹力.MATLAB神经网络应用设计[M].北京：科学出版社,2000：88-95
[30]闻新,周露,李翔.MATLAB神经网络仿真与应用[M].北京：科学出版社,2003：53-62
[3l]许东,吴铮.基于MATLAB6.X的系统分析与设计神经网络[M].西安：西安电子科技大学出版社,2002：165-172
[32]曾祥照.X射线实时成像检测技术现场测评方法[J].无损探伤,2000,13(5)：27-32
[33]戚大伟.木材无损检测图像处理系统的研究[J].林业科学,2001,37(6)：92-96
[34]Daniil Ryabko. On Computability of Pattern Recognition Problems[J]. Lecture Notes in Computer Science,2005,37(34):148
[35]Hardcastle M J, Worrall D M. ROSAT X-ray observations of 3CRR radio sources[J]. Mon Not Roy Astron Soc,1999,309(7):969-990
[36]Brinkmann W, Siebert J, Boller T. The X-ray AGN content of the Molonglo 408 MHz Survey:Bulk properties of previously optically identified sources[J]. Astron Astrophys, 1994,281(9):355-374
[37]Rogerio Ferreira de Paiva, John Lynch, Elisabeth Rosenberg, et al. A beam hardening correction for X-ray microtomography[J]. NDT & International,1998,31(1):17-22
[38]Wu Q W, Yuan F, Cao X W. Origin of the X-ray emission in the nuclei of FRIs[J]. J Phys-Conference Ser,2008,131(2):20-60
[39]C. David, B. Kaulich, R. Barrett, M. Salome, and J. Susini. High-resolution lenses for sub-100nm x-ray fluorescence microscopy [J]. APPLIED PHYSICS LETTERS,2000,77(23): 3851-3853
[40]B. Lai, W. B. Yun, D. Legnini, et al. Hard x-ray phase zone plate fabricated by lithographic techniques[J]. Appl.Phys.Lett.,1992,61(16):1877-1879
[41]Hardcastle M J, Evans D A, Croston J H. The X-ray nuclei of intermediate-redshift radio sources[J]. Mon Not Roy Astron Soc,2006,370(11):1893-1904
[42]Hanke R, Bobel F. Determination of material flaw size by intensity evaluation of polychromatic X-ray t ransmission[J]. NDT & E International,1992,25(2):87
[43]Leiliveld C J, Maas J G, Van Eijk C W E, et al. On the significance of scattered radiation in ndust rial X-ray computerized tomographic imaging. IEEE Transaction on Nuclear Science,1994,41(1):290-294
[44]阮炜,郑明山.X射线检测钢丝绳成像系统设计[J].自动化技术与应用,2005,24(12)：67-83
[45]Siebert J, Brinkmann W, Morganti R, et al. The soft X-ray properties of a complete ample of radio sources[J]. Mon Not Roy Astron Soc,1996,279:1331-1344
[46]Hardcastle M J, Harris D E, Worrall D M, et al. The origins of X-ray emission from the hot spots of FR Ⅱ radio sources[J]. Astrophys J,2004,612:729-748
[47]Fabbiano G, Miller L, Trinchieri G, et al. An X-ray survey of a complete sample of 3CR radio galaxies[J]. Astrophys J,1984,277:115-131
[48]王华.X射线数字成像系统及其应用[J].计测技术,2005,25(6)：28-32
[49]Canosa C M, Worrall D M, Hardcastle M J, et al. X-ray observations of low-power radio galaxies from the B2 catalogue[J]. Mon Not Roy Astron Soc,1999,310:30-38
[50]朱红.数字图像处理[M].北京：科学出版社,2006：10-19
[51]Rafael C.Gonzalez, Richard E.Woods. Digital image processing [M]. BEIJING:Publishing House of Electronics Industry,2003:496
[52]龚声蓉,刘纯平,王强.数字图像处理与分析[M].北京：清华大学出版社,2006：25-30
[53]吴传孙,邹扬德,周定康.二值化算法在车牌识别中的应用[J],计算机与现代化,2003,18(6)：13-16
[54]姚敏.数字图像处理[M].北京：机械工业出版社,2006：1-3
[55]王家文,李仰军.MATLAB7.0图形图像处理[M].国防工业出版社,2006：1-2
[56]E.T. Lee, P. K. Kim, H. J. Kim. Automatic Recognition of a Car License Plate Using[J]. Color Image Processing.2001,6(1):50-58
[57]何东健,耿楠,张义宽.数字图像处理[M].西安：西安电子科技大学出版社,2003：3-5
[58]Luminita A.Vese, Tony F.Chan. A Multiphase Level Set Framework for Image Segmentation Using the Mumford and Shah Model[J]. International Journal of Computer Vision,2002,50(3):271-293
[59]J.W.Funck, Y.Zhong, D.A.Butler, C.C.Brunner, J.B.Forrer. Image segmentation algorithms applied to wood defect detection[J]. Computers and Electronics in Agriculture,2003,41: 157-179
[60]Pew-Thian Yap, Raveendran Paramesran, Seng-Huat Ong. Image Analysis by Krawtchouk Moments. IEEE Trans. on Image Processing,2003,12(11):1367-1377
[61]杨高波,杜青松.MATLAB图像/视频处理应用及实例[M].北京：电子工业出版社,2010：92-105
[62]Berend Jan van der Zwaag, Kees Slump, Lambert Spaanenburg. On the Analysis of Neural Networks for Image Processing[J]. Lecture Notes in Computer Science,2003,27(5): 950-957
[63]Hong Yan. Human Face Image Processing Techniques[J]. International Journal of Image and Graphics,2001,1(1):197-215
[64]Rafael C. Gonzalez, Richard E. Woods, Steven L. Eddins著.阮秋琦等译.数字图像处理(MATLAB版)[M].北京：电子工业出版社,2005：55-61
[65]王爱玲,叶明生,邓秋香.MATLABR2007图像处理技术与应用[M].北京：电子工业出版社,2008：166-170
[66]J.Canny. A computational approach to edge detection. IEEE Trans.Patern Analysis and Machine Intelligence,1996,8(6)679-698
[67]J.W.Funck, Y.Zhong, D.A.Butler, C.C.Brunner, J.B.Forrer. Image segmentation algorithms applied to wood defect detection[J]. Computers and Electronics in Agriculture,2003, 35(41):157-179
[68]T.Tanaka, T.Joke, T.Oka. Cell nucleus segmentation of skin tumor using image processing[C]. Engineering in Medicine and Biology Society 2001 Proceedings of the 23rd Annual International Conference of the IEEE,2001:2716-2719
[69]Rishi R R, Chaudhuri P, Thresholding in edge detection:a statistical approach. IEEE Tran on Image Processing,2004,13(7):927-936
[70]K. H. Liang, T. Tjahjadi, Y H. Yang. Roof edge detection using regularized cubic B-spline fitting[J]. Pattern Recognition,1997,30 (5):719-728
[71]D.Demigny. On Optimal Linear Filtering for Edge Detection. IEEE Trans. Image Processing,2002,11(7):728-737
[72]A.Nedzved, S.Ablameyko, I.Pitas. Morphological segmentation of histology cell image[C]. IEEE pattern recognition 2000 proceedings 15th international conference on,2000,5(1): 500-503
[73]陈强,朱立新.结合Canny算了的图像二值化[J].计算机辅助设计与图形学学报,2005,17(6)：1302-1306
[74]高朝晖,黄卫.基于彩色图像车牌分割研究.公路交通科技[J],2004,21(8)：114-117
[75]Wells W M, Grimson W E L., Kikinis R., Jolesz F A. Adaptive Segmentation of MRI Data. IEEE Trans on Medical Imaging,1996,15 (4):429-442
[76]Wermser D, Haussmann G, Liedtke C E. Segmentation of blood smears by hierarchical thresholding[J].Computer Vision Graphics and Image Processing,1984,25(2):151-168
[77]付峰,应义斌.生物图像阈值分割方法研究[J].浙江大学学报,2003：29(1),108-112
[78]Corneloup G, Moysan J. BSCAN. Image Segmentation by Thresholding Using Cooccurence Matrix Analysis[J]. Pattern Recognition.1996,29(2):281-296
[79]Liou S P, Jain R C. An Approach to Three Dimensional Image segmentation[J]. Computer Vision Graphics and Image Processing,1991,46 (3):237-252
[80]李坚.木材科学[M].哈尔滨：东北林业大学出版社,1994：283-285
[81]赵建民.木材概论[M].北京：高等教育出版社,2002：90-100
[82]刘继敏,史忠植.一种基于形状的图像信息检索方法[J].软件学报,2001,11(1)：110-115
[83]刘进,张天序.图像不变矩的轮廓链快速算法[J].华中科技大学学报(自然科学版),2003,31(1)：67-69
[84]童卓,李霆.基于遗传算法的图像不变矩匹配[J].计算机工程与科学,2002,24(3)：14-17
[85]杨静,丘江,王岩飞.线性不变矩及其在图像识别中的应用算法研究[J].光子学报,2003,32(3)：336-339
[86]王波涛,孙景鳌,蔡安妮.相对矩及在几何形状识别中的应用[J].中国图象图形学报,2001,6A(3)：296-300
[87]YS.Kim, W.Y.Kim. Content-based trademark retrieval system using a visually salient feature[J]. Image and Vision Computing.1998,16:931-939
[88]Richard J. Prokop and Anthony P.Reeves. A Survey of Moment-Based techniques for Unoccluded Object Representation and Recognition [J]. CVGIP,1992,54(5):438-460
[89]杜亚娟,张洪才,潘泉.基于矩特征的三维飞机目标识别[J].数据采集与处理, 2000,15(3)：390-394
[90]初秀民,王荣本,储江伟,等.基于不变矩特征的沥青路面破损图像识别[J].吉林大学学报,2003,33(1)：1-7
[91]Flusser J, Suk T. Pattern recognition by affine moment invariants[J].Pattern Recognition,1993,26(1):167-174
[92]HU M K.Visual Pattern Recognition by Moment In variants. IEEE Trans on Information Theory,1962,22(8):179-187
[93]HU M K.Improved Moment Invariants for Shape Discrim Ination[J]. Pattern Recognition, 1993,26(5):683-686
[94]孙君顶,赵珊.图像低层特征提取与检索技术[M].北京：电子工业出版社,2009：95-99
[95]Chen C C. Improved momentinvariant for shape discrimination[J]. Pattern Recognition, 1993,26(5):683-686
[96]蒋宗礼.人工神经网络导论[M].北京：高等教育出版社,2001：16-18
[97]魏海坤.神经网络结构设计的理论与方法[M].北京：国防工业出版社,2005：11-15
[98]Malki, S., Spaanenburg, L., Ray, N. Neural vision sensors for surface defect detection. IEEE International Joint Conferenc,2004,4(25):3155-3160
[99]Hou L, Wu W, Zhu B, Li F. A segmentation method for merged characters using self-organizing map neural networks. Journal of Information and Computational Science,2006, 3(2):219-226
[100]Shishir Bashyal. Classification of Psychiatric Disorders Using Artificial Neural Network[J]. Lecture Notes in Computer Science,2005,34(9):386-394
[101]Alireza Givehchi, Gisbert Schneider. Impact of Descriptor Vector Scaling on the Classification of Drugs and Nondrugs with Artificial Neural Networks[J]. Journal of Molecular Modeling,2004,10(3):204-211
[102]Jun Kong, Wei wu. Online gradient methods with a punishing term for neural networks[J]. Northeastern Mathematical Journal,2001,17(3):371-378
[103]Hirohito Shintani, Hirofumi Nagashino, Masatake Akutagawa, et al. A Neural Network Model for Pattern Recognition[J]. Lecture Notes in Computer Science,2003,27(7): 814-821
[104]Yasuo Saito, Toshiharu Hatanaka, Katsuji Uosaki, et al. Neural Network Application to Eggplant Classification[J]. Lecture Notes in Computer Science,2003,27(7):933-940
[105]N. Ampazis, S.T. Perantonis. Two highly efficient second order algorithms for training feedforward networks. IEEE Trans. Neural Networks,2002,13:1064-1284
[106]吴成东,刘文涵,傅小菲,等.基于粗网络神经网络的车牌字符识别方法[J].沈 阳建筑大学学报(自然科学版),2007,23(4)：693-697
[107]Kwang Kyu Seo, Sung Hwan Min, Hun Woo Yoo. Artificial Neural Network Based Life Cycle Assessment Model for Product Concepts Using Product Classification Method[J]. Lecture Notes in Computer Science,2005,34(8):458
[108]董剑龙,徐美芳.神经网络在缺陷识别中的应用[J].山西电子技术,2005,15(6)：24-26
[109]Haralambos Sarimveis, Philip Doganis, Alex Alexandridis. A classification technique based on radial basis function neural networks[J].Advances in Engineering Software,2006, 37(4):218-221
[110]C.Harpham,C.W.Dawson. The effect of different basis functions on a radial basis function network for time series prediction:A comparative study [J]. Neurocomputing, 2006,69(18):2161-2170
[111]Meng Joo Er,Weilong Chen,Shiqian Wu. High-Speed Face Recognition Based on Discrete Cosine Transform and RBF Neural Networks. IEEE Transaction on Neural networks,2005,16(3):679-691
[112]植浚文,戴青云,黄雪贤.一种基于径向基神经网络的车牌字符识别方法[J].计算机应用,2005,25(12)：311-313
[113]王春,刘波,周新志.采用BP神经网络的车牌字符识别方法研究[J].中国测试技术,2005,31(1)：26-28
[114]杨宏伟,钟新辉,胡建强.基于粗糙集和神经网络的润滑油中磨损磨粒的识别[J].润滑与密封,2007,32(1)：162-164
[115]林培群,徐建闽,傅惠.多分支BP网络模型及其在车型分类中的应用[J].人工智能,2005,21(12)：182-184
[116]王科俊,李国斌.几种变学习率的快速BP算法比较研究[J].哈尔滨工程大学学报,1997,18：31-35
[117]Guang-Bin Huang, P.Saratchandran, Narasimhan Sundararajan. An Efficient Sequential Learning Algorithm for Growing and Pruning RBF(GAP-RBF) Networks. IEEE Trans.on Cybernetics.2004,34(6):2284-2292
[118]植浚文,戴青云,黄雪贤.一种基于径向基神经网络的车牌字符识别方法[J].计算机应用,2005,25(12)：311-313
[119]Ali Ghodsi, Dale Schuurmans. Automatic basis selection techniques for RBF Networks[J]. Neural Networks,2003, (16):809-816
[120]Douglas Turnbull,Charles Elkan. Fast Recognition of Musical Genres Using RBF Networks. IEEE Transactions on Knowledge and Data Engineering,2005,17(4):580-584
[121]Alex Alex Haralambos Sarim, Georage Bafas. A new algorithm for online structure and parameter adaptation of RBF networks[J]. Neural Networks,2003,21(16):1003-1017
[122]王晓东,秦超英.基于径向基神经网络的目标识别研究[J].西南民族大学学报,2006,12(3)：195-197
[123]田凯,郑丽颖,王科俊.一种用于模式识别的新型神经网络模型[J].哈尔滨工程大学学报,2002,23(6)：82-84