胶片薄膜光学小波滤波器的研究及其实现
|
摘要
光学小波变换是近年来发展起来的实时处理信号的一种方法,它结合了小波变换和光信息处理的优点,以并行性和高速实时性为特点,为图像数据压缩提供了一个有效途径。这在需要大量图像数据传输的系统,如太空遥感、森林防火等领域有很大的应用前景。光学小波变换在模式识别和纹理分割方面已取得一定的成果,而对于用光学小波变换实现图像数据压缩的报道很少,其主要原因是它对整个光学变换系统的精度要求比其它应用高得多。
在典型的光学4f系统的谱面上加滤波片是实现光学小波变换的常用方法。可以用光栅、空间光调制器、计算全息、胶片作为滤波片。论文研究了在4f系统的谱面上加胶片作为滤波片来实现小波变换的方法。主要工作如下:
(1)研究了胶片的制作:分析了激光打印机、热升华打印机、胶片记录仪的优缺点,结合课题要求,确定胶片记录仪作为胶片制作仪器。
(2)分析了胶片的误差:设计了一系列的样本图像作为色标,通过扫描仪和Matlab对胶片制作的样本图像进行定量定性的分析,得到不同颜色通道的误差特性曲线。
(3)通过多项式拟合,寻求胶片图像均值与标准图像均值间的函数f(x),使f(x)与标准的数据点最接近。
(4)胶片一旦制作,难以修改,针对胶片的加工工艺误差,利用神经网络自学习特性对胶片制作误差进行分析,用胶片制作误差修正神经网络权值。若用网络输出结果对标准图像进行预畸变处理,则制作后胶片图像的灰度值趋进标准图像的灰度值。
研究结果表明:以24bit的彩色图像为分析基础,胶片图像的灰度级与原始图像的灰度级之间并非线性对应,R、G、B三通道的精度略有差异,R通道的性能最优,B通道的性能最差,当各通道灰度级以相同规律变化时,引入的误差最小。经多项式拟合拟合后,均值的平均峰值信噪比可提高10~20dB。而文中确定的网络模型能够稳定地收敛于较高的精度,为后续的误差补偿提供了依据。
Optical wavelet transform is a real-time signal processing method which has developed in recent years. It combines the advantage of both wavelet transform and optical information. It is an effective approach to realize image data compression with the characteristic of parallelism and high-speed. It plays a significant role in some application fields which need to transmit abundant image data; for example, sunspace remote sensing and forest fireproofing. By now, it has got achievement on pattern recognition and texture segmentation. Because image data compression makes special demands on precision of the whole optical system, the reports on this field are rare. Placing a filter on the Fourier plane in an optical 4f system is a typical way to implement optical wavelet transform. Grating, spatial light modulator, computer- generated hologram and film can be used as the filter. This thesis focused on the research of the filter made of film. The whole thesis mainly consists of the following contents:
Comparing with laser printer and dye-sublimation printer, film recorder of converting digital image into film was determined judging by the need of this project. Considering inaccuracy brought in by man-made factors or hardware devices in the facture of optical filter, it is necessary to measure the error of it. A feasible experiment scheme for measuring the error was provided.
For the error, it is impractical to compensate after fabrication. In order to get relative accurate gray level on films, this paper borrowed ideas from predistortion applied to modify original image according to the nonlinear effects in manufacturing process. Therefore, a primary task of the paper is to set up the error model through artificial neural network, and then compensate the error in the system.
It is verified by lots of experiments that based on color images of 24 bits. The relation between original image gray level and the one made by film is non-linear, and gray levels of film image varies with different grey level combination of R、G、B channel. Meanwhile, the precision of each channel is different. The performance of R channel is the best, and B channels’is the worst. By polynomial curve fitting, the PSNR of film will increase 10~20dB.
Finally, the author made a conclusion and pointed out the problems that need to be explored further and the future research emphasis.
在典型的光学4f系统的谱面上加滤波片是实现光学小波变换的常用方法。可以用光栅、空间光调制器、计算全息、胶片作为滤波片。论文研究了在4f系统的谱面上加胶片作为滤波片来实现小波变换的方法。主要工作如下:
(1)研究了胶片的制作:分析了激光打印机、热升华打印机、胶片记录仪的优缺点,结合课题要求,确定胶片记录仪作为胶片制作仪器。
(2)分析了胶片的误差:设计了一系列的样本图像作为色标,通过扫描仪和Matlab对胶片制作的样本图像进行定量定性的分析,得到不同颜色通道的误差特性曲线。
(3)通过多项式拟合,寻求胶片图像均值与标准图像均值间的函数f(x),使f(x)与标准的数据点最接近。
(4)胶片一旦制作,难以修改,针对胶片的加工工艺误差,利用神经网络自学习特性对胶片制作误差进行分析,用胶片制作误差修正神经网络权值。若用网络输出结果对标准图像进行预畸变处理,则制作后胶片图像的灰度值趋进标准图像的灰度值。
研究结果表明:以24bit的彩色图像为分析基础,胶片图像的灰度级与原始图像的灰度级之间并非线性对应,R、G、B三通道的精度略有差异,R通道的性能最优,B通道的性能最差,当各通道灰度级以相同规律变化时,引入的误差最小。经多项式拟合拟合后,均值的平均峰值信噪比可提高10~20dB。而文中确定的网络模型能够稳定地收敛于较高的精度,为后续的误差补偿提供了依据。
Optical wavelet transform is a real-time signal processing method which has developed in recent years. It combines the advantage of both wavelet transform and optical information. It is an effective approach to realize image data compression with the characteristic of parallelism and high-speed. It plays a significant role in some application fields which need to transmit abundant image data; for example, sunspace remote sensing and forest fireproofing. By now, it has got achievement on pattern recognition and texture segmentation. Because image data compression makes special demands on precision of the whole optical system, the reports on this field are rare. Placing a filter on the Fourier plane in an optical 4f system is a typical way to implement optical wavelet transform. Grating, spatial light modulator, computer- generated hologram and film can be used as the filter. This thesis focused on the research of the filter made of film. The whole thesis mainly consists of the following contents:
Comparing with laser printer and dye-sublimation printer, film recorder of converting digital image into film was determined judging by the need of this project. Considering inaccuracy brought in by man-made factors or hardware devices in the facture of optical filter, it is necessary to measure the error of it. A feasible experiment scheme for measuring the error was provided.
For the error, it is impractical to compensate after fabrication. In order to get relative accurate gray level on films, this paper borrowed ideas from predistortion applied to modify original image according to the nonlinear effects in manufacturing process. Therefore, a primary task of the paper is to set up the error model through artificial neural network, and then compensate the error in the system.
It is verified by lots of experiments that based on color images of 24 bits. The relation between original image gray level and the one made by film is non-linear, and gray levels of film image varies with different grey level combination of R、G、B channel. Meanwhile, the precision of each channel is different. The performance of R channel is the best, and B channels’is the worst. By polynomial curve fitting, the PSNR of film will increase 10~20dB.
Finally, the author made a conclusion and pointed out the problems that need to be explored further and the future research emphasis.
引文
[1] A.Huang. Arichitectural consideration involved in the design of an optical digital computer. Proceeding of the IEEE. 1984,72(7):780-786.
[2] David Mendlovic,Ido Ouzieli,Irena Kiryuschev et al.. Two-dimensional wavelet transform achieved by computer-generated multireference matched filter and Dammann grating. Appl.Opt. 1995.12,34(35):8213-8219.
[3] 谢嘉宁. 光学信息处理过程的计算机仿真方法研究. 西北工业大学硕士论文. 2002.3.
[4] 王文陆. 光学子波变换及其在图像处理中的应用. 清华大学博士论文. 1995.6.
[5] A.Vander Lugt. Signal detection by complex spatial filtering. IEEE TRANSCTIONS ON INFORMATION THEORY. 1964.4:139-145.
[6] Yunlong Sheng,Danny Roberge,Harold H.Szu. Optical wavelet transform. Opt. Eng. 1992.9, 31(9):1840-1845.
[7] Wang W L,Jin G F,Yan Y B et al.. Image Feature Extraction with optical Haar wavelet transform. Opt.Eng. 1995,134(4):1238-1242.
[8] Peter G.Block,Steven K.Rogers,and Dennis W. Ruck. Optical wavelet transform from computer-generated holography. Appl.Opt. 1994.8,33(23):5275-5278.
[9] Y L Sheng,Taiwei Lu,Danny Roberge,H John Caulfield. N4 Inplementadon of a 2-D wavelet transform. Opt.Eng. 1992,l3l(9):1859-1864.
[10] Meiyuan Wen,Shizhuo Yin,Purwosumarto Purwardi et al.. Wavelet matched filter using a photorefractive crystal. Opt.Commun. 1993:325-330.
[11] Ido Ouzieli,David Mendlovic. Two-dimensional wavelet processor. Appl.Opt. 1996.10,35(29):5839-5846.
[12] 赵维义,宋菲君,俞蕾等. 光学小波变换. 物理. 2000.3,29(3):162-168.
[13] 华家宁,陈家壁. 光学小波变换初探. 南京师大学报(自然科学版). 1994,17(4):31-37.
[14] 聂守平,杨罕,李爱民等. 一种新颖的光信息处理方法——子波变换. 南京理工大学学报. 1995.2,19(1):13-16.
[15] D.Mendlovic. Continuous two-dimensinal on-axis optical wavelet transformer and wavelet processor by wavelet processor with white-light illumination. Appl.Opt. 1998.3,37(8):1279-1282.
[16] 陈修建,田逢春. 光学小波图像传输系统中的小波滤波器误差测量. 重庆大学学报. 2004,27(8):79-82.
[17] 马晶,谭立英,冉启文. 光学小波滤波理论初探. 中国激光. 1999. 4,26(4):343-346.
[18] 陈梨. 用光栅法实现光学小波滤波器. 重庆大学硕士论文. 2006:9-18.
[19] 冉启文,潭立英. 小波分析与分数傅立叶变换及应用. 国防工业出版社. 2002 年.
[20] 冯文毅,黄高贵,严瑛白等. 多通道体全息子波相关器及其在指纹识别中的应用. 光学学报. 2000,20(3):370-375.
[21] A.D.Fisher,J.N.Lee. The current status of two-dimensional spatial light modulator technology. SPIE. 1986,634:352-371.
[22] 张以谟,秦玉文,黄战华等. 实时高效联合变换相关器. 光学学报. 1995.7,15( 7):883-887.
[23] 陈怀新,陈祯培,陈晓伟. 基于相移功率谱相减的二元联合变换相关识别. 中国激光. 1998.4,25(4):353-358.
[24] F.T.S. Yu,S. Jutamulia,T.W. Liu et al.. Adaptive real-time pattern recognition using a liquid crystal TV based joint transform correlator. Appl.Opt. 1987,26(8):1370-1372.
[25] Y. Sheng,G. Paul-Hus. Optical on-axis imperfect phase-only correlator using liquid-crystal television. Appl.Opt. 1993,32(29):5782-5785.
[26] 陈家祯,郑子华,陈利永. MATLAB 中二元傅里叶变换计算全息图的算法. 福建师范大学学报(自然科学版). 2003,19(13):22-25.
[27] 康果果,谢敬辉,齐月静等. 用 Matlab 软件和 LCD 实现数字全息图的制作和再现. 北京理工大学学报. 2005.6,25(6):509-511.
[28] 陈鹤鸣,贾小兵,施伟华. 应用计算全息制作的子波匹配滤波器实现光学子波变换的研究. 光学学报. 1998.8,18(8):1087-1090.
[29] 虞祖良,金国藩. 计算机制全息图. 北京清华大学出版社. 1984 年.
[30] 陈祯培,陈怀新,戴建明. 减少计算全息再现散斑的多图像平均方法. 光电子?激光. 1997.10,8(5):383-386.
[31] 吕德章. 降低 CGH 再现像量化噪声的新方法. 光学仪器. 1999.9,21(3):25-28.
[32] 陈怀新,陈祯培,苏红军. 提高计算全息图再现质量的迭代傅立叶变换算法. 光电工程. 1997.6,24(3):62-66.
[33] 高峰,朱建华,黄奇忠等. 电子束直写计算全息图. 中国激光. 2001.6,28(6):556-558.
[34] 吕乃光. 傅里叶光学. 机械工业出版社. 1988 年.
[35] 苏显渝,李继陶. 信息光学. 北京科学出版社. 1999 年.
[36] 陈怀新,隋展,陈祯培等. 液晶电视(LCTV)的光学调制特性及其应用. 中国激光,2000.8,27(8):741-745.
[37] 高泽华,黄献烈,赖虹凯. 使用液晶光阀的实时计算机制全息图的产生. 厦门大学学报(自然科学版). 1998.1,37(1):37-40.
[38] 张援朝. 办公设备维修讲座(四)异彩纷呈的打印机世界:打印机的分类、原理、结构和维护. 家电科技:维修与培训. 2003,(4):7-11.
[39] 董轩宙,唐应武,张复实等. 数字图像的胶片记录. 感光材料. 1998,(3):19-21.
[40] 雷振宇,何金泉. CRT 型胶片记录仪的原理和使用. 影视技术. 2000,(8):16-18.
[41] 张兆礼,赵春晖,梅晓丹. 现代图像处理技术及 Matlab 实现. 人民邮电出版社. 2001 年.
[42] 扫描仪的分辨率与测试精度. HP LFP Forum:13-20.
[43] 张景禹. 彩色胶片. 北京轻工业出版社. 1987 年.
[44] Dekbruger D,Hunt B R. Image processing and neural networks for recognition of cartographic area features. Pattern Recognilion. 1994,27(4):461-483.
[45] Bajaj R,Chaudhury S. Signature verification using multiple neural classifiers. Pattern Recognition. 1997,30(1):1-7.
[46] Penedo M G,Carreira M J,Mosquera A. Computer-aided diagnosis: A neural-network based approach to lung nodule detection. IEEE Transactions on Medical Imaging. 1998,17(6):872- 880.
[47] 匡锦瑜,丁俊民,陈士碌. 图像的小波系数神经网络预测编码. 北京师范大学学报. 1999,35(4):464-467.
[48] Kim H B,Nam K H. Object recognition of one-DOF tools by a back propagation neural network. IEEE Transactions on Neural Networks. 1995,6(2):484-487.
[49] Robert Hecht-Nielson. Theory of the Back-Propagation Neural Network. IJCNN. 1989,I:583-604.
[50] 从爽. 面向 MATLAB 工具箱的神经网络理论与应用. 中国科技大学出版社. 2003 年.
[51] 许锋,卢建刚,孙优贤. 神经网络在图像处理中的应用. 信息与控制. 2003.8,32(4):84-96.
[52] Joon K P. Image restoration using a modified Hopfield network. IEEE Transactions on Image Processing. 1992,1(1):49-63.
[53] 杨行峻. 人工神经网络与盲信号处理. 北京清华大学出版社. 2003 年.
[54] 邹广宇,王宏峰,汪定伟. 基于神经元网络模型的城市用水量预测. 信息与控制. 2004,33(3):364-368.
[55] Freeman J A,Skapura D M. Neural Networks:Algorithms,Applications and Programming Techniques. Red-wood City,CA,USA:Addison-Wesley. 1991.
[56] 闻新,周露,李翔等. MATLAB 神经网络仿真与应用. 北京科学出版社. 2003 年.
[57] 章剑雄. 复杂系统的神经网络建模及仿真研究. 浙江工业大学硕士学位论文. 2003:37-40.
[58] 李峥. 基于计算机视觉的蔬菜颜色检测系统研究. 吉林大学硕士学位论文. 2004:57-60.
[59] 勒蕃. 神经网络与神经计算机原理应用. 成都西南交通大学出版社. 1991 年.
[60] 蒋宗礼. 人工神经网络导论. 北京高等教育出版社. 2001 年.
[61] Vrhel MJ,Trussell HJ. Color scanner calibration via a neural network. IEEE International Conference on Acoustics,Speech,and Signal Processing. 1999,(6):3465-3468.
[62] Shiro Usui,Yoshfumi Arai,Shigeki Nakauchi. Neural networks for device-independent digital color imaging. Information Sciences. 2000,123(122):115-125.
[63] 韩柯,王汇源. 基于主元分析和神经网络的人脸识别方法. 山东大学学报(工学版). 2004.4,34(2):55-58.
[64] 肖幸娟,王能超,周叶. 两种序的二维 Walsh 变换在人脸识别中的应用. 华中科技大学学报(自然科学版). 2004.11,32(11):109-110.
[65] 高大启. 有教师的线性基本函数前向三层神经网络结构研究. 计算机学报. 1998.1,21(1):80-86.
[66] Sivakumar K,Desai U B. Image restoration using a multilayer perceptron with a multilevel sigmoid function. IEEE Transaction on Signal Processing. 1993.5,41(5):2018-2022.
[67] Asmatullah,Mirza A.M.,Khan A.. Blind image restoration using multilayer backpropagator. Proceedings IEEE INMIC. 2003:55- 58.
[68] 徐艳芳,刘文耀,左坤隆等. 彩色扫描仪的特征化. 光学精密工程. 2004.2,12(1):15-20.
[69] 李丽,王振领. MATLAB工程计算及应用. 北京人民邮电出版社. 2001年.
[2] David Mendlovic,Ido Ouzieli,Irena Kiryuschev et al.. Two-dimensional wavelet transform achieved by computer-generated multireference matched filter and Dammann grating. Appl.Opt. 1995.12,34(35):8213-8219.
[3] 谢嘉宁. 光学信息处理过程的计算机仿真方法研究. 西北工业大学硕士论文. 2002.3.
[4] 王文陆. 光学子波变换及其在图像处理中的应用. 清华大学博士论文. 1995.6.
[5] A.Vander Lugt. Signal detection by complex spatial filtering. IEEE TRANSCTIONS ON INFORMATION THEORY. 1964.4:139-145.
[6] Yunlong Sheng,Danny Roberge,Harold H.Szu. Optical wavelet transform. Opt. Eng. 1992.9, 31(9):1840-1845.
[7] Wang W L,Jin G F,Yan Y B et al.. Image Feature Extraction with optical Haar wavelet transform. Opt.Eng. 1995,134(4):1238-1242.
[8] Peter G.Block,Steven K.Rogers,and Dennis W. Ruck. Optical wavelet transform from computer-generated holography. Appl.Opt. 1994.8,33(23):5275-5278.
[9] Y L Sheng,Taiwei Lu,Danny Roberge,H John Caulfield. N4 Inplementadon of a 2-D wavelet transform. Opt.Eng. 1992,l3l(9):1859-1864.
[10] Meiyuan Wen,Shizhuo Yin,Purwosumarto Purwardi et al.. Wavelet matched filter using a photorefractive crystal. Opt.Commun. 1993:325-330.
[11] Ido Ouzieli,David Mendlovic. Two-dimensional wavelet processor. Appl.Opt. 1996.10,35(29):5839-5846.
[12] 赵维义,宋菲君,俞蕾等. 光学小波变换. 物理. 2000.3,29(3):162-168.
[13] 华家宁,陈家壁. 光学小波变换初探. 南京师大学报(自然科学版). 1994,17(4):31-37.
[14] 聂守平,杨罕,李爱民等. 一种新颖的光信息处理方法——子波变换. 南京理工大学学报. 1995.2,19(1):13-16.
[15] D.Mendlovic. Continuous two-dimensinal on-axis optical wavelet transformer and wavelet processor by wavelet processor with white-light illumination. Appl.Opt. 1998.3,37(8):1279-1282.
[16] 陈修建,田逢春. 光学小波图像传输系统中的小波滤波器误差测量. 重庆大学学报. 2004,27(8):79-82.
[17] 马晶,谭立英,冉启文. 光学小波滤波理论初探. 中国激光. 1999. 4,26(4):343-346.
[18] 陈梨. 用光栅法实现光学小波滤波器. 重庆大学硕士论文. 2006:9-18.
[19] 冉启文,潭立英. 小波分析与分数傅立叶变换及应用. 国防工业出版社. 2002 年.
[20] 冯文毅,黄高贵,严瑛白等. 多通道体全息子波相关器及其在指纹识别中的应用. 光学学报. 2000,20(3):370-375.
[21] A.D.Fisher,J.N.Lee. The current status of two-dimensional spatial light modulator technology. SPIE. 1986,634:352-371.
[22] 张以谟,秦玉文,黄战华等. 实时高效联合变换相关器. 光学学报. 1995.7,15( 7):883-887.
[23] 陈怀新,陈祯培,陈晓伟. 基于相移功率谱相减的二元联合变换相关识别. 中国激光. 1998.4,25(4):353-358.
[24] F.T.S. Yu,S. Jutamulia,T.W. Liu et al.. Adaptive real-time pattern recognition using a liquid crystal TV based joint transform correlator. Appl.Opt. 1987,26(8):1370-1372.
[25] Y. Sheng,G. Paul-Hus. Optical on-axis imperfect phase-only correlator using liquid-crystal television. Appl.Opt. 1993,32(29):5782-5785.
[26] 陈家祯,郑子华,陈利永. MATLAB 中二元傅里叶变换计算全息图的算法. 福建师范大学学报(自然科学版). 2003,19(13):22-25.
[27] 康果果,谢敬辉,齐月静等. 用 Matlab 软件和 LCD 实现数字全息图的制作和再现. 北京理工大学学报. 2005.6,25(6):509-511.
[28] 陈鹤鸣,贾小兵,施伟华. 应用计算全息制作的子波匹配滤波器实现光学子波变换的研究. 光学学报. 1998.8,18(8):1087-1090.
[29] 虞祖良,金国藩. 计算机制全息图. 北京清华大学出版社. 1984 年.
[30] 陈祯培,陈怀新,戴建明. 减少计算全息再现散斑的多图像平均方法. 光电子?激光. 1997.10,8(5):383-386.
[31] 吕德章. 降低 CGH 再现像量化噪声的新方法. 光学仪器. 1999.9,21(3):25-28.
[32] 陈怀新,陈祯培,苏红军. 提高计算全息图再现质量的迭代傅立叶变换算法. 光电工程. 1997.6,24(3):62-66.
[33] 高峰,朱建华,黄奇忠等. 电子束直写计算全息图. 中国激光. 2001.6,28(6):556-558.
[34] 吕乃光. 傅里叶光学. 机械工业出版社. 1988 年.
[35] 苏显渝,李继陶. 信息光学. 北京科学出版社. 1999 年.
[36] 陈怀新,隋展,陈祯培等. 液晶电视(LCTV)的光学调制特性及其应用. 中国激光,2000.8,27(8):741-745.
[37] 高泽华,黄献烈,赖虹凯. 使用液晶光阀的实时计算机制全息图的产生. 厦门大学学报(自然科学版). 1998.1,37(1):37-40.
[38] 张援朝. 办公设备维修讲座(四)异彩纷呈的打印机世界:打印机的分类、原理、结构和维护. 家电科技:维修与培训. 2003,(4):7-11.
[39] 董轩宙,唐应武,张复实等. 数字图像的胶片记录. 感光材料. 1998,(3):19-21.
[40] 雷振宇,何金泉. CRT 型胶片记录仪的原理和使用. 影视技术. 2000,(8):16-18.
[41] 张兆礼,赵春晖,梅晓丹. 现代图像处理技术及 Matlab 实现. 人民邮电出版社. 2001 年.
[42] 扫描仪的分辨率与测试精度. HP LFP Forum:13-20.
[43] 张景禹. 彩色胶片. 北京轻工业出版社. 1987 年.
[44] Dekbruger D,Hunt B R. Image processing and neural networks for recognition of cartographic area features. Pattern Recognilion. 1994,27(4):461-483.
[45] Bajaj R,Chaudhury S. Signature verification using multiple neural classifiers. Pattern Recognition. 1997,30(1):1-7.
[46] Penedo M G,Carreira M J,Mosquera A. Computer-aided diagnosis: A neural-network based approach to lung nodule detection. IEEE Transactions on Medical Imaging. 1998,17(6):872- 880.
[47] 匡锦瑜,丁俊民,陈士碌. 图像的小波系数神经网络预测编码. 北京师范大学学报. 1999,35(4):464-467.
[48] Kim H B,Nam K H. Object recognition of one-DOF tools by a back propagation neural network. IEEE Transactions on Neural Networks. 1995,6(2):484-487.
[49] Robert Hecht-Nielson. Theory of the Back-Propagation Neural Network. IJCNN. 1989,I:583-604.
[50] 从爽. 面向 MATLAB 工具箱的神经网络理论与应用. 中国科技大学出版社. 2003 年.
[51] 许锋,卢建刚,孙优贤. 神经网络在图像处理中的应用. 信息与控制. 2003.8,32(4):84-96.
[52] Joon K P. Image restoration using a modified Hopfield network. IEEE Transactions on Image Processing. 1992,1(1):49-63.
[53] 杨行峻. 人工神经网络与盲信号处理. 北京清华大学出版社. 2003 年.
[54] 邹广宇,王宏峰,汪定伟. 基于神经元网络模型的城市用水量预测. 信息与控制. 2004,33(3):364-368.
[55] Freeman J A,Skapura D M. Neural Networks:Algorithms,Applications and Programming Techniques. Red-wood City,CA,USA:Addison-Wesley. 1991.
[56] 闻新,周露,李翔等. MATLAB 神经网络仿真与应用. 北京科学出版社. 2003 年.
[57] 章剑雄. 复杂系统的神经网络建模及仿真研究. 浙江工业大学硕士学位论文. 2003:37-40.
[58] 李峥. 基于计算机视觉的蔬菜颜色检测系统研究. 吉林大学硕士学位论文. 2004:57-60.
[59] 勒蕃. 神经网络与神经计算机原理应用. 成都西南交通大学出版社. 1991 年.
[60] 蒋宗礼. 人工神经网络导论. 北京高等教育出版社. 2001 年.
[61] Vrhel MJ,Trussell HJ. Color scanner calibration via a neural network. IEEE International Conference on Acoustics,Speech,and Signal Processing. 1999,(6):3465-3468.
[62] Shiro Usui,Yoshfumi Arai,Shigeki Nakauchi. Neural networks for device-independent digital color imaging. Information Sciences. 2000,123(122):115-125.
[63] 韩柯,王汇源. 基于主元分析和神经网络的人脸识别方法. 山东大学学报(工学版). 2004.4,34(2):55-58.
[64] 肖幸娟,王能超,周叶. 两种序的二维 Walsh 变换在人脸识别中的应用. 华中科技大学学报(自然科学版). 2004.11,32(11):109-110.
[65] 高大启. 有教师的线性基本函数前向三层神经网络结构研究. 计算机学报. 1998.1,21(1):80-86.
[66] Sivakumar K,Desai U B. Image restoration using a multilayer perceptron with a multilevel sigmoid function. IEEE Transaction on Signal Processing. 1993.5,41(5):2018-2022.
[67] Asmatullah,Mirza A.M.,Khan A.. Blind image restoration using multilayer backpropagator. Proceedings IEEE INMIC. 2003:55- 58.
[68] 徐艳芳,刘文耀,左坤隆等. 彩色扫描仪的特征化. 光学精密工程. 2004.2,12(1):15-20.
[69] 李丽,王振领. MATLAB工程计算及应用. 北京人民邮电出版社. 2001年.