可调对比度目标源装置中对比度的标定

王素华; 沈湘衡; 叶露

doi:10.3788/OPE.20122005.0949

您当前的位置：

首页 >

文章列表页 >

可调对比度目标源装置中对比度的标定

现代应用光学 | 更新时间：2020-08-12

- 可调对比度目标源装置中对比度的标定
- Calibration of contrast for adjustable contrast optical target equipment
- 光学精密工程 2012年20卷第5期页码：949-956
- 作者机构：
  
  1. 中国科学院长春光学精密机械与物理研究所,吉林长春,中国,130033
  2. 中国科学院研究生院北京,100039
  3. 长春职业技术学院,吉林长春 130033
- 作者简介：
- 基金信息：
  
  中国科学院创新基金资助项目(No.YZ200904)()
- DOI：10.3788/OPE.20122005.0949
  中图分类号： TB96;TP391.4
- 收稿日期：2011-11-05，
  
  修回日期：2011-11-29，
  
  网络出版日期：2012-05-10，
  
  纸质出版日期：2012-05-10
- 稿件说明：
移动端阅览
王素华, 沈湘衡, 叶露. 可调对比度目标源装置中对比度的标定[J]. 光学精密工程, 2012,20(5): 949-956

WANG Su-hua, SHEN Xiang-heng, YE Lu. Calibration of contrast for adjustable contrast optical target equipment[J]. Editorial Office of Optics and Precision Engineering, 2012,20(5): 949-956
王素华, 沈湘衡, 叶露. 可调对比度目标源装置中对比度的标定[J]. 光学精密工程, 2012,20(5): 949-956 DOI： 10.3788/OPE.20122005.0949.

WANG Su-hua, SHEN Xiang-heng, YE Lu. Calibration of contrast for adjustable contrast optical target equipment[J]. Editorial Office of Optics and Precision Engineering, 2012,20(5): 949-956 DOI： 10.3788/OPE.20122005.0949.

摘要

搭建了可调对比度目标源装置

研究了图像对比度和光学对比度的关系

提出了用改进的BP神经网络标定对比度的方法。首先

设计了用于对比度标定的BP神经网络模型。然后

利用LM(Levenberg-Marquardt)算法结合缩放法改进神经网络以提高其收敛速度及泛化能力。最后

通过可调对比度目标源装置实验平台

由测量的辐照度得出了对应的图像对比度数据

使该装置可以通过调节辐照度实时获得规定的对比度。与传统BP神经网络方法相比

改进后的BP神经网络收敛速度快

泛化能力强。标定精度比经典BP算法提高了100倍

比最速下降法提高了10倍。训练次数仅需2 876次时

对比度的标定值与目标值的误差最大值是0.01%

训练均方误差收敛为0.000 459 441

测试误差收敛为0.000 467 003

满足了对检验装置中对比度标定的需要。

Abstract

An adjustable contrast optical target equipment was constructed. After researching the relationship between image contrast and optical contrast

a contrast calibration method by the improved Back Propagation(BP) neural network was proposed. Firstly

the BP neural network model was designed for calibrating the contrast. Then

by combining the Levenberg-Marquardt(LM) with Shrinking-Magnifying Approach

the BP neural network was improved to optimize the convergence speed and generalization ability. Finally

based on the experimental platform of the adjustable-contrast target

the image contrast was obtained by measured radiation data. Comparing with the traditional BP algorithm

the improved one has a better convergence speed and generalization ability. Its calibration accuracy has been improved by 100 times and by 10 times as compared with those of the traditional BP network and the steepest descent method

respectively. When the training times is to be only 2 876 times

the maximum error between calibration value and target calibration value for the contrast is 0.01%

the training mean square error converges is 0.000 459 441

and the test error converges is 0.000 467 003. These results demonstrate that the algorithm is feasible and can meet the demands for contrast calibration in the equipment.

关键词

Keywords

references

叶露,谷立山,沈湘衡. 可调对比度光学无穷远目标源设计 [J]. 应用光学,2010,31(5):681-684. YE L, GU L SH, SHEN X H. Design of adjustable contrast optical target [J]. Jouranal of Applied Optics, 2010,31(5): 681-684. (in Chinese)[2] 于国着. 小波变换在低对比度目标相关探测的应用 . 长春:长春理工大学 ,2009. YU G ZH. Application of wavelet transform in low contrast target correlation detection . Changchun: Changchun University of Science and Technology, 2009. (in Chinese)[3] 黄杰贤,李迪,叶峰,等. 挠性印制电路板焊盘表面缺陷的检测 [J]. 光学精密工程,2010,18(11):2444-2453. HUANG J X, LI D, YE F, et al.. Detection of surface defection of solder on flexible printed circuit[J]. Opt. Precision Eng., 2010, 18(11): 2444-2453. (in Chinese)[4] 王刚,禹秉熙. 基于对比度的空中红外点目标探测距离估计方法 [J]. 光学精密工程,2002,10(3):276-280. WANG G, YU B X. Approach to estimate infrared point-target detection range against sky backgroung based on contrast[J]. Opt. Precision Eng., 2002, 10(3): 276-280. (in Chinese)[5] 马国锐,王长力,眭海刚,等. 弱小目标可见光传感器成像特性研究 [J]. 无线电工程,2010,40(1):48-51. MA G R, WANG CH L, SUI H G, et al.. Small target imaging mechanism of visual CCD Sensor[J]. Radio Engineering, 2010, 40(1): 48-51. (in Chinese)[6] HAYKIN S. Neural Networks and Learning Machines[M]. Third Edition. Prentice Hall, 2009.[7] 郭旭东,严荣国,颜国正. 胶囊内窥镜无线遥测定位的校正 [J]. 光学精密工程,2010,18(12):2650-2655. GUO X D, YAN R G, YAN G ZH. Calibration method for wirelessly localizing capsule endoscopy[J]. Opt. Precision Eng., 2010, 18(12): 2650-2655. (in Chinese)[8] FENG N, WANG F, QIU Y H. Novel approach for promoting the generalization ability of neural networks[J]. International Journal of Information and Communication Engineering, 2006, 2(2):131-135.[9] MORé J J. The levenberg-marquardt algorithm implementation and theory[J]. Numerical Analysis, 1978, 630:105-116.[10] RANGANATHAN A. The levenberg-marquardt algorithm . (2007-07-15). http: // www. cc. gatech. edu/ ~ananth/docs/lmtut.pdf.[11] ZHANG N, SHEN X H. System identificatioin of tracking error and evaluation of tracking[J]. SPIE 2009,7383:73832F-1-73832F-9.[12] FENG L. Research on the estimating model of the stock market price based on the LM-BP neural network . 2010 Fourth Internamtional Conference on Genetic and Evolutionary Computing, 2010: 562-565.[13] 王自强,李银妹,楼立人,等. BP神经网络用于光镊力的非线性修正 [J]. 光学精密工程,2008,16(1):6-10. WANG Z Q, LI Y M, LOU L R, et al.. Application of BP neural network to nonlinearity correction of optical tweezer force[J]. Opt. Precision Eng., 2008, 16(1): 6-10. (in Chinese)[14] WU J D, LI N, YANG H J. Risk evaluation of heavy snow disasters using BP artificial neural network: the case of Xilingol in Inner Mongolia[J]. Springer, 2008,22:719-725.[15] SHI CH B, JJA X D, LI S, et al.. A BP neural network model for the sea state recognition using laser altimeter[J]. SPIE, 2009,7382: 738251-1-738251-7.

浏览量

616

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于神经网络的全斯托克斯光电探测器

基于圆卷积神经网络的粘连导电粒子检测

面向领域自适应的部分最优传输高光谱图像分类

基于热传导和卷积神经网络的磨床主轴热误差预测

基于刚度预测-神经网络的调姿平台误差补偿