基于无参考质量评价模型的静脉图像采集方法
|
摘要
由于不同人的手背静脉属性间存在较大差异,因此对于不同静脉对象,在固定采集系统参数条件下很难都采集到高质量的静脉图像,这里提出了一种针对静脉特点的质量评价模型,并设计了基于评价结果的自寻优静脉图像采集方法.首先,提出了基于关键信息熵的测度函数,衡量了静脉信息的完整性;其次,提出了基于轮廓波分解的测度函数,用于评价静脉方向性信息的丰富性;再次,将两种测度函数融合,构成了客观的无参考的质量评价模型;最后,在图像自寻优过程中,提出了迭代淘汰机制,克服了最速下降法在寻优过程易陷入局部最优的缺陷.实验表明,提出的质量评价模型是可控的,且满足人眼视觉系统的视觉特性,同时,通过提出的迭代淘汰机制,降低了寻优过程的迭代次数,保证了采集系统的实时性要求.
Because of dorsal vein properties difference between different persons,for different vein objects not all vein images with high qualities can be acquired when fixing acquisition system parameters,in the paper a newimage quality assessment model is proposed according to vein characteristics,moreover,vein image can be acquired through self optimization based on assessment result. First,a measure function based on key information entropy is proposed,which measures vein information completeness; Second,another measure function based on Contourlet decomposition is proposed,which is used to judge whether vein directional information is rich or not; Third,the objective vein image quality assessment model without reference is formed by fusing the two measure functions. In final,iteration elimination method is proposed in order to overcome the defect of steepest descent method,which is easy to fall into local optimum during optimization process. Experiments showthe proposed quality assessment model is controllable and can meet the characteristics of human visual system,in the meantime the number of iterations can be reduced effectively through the proposed iteration elimination method,and the real-time requirement of acquisition system can be ensured.
Because of dorsal vein properties difference between different persons,for different vein objects not all vein images with high qualities can be acquired when fixing acquisition system parameters,in the paper a newimage quality assessment model is proposed according to vein characteristics,moreover,vein image can be acquired through self optimization based on assessment result. First,a measure function based on key information entropy is proposed,which measures vein information completeness; Second,another measure function based on Contourlet decomposition is proposed,which is used to judge whether vein directional information is rich or not; Third,the objective vein image quality assessment model without reference is formed by fusing the two measure functions. In final,iteration elimination method is proposed in order to overcome the defect of steepest descent method,which is easy to fall into local optimum during optimization process. Experiments showthe proposed quality assessment model is controllable and can meet the characteristics of human visual system,in the meantime the number of iterations can be reduced effectively through the proposed iteration elimination method,and the real-time requirement of acquisition system can be ensured.
引文
[1]殷俊,杨万扣.核稀疏保持投影及生物特征识别应用[J].电子学报,2013,41(4):639-645.Yin J,Wang W K.Kernel sparsity preserving projections and its application to biometrics[J].Acta Electronica Sinica,2013,41(4):639-645.(in Chinese)
[2]Cross J M,Smith C L.Thermographic imaging of subcutaneous vascular netw ork of the back of the hand for biometric identification[A].Proceedings of IEEE 29th International Carnahan Conference on Security Technology[C].England,1995.20-35.
[3]Lin C L,Fan K C.Biometric verification using thermal images of palm-dorsa vein patterns[J].Transactions on Circuits and Systems for Video Technology,2004,14(2):199-213.
[4]Crisan C,Tarnovan L G,Crisan T E.A low cost vein detection system using near infrared radiation[A].2007 IEEE Sensors Applications Symposium[C].SAS,San Diego,CA,United states,2007.51-56.
[5]Wang J G,Yau W Y,Suwandy A,et al.Person recognition by fusing palmprint and palmvein images based on“Laplacianpalm”representation[J].Pattern Recognition,2008,41(5):1514-1527.
[6]Zhao S,Wang Y D,Wang Y H.Biometric identification based on low-quality hand vein pattern images[A].Proceedings of the 7th International Conference on M achine Learning and Cybernetics[C].ICM LC,Kunming,China.2008,1172-1177.
[7]王云新,刘铁根,朱均超等.嵌入式人体手背静脉图像采集系统的研制[J].仪器仪表学报,2009,30(2):308-312.Wang Y X,Liu T G,Zhu J C,et al.Development of an embedded image acquisition system for human dorsal hand vein[J].Chinese Journal of Scientific Instrument,2009,30(2):308-312.(in Chinese)
[8]Yuan X,Song Y D,Wei X Y.Parallel sub-neural network system for hand vein pattern recognition[J].Chinese Optics Letters,2011,9(5):051002.
[9]Liu Z,Song S L.An embedded real-time finger-vein recognitionsystem for mobile devices[J].IEEE Transactions on Consumer Electronics,2012,58(2):522-527.
[10]Zhao Y.Acquisition and preprocessing of hand vein image[A].2011 International Conference on Multimedia Technology[C].Hangzhou,China,2011.5727-5729.
[11]苑玮琦,杨国天,李威.基于波长选择的手掌静脉采集系统研究[J].激光与红外,2011,41(2):234-239.Yuan W Q,Yang G T,Li W.Research on palm vein acquisition system based on w avelength choice[J].Laser&Infrared,2011,41(2):234-239.(in Chinese)
[12]Zanin M,Zunino L,Rosso O A,et al.Permutation entropy and its main biomedical andeconophysics applications:a review[J].Entropy,2012,14(8):1553-1577.
[13]Niu J,Sivakumar B.Scale-dependent synthetic streamflow generation using a continuous w avelet transform[J].Journal of Hydrology,2013,496(24):71-78.
[14]Wang X Y,Yang H Y,Zhang Y,et al.Image denoising using SVM classification in nonsubsampled contourlet transform domain[J].Information Science,2013,246(10):155-176.
[15]Zargari F,Mosleh A.A new content based image retrieval method using contourlet transform[J].Journal of Computer&Robotics,2010,2(1):45-51.
[16]Doel K V D,Ascher U.The chaotic nature of faster gradient descent methods[J].Journal of Scientific Computing,2012,51(3):560-581.
[2]Cross J M,Smith C L.Thermographic imaging of subcutaneous vascular netw ork of the back of the hand for biometric identification[A].Proceedings of IEEE 29th International Carnahan Conference on Security Technology[C].England,1995.20-35.
[3]Lin C L,Fan K C.Biometric verification using thermal images of palm-dorsa vein patterns[J].Transactions on Circuits and Systems for Video Technology,2004,14(2):199-213.
[4]Crisan C,Tarnovan L G,Crisan T E.A low cost vein detection system using near infrared radiation[A].2007 IEEE Sensors Applications Symposium[C].SAS,San Diego,CA,United states,2007.51-56.
[5]Wang J G,Yau W Y,Suwandy A,et al.Person recognition by fusing palmprint and palmvein images based on“Laplacianpalm”representation[J].Pattern Recognition,2008,41(5):1514-1527.
[6]Zhao S,Wang Y D,Wang Y H.Biometric identification based on low-quality hand vein pattern images[A].Proceedings of the 7th International Conference on M achine Learning and Cybernetics[C].ICM LC,Kunming,China.2008,1172-1177.
[7]王云新,刘铁根,朱均超等.嵌入式人体手背静脉图像采集系统的研制[J].仪器仪表学报,2009,30(2):308-312.Wang Y X,Liu T G,Zhu J C,et al.Development of an embedded image acquisition system for human dorsal hand vein[J].Chinese Journal of Scientific Instrument,2009,30(2):308-312.(in Chinese)
[8]Yuan X,Song Y D,Wei X Y.Parallel sub-neural network system for hand vein pattern recognition[J].Chinese Optics Letters,2011,9(5):051002.
[9]Liu Z,Song S L.An embedded real-time finger-vein recognitionsystem for mobile devices[J].IEEE Transactions on Consumer Electronics,2012,58(2):522-527.
[10]Zhao Y.Acquisition and preprocessing of hand vein image[A].2011 International Conference on Multimedia Technology[C].Hangzhou,China,2011.5727-5729.
[11]苑玮琦,杨国天,李威.基于波长选择的手掌静脉采集系统研究[J].激光与红外,2011,41(2):234-239.Yuan W Q,Yang G T,Li W.Research on palm vein acquisition system based on w avelength choice[J].Laser&Infrared,2011,41(2):234-239.(in Chinese)
[12]Zanin M,Zunino L,Rosso O A,et al.Permutation entropy and its main biomedical andeconophysics applications:a review[J].Entropy,2012,14(8):1553-1577.
[13]Niu J,Sivakumar B.Scale-dependent synthetic streamflow generation using a continuous w avelet transform[J].Journal of Hydrology,2013,496(24):71-78.
[14]Wang X Y,Yang H Y,Zhang Y,et al.Image denoising using SVM classification in nonsubsampled contourlet transform domain[J].Information Science,2013,246(10):155-176.
[15]Zargari F,Mosleh A.A new content based image retrieval method using contourlet transform[J].Journal of Computer&Robotics,2010,2(1):45-51.
[16]Doel K V D,Ascher U.The chaotic nature of faster gradient descent methods[J].Journal of Scientific Computing,2012,51(3):560-581.