跳到主要內容

臺灣博碩士論文加值系統

(3.236.110.106) 您好!臺灣時間:2021/07/25 23:11
字體大小: 字級放大   字級縮小   預設字形  
回查詢結果 :::

詳目顯示

我願授權國圖
: 
twitterline
研究生:楊緒屏
研究生(外文):Shiuh-Pyng Yang
論文名稱:利用靜脈圖形做掌背辨識
論文名稱(外文):Back-palm recognition based on vein pattern
指導教授:王敬文
指導教授(外文):Jing-Wein Wang
學位類別:碩士
校院名稱:國立高雄應用科技大學
系所名稱:光電與通訊研究所
學門:工程學門
學類:電資工程學類
論文種類:學術論文
論文出版年:2009
畢業學年度:97
語文別:英文
論文頁數:101
中文關鍵詞:中值濾波器斷開掌背靜脈中值濾波器斷開閉合骨架化二元樹正Y圖形倒Y圖形平均坐標掌背靜脈
外文關鍵詞:Back-palm veinMedian filterOpening
相關次數:
  • 被引用被引用:0
  • 點閱點閱:483
  • 評分評分:
  • 下載下載:0
  • 收藏至我的研究室書目清單書目收藏:1
隨著科技的進步,人們開始使用各種生物特徵唯一的特性來當作身份識別的工具,如虹膜、指紋、人臉、血管等等,皆為驗證身份重要的工具。掌背靜脈特徵具有獨特性與唯一性,且不隨時間而改變,其性質適合一對多辨識,且較指紋與人臉更適合作為辨識系統,綜觀所有生物辨識技術,掌背靜脈領域在未來是值得研究與探討。掌背靜脈的生物特徵位於皮下組織3公釐位置,不易被複製與被使用犯罪。當使用者將手置於我們的彩色攝影機取樣系統(SONY SSC-E473)中,近紅外線光照射掌背靜脈,靜脈中紅血球吸收紅外線光反映於影像中呈現黑色粗糙線條,其餘部份則呈現青綠色,當完成分類後與原先儲存影像比對即可完成初步辨識。我們在這篇論文裡提出一個新的掌背靜脈血管特徵抽取、分類與辨識方法,首先將近紅外線攝影機截取的影像,使用高通加強濾波器作影像加強,並想出新的前景取出公式,參考拉卜拉斯與高斯觀念設計一個11×11遮罩濾波器,藉以呈現靜脈血管與背景之對比,再由中值濾波器清除大部份散佈於圖形雜訊。為避免外形輪廓影響後續靜脈血管圖形抽取,先將前後景二值化處理,以比對方式和3×3遮罩去除輪廓邊緣,接著用形態學斷開方法去除微小雜訊,並由連通標籤抽取靜脈血管。參考形態學閉合方法使抽出的靜脈血管平滑化,後以細線化方式使靜脈血管呈現骨架化,接著用修剪方法去除骨架邊緣細小突出部份。為使掌背靜脈血管分類與辨識,設計正Y與倒Y圖形找出骨架影像中三角交叉點,利用二元樹方法與總數相同Y圖形資料夾,比較各別計數正Y與倒Y圖形數目分類,此種方法達成穩定正確分類與初步辨識目地,為減少日後靜脈血管辨識錯誤,依據正Y與倒Y圖形坐標,算出正Y與倒Y圖形各別與平均坐標,最後再由各別、整體平均坐標與正Y與倒Y坐標,算出各別與整體相互距離之分數做比對,以達到提昇掌背靜脈血管辨識之目標。在我們的實驗中,採用25位於各個年齡登入者,其中5位女生、20位男生。在不同時間,每人左手與右手各取樣10張影像,比對抽取掌背靜脈圖形與原始影像之結果為正確抽取率為93.35%,其中右手正確抽取率為94.8%,左手正確抽取率為91.9%。
For the characteristic of uniqueness on the human, number of biometric technologies have been developed and are used in personal identification like as iris, fingerprint, face detection and vein etc. The back-palm vein pattern is unique to individuality which pattern does not change over time apart from size. This feature makes it suitable for one-to-many matching, for which fingerprint and face recognition may not be robust. Among all the biometric techniques, back-palm vein recognition is a topic worthy to receive further investigation since this technology overcomes aversion to fingerprinting and related privacy concerns, which its traditional association to criminal activity is non-existent. Back-palm vein recognition works by identifying the subcutaneous vein patterns in an individual's back-palm and is difficult to replicate because they lie under the skin surface. When a user's back-palm is placed under the color SONY camera (SSC-E473) in our system, a near-infrared (NIR) light maps the location of the veins. The red blood cells present in the veins absorb the rays and show up on the map as black rugged lines, whereas the remaining back-palm structure shows up as light cyan. After the vein classification, it is compared with previously stored patterns and a match is made. In this thesis, we present a new method for vein extraction and classification of the back-palm vein. High-boost filter enhanced original blur image with NIR radiation CCD (charged-couple-device) camera. For enhancing vein data, a novel thresholding is designed to separate foreground data and background information. A 11×11 mask from Laplacian of Gaussian concept contrast between vein and surrounding areas. For avoiding the contour of the back palm affect the vein extraction, median filter is used priority to remove most of the noise. We use AND gate to contrast the result between binarization and median image. Thresholding of 3×3 mask is utilized to clear the noise around contour of the back-palm image. Sequentially, morphological opening is used to eliminate little noise before vein extraction with connected component labeling. After above movement, morphological closing, thinning and pruning present skeleton and clear protrude on the skeleton image. Finally, both positive and opposite Y types are designed to find out the tri-intersection in pruning image, then utilize binary tree and counting independent numbers about positive and opposite Y types for same number in classified file to classification. This method can achieve stable classification and original recognizable purpose. For reducing the recognizable error in the future, we design average coordinate method to set up three datum points with positive Y, opposite Y, total Y and calculate independent score with mutual Euclidean distance between average coordinate and Y points. In our experiment, 25 enrolled user (five girls and twenty boys) of different gender, each has both 10 images for left and right hand at different intervals. The performance of the accurate extraction ratio is 93.35% (94.8% on right hand, 91.9% on left hand) between back-palm vein pattern and original images.
Chinese abstract ---------------------------------------------------- iii
English abstract ---------------------------------------------------- iv
Gratitude ---------------------------------------------------- vi
Content ---------------------------------------------------- vii
List of figure ---------------------------------------------------- viii
List of table ---------------------------------------------------- xi
Chapter 1、 Preface------------------------------------------- 1
1.1 Introduction------------------------------------- 1
1.2 Motivate----------------------------------------- 3
1.3 Related literatures research------------------------- 4
1.4 Structure of the thesis------------------------------- 5
Chapter 2、 Pre-processing---------------------------------- 9
2.1 High-boost filter-------------------------------- 16
2.2 ROI (region of interest) Segmentation------------ 23
2.3 Vein pattern extraction------------------------------ 30
2.4 Median filter ----------------------------------------- 34
Chapter 3、 Vein extraction--------------------------------------- 38
3.1 Binarization and ROI---------------------------------- 40
3.2 Thresholding (3×3 mask) and opening------------ 47
3.3 Connected component labeling--------------------------- 54
Chapter 4 Post-processing--------------------------------- 58
4.1 Closing------------------------------------------------- 59
4.2 Thinning----------------------------------------------------- 62
4.3 Pruning------------------------------------------------------- 68
Chapter 5 Classification----------------------------------------- 73
Chapter 6 Experiment result and discussion------------------ 86
Chapter 7 Conclusion and Future work------------------------ 98
Reference ---------------------------------------------------- 99
Reference
[1] J. M. Cross and C. L. Smith, “Thermographic imaging of the subcutaneous vascular network of the back of the hand for biometric identification,” Proceedings of 29th International Carnahan Conference on Security Technology, Institute of Electrical and Electronics Engineers, 20-35, 1995.
[2] C. L. Lin, and K. C. Fan, “Biometric verification using thermal images of palm -dorsa vein patterns,” IEEE Transitions on Circuit and Systems for Video Technology, vol.14, no. 2, pp. 199-213, Feb. 2004.
[3] S. K. Im, H. M. Park, Y. W. Kim, S. C. Han, S. W. Kim and C. H Kang, “An biometric identification system by extracting hand vein patterns,” Journal of the Korean Physical Society, vol. 38, no. 3, pp. 268-272, March 2001.
[4] S. K. Im, H. S. Choi and S.W. Kim, “A direction-based vascular pattern extraction algorithm for hand vascular pattern verification,” ETRI journal, vol. 25, pp. 101-108, April 2003.
[5] A. M. Badawi “Hand vein biometric verification prototype: A testing performance and patterns similarity,” IPCV 06 June 2006.
[6] K. J. Wang, Q. C. Guo, D. Y. Zhuang, Z. Y. Li and H. G. Chu, “The study of hand vein image processing method,” Proceeding s of the 6th World Congress on Intelligent Control and Automation, vol. 2, pp. 10197-10201, June 2006.
[7] K. J. Wang, Y. Zhang, Z. Yuan and D. Y. Zhuang, “Hand vein recognition based on multi supplemental features of multi-classifier fusion decision,” Proceedings of the 2006 IEEE International Conference on Mechatronics and Automation June 25-28, Luoyang, Chine.
[8] S. Malki, Y. Fuqiang and L. Spaanenburg, “Vein feature extraction using DT-CNNs,” 2006 10th International Workshop on Cellular Neutral Networks and Their Applications, Istanbul, Turkey, Aug. 2006.
[9] Z. L. Wang, B. C. Zhang, W. P. Chen and Y. S. Gao, “A performance evaluation of shape and texture based methods for vein recognition,” Image and Signal Processing 2008 CISP08, vol. 2 27-30, pp 659-661, May 2008.
[10] W. X. Kang, and F. Q. Deng, “Vein recognition method base on angle matrix of feature point,” Control Conference 2008, CCC 2008, 27th Chinese, pp 569-572 July 2008.
[11] S. A. Coleman, B. W. Scotney and M. C. Herron, “A systematic design procedure for scalable near-circular Laplacian of Gaussian operators,” Proceedings of the 17th international Conference on Pattern Recognition (ICPR’04), vol. 1, pp. 700-703, Aug. 2004.
[12] S. D. Joan. C. M. Liu, C. C. Chang and Z. Chen, “A new algorithm and its VLSI architecture design for connected component labeling,” IEEE International Symposium on Circuits and Systems, vol.2, pp. 565-568, May 1994.
[13] W. N. Leung, C. M. Ng and P.C. Yu, “Contour following IParallel thinning for simple binary images,” IEEE Systems, Man, and Cybernetics, vol, 3, pp. 1650-1655, Oct. 2000.
[14] B. Fei and J. Liu, “Binary tree of SVM: a new fast multiclass training and classification algorithm,” IEEE Transactions on Neural Networks, vol. 17, no. 3, May 2006.
[15] G. Sun, Z. Wang and M. Wang, “A new multi-classification method based on binary tree support vector machine,” Innovative Computing Information and Control, pp. 74-77, June 2008.
[16] S. Aso, S. Yokota and H. Hashimoto, “Control and stability for robotic swarm based on center of gravity of local swarm,” IEEE International Symposium on Industrial Electronics, pp. 1341-1346, July 2008.
[17] B. Beckman, J. Pieper, D. Mackay, M. Trentini and D. Erickson, “Two dimensional dynamic stability for reconfigurable robots designed to traverse rough terrain,” IEEE Intelligent Robots and Systems, pp. 1341-1346, Sept. 2008.
[18] T. Y. Hsieh, K. J. Lee and M. A. Breuer, “An error rate based test methodology to support error-tolerance,” IEEE Transactions on Reliability, vol. 57, no. 1, pp. 204-214, March 2008.
QRCODE
 
 
 
 
 
                                                                                                                                                                                                                                                                                                                                                                                                               
第一頁 上一頁 下一頁 最後一頁 top