基於形態學之手掌身份辨識

Hand Recognition Based on Morphology

王蔚強(Wei-Chiang Wang)

指導教授 ： 陳文雄

共同指導教授 ： 謝志明(Chih-Ming Hsieh)

國立暨南國際大學/科技學院/電機工程學系/碩士(2008年)

https://doi.org/10.6837/NCNU.2008.00118

摘要

有鑑於現有的研究，多半是分別依掌紋與掌形不同型式的特徵進行擷取，或者結合掌紋與掌形的特徵以進行辨識。本論文提出一套生物辨識的系統，此系統基於形態學的方法，融合掌紋與掌形以做為身份辨識的基準。本系統利用了手掌的整個正面的影像，並且省去了大小正規化的程序。 本系統可分為四個模組，分別是影像擷取、前處理、特徵萃取與辨識模組。前處理的目的是依照個人中指的長度定義手掌區域。特徵萃取模組則利用影像形態學與Voronoi diagram 的概念，將手掌正面的影像依照掌形，切割成若干不規則形狀的小區塊，接著以小區塊中灰階之統計特性做為特徵值。接著本論文提出一套兩階段辨識模組，分別以掌形與掌紋為特徵。第一階段利用手掌之帶狀區域的數目進行粗糙辨識，第二階段利用區塊中的灰階之平均或變異量當作為特徵碼以進行細部辨識。本論文以實驗室自建的手掌影像資料庫做實驗，獲得令人滿意的結果。錯誤接受率（falseacceptance rate; FAR）為0.0035%；錯誤拒絕率（false rejection rate; FRR）為5.7692%。 最後以F-ratio 檢視手掌中那些區域的統計特徵較具鑑別度。

關鍵字

生物辨識掌紋 ； 掌形 ； 影像形態學 ； 兩階段辨識

並列摘要

In the past time, most of palmprint and hand geometry recognition methods were proposed individually. Besides, the fusion of palmprint and hand geometry was also proposed. This thesis presents a novel biometric recognition system fusing the palmprint and hand-shape of a human hand based on morphology. And our method utilizes the image of the front of the whole palm. In addition, we do not need a process of size normalization in our system.
The proposed system consists mainly of four modules: image acquisition, pre-processing, feature extraction, and recognition modules. The pre-processing module utilizes the length of middle finger to define the region of palm. The feature extraction module employs the image morphology and concept of Voronoi diagram to divide the palm image into a number of irregular blocks in accordance with the hand geometry. Furthermore, some statistical measurements of the gray level for each block are extracted as feature values. Then, we present a two-stage recognition module based on the features from hand-shape and palmprint of a hand. First stage utilizes the number of stripe regions of a palm to be the feature value. Second stage encodes the means or variances of the pixel gray levels of the blocks as feature values. The experimental results show that the proposed
system has an encouraging performance on our own hand database. The false acceptance rate (FAR) and false rejection rate (FRR) down to 0.0035% and 5.7692%, respectively. Finally we look over with F-ratio in which areas with statistical characteristics in front of the whole palm have discriminative characteristics.

並列關鍵字

Biometrics ； palmprint ； hand geometry ； morphology ； Voronoi diagram ； two-stage recognition module ； F-ratio

參考文獻

[1] A. Kong, D. Zhang and M. Kamel, “Palmprint identification using feature-level fusion,” Pattern Recognition, vol. 39, no. 3, pp. 478-487, 2006.

[2] A. Kumar, D. C. M. Wong, H. C. Shen and A. K. Jain, “Personal verification using palmprint and hand geometry biometrics,” Proc. of 4th International Conference on Audio- and Video-based Biometric Person Authentication, 2003.

[3] A. Kumar and D. Zhang, “Combing fingerprint, palmprint and hand-shape for user authentication,” 18th International Conference on Pattern Recognition, vol. 4, pp. 549-552, 2006a.

[4] A. Kumar and D. Zhang, “Personal recognition using hand shape and texture,” IEEE Transaction on Image Procdssing, vol. 15, no. 8, pp. 2454-2461, 2006b.

[5] A. Kumar and H. C. Shen, “Recognition of palmprints using wavelet-based features,” Proc. of International Conference on Systems and Cybernetics, 2002.