ارسل ملاحظاتك

ارسل ملاحظاتك لنا







الرئيسية > New Chaff Point Based Fuzzy Va...>الوصف

New Chaff Point Based Fuzzy Vault for Multimodal Biometric Cryptosystem Using Particle Swarm Optimization

المصدر: مجلة جامعة الملك سعود - علوم الحاسب والمعلومات
الناشر: جامعة الملك سعود
المؤلف الرئيسي: Amirthalingam, Gandhimathi (Author)
مؤلفين آخرين: Radhamani, G. (Co-Author)
المجلد/العدد: مج28, ع4
محكمة: نعم
الدولة: السعودية
التاريخ الميلادي: 2016
الصفحات: 381 - 394
DOI: 10.33948/0584-028-004-002
ISSN: 1319-1578
رقم MD: 973970
نوع المحتوى: بحوث ومقالات
اللغة: الإنجليزية
قواعد المعلومات: science
مواضيع:
علوم الحاسوب | التشفير | الخوارزميات
كلمات المؤلف المفتاحية:
Modified Region Growing Method | Local Gabor XOR Pattern | Chaff Points | Particle Swarm Optimization Algorithm | Fuzzy Vault
رابط المحتوى:
PDF (صورة)     
صورة الغلاف QR قانون
حفظ في:
المستخلص: An effective fusion method for combining information from single modality system requires Multimodal biometric crypto system. Fuzzy vault has been widely used for providing security, but the disadvantage is that the biometric data are easily visible and chaff points generated randomly can be easily found, so that there is a chance for the data to be hacked by the attackers. In order to improve the security by hiding the secret key within the biometric data, a new chaff point based fuzzy vault is proposed. For the generation of the secret key in the fuzzy vault, grouped feature vectors are generated by combining the extracted shape and texture feature vectors with the new chaff point feature vectors. With the help of the locations of the extracted feature vector points, x and y co-ordinate chaff matrixes are generated. New chaff points can be made, by picking best locations from the feature vectors. The optimal locations are found out by using particle swarm optimization (PSO) algorithm. In PSO, extracted feature locations are considered particles and from these locations, best location for generating the chaff feature point is selected based on the fitness value. The experimentation of the proposed work is done on Yale face and IIT Delhi ear databases and its performance are evaluated using the measures such as Jaccard coefficient (JC), Genuine Acceptance Rate (GAR), False Matching Rate (FMR), Dice Coefficient (DC) and False Non Matching Rate (FNMR). The results of the implementation give better recognition of person by facilitating 90% recognition result.
ISSN: 1319-1578

عناصر مشابهة


© 2024 المنظومة. جميع الحقوق محفوظة.