	LOGIN
	>> Home
	>> Topics
	>> Students
	>> Partners
	>> Statistics

Information for topics

Topic Id:

ID topic:	406
Partner Email:	ktutuncu@selcuk.edu.tr
Project Title:	AUTOMATIC FINGERPRINT RECOGNITION BY USING SOFT COMPUTING TECHNIQUES
Abstract:	This study is realized to recognize the people by a new generated method with the assistance of soft computing methods. 1000 fingerprint images that belong to 10 people are used from the NIST-4 database. The quality of the images of the fingerprints that are used in the process of the recognition affects the performance of recognition. These fingerprint images include interferences due to the external effects. For this reason the images of fingerprints are processed initially for image enhancement. At the second stage of the study, the feature vector called as Fingercode is obtained by means of a generated method. The obtained feature vectors are trained by artificial neural networks and the fingerprint recognition is performed. For this purpose, from totally 1000 fingerprints, 700 fingerprints are used in the process of ANN training , 150 fingerprints are used in the process of validation and 150 fingerprints are used in the process of testing. In the last period of the study, the size of the feature vector is taken into account. Thus, the reduction of the size is needed. Because of this reason the topics of reduction of feature, data and size are investigated. Thus, by eliminating the features with interference, selection of the features are made. The selected features are retrained and recognition process is performed by using ANN.
Advisor:	Prof. Dr. Novruz ALLAHVERDI
Link:
Degree:	Master
Keywords:
Computer Software
Algorithms & problem solving
Artificial intelligence & Neural networks
Computer vision
Image processing
Information systems