Biometric palm scanning methods for identification using new approaches

Biometrics is a unique and measurable feature for identification. Biometric systems are divided into two categories: identification and verification. The first category's aim is recognizing the identification of the individual among the others in the database. In such systems, to the question "who am I?" will answer. In the second, the person offers his identity and the aim of the system is the confirming this identification and in the systems, to the question "do what I say I am?" will answer. It is clear that any biometric identification system is not the absolutely best solution for identification. International biometric Group does an interesting comparison done between different systems based on four parametric, distinctiveness, and cost of identification system, time and spending effort by the user in the identification and the rate of user's comfort during identification . An ideal biometric system is a system in which contains all four enumerated parameters in the farthest place according to the center of the diagram . In another study has gained the highest percentage compatibility by taking six biometric techniques (face, fingerprint, hand geometry, voice, eyes and signature) with machine-readable travel documents (MRTD), facial features. In this study, parameters such as registration, refreshment, hardware requirement and public acceptance is considered. Palm recognition system is a biometric system using intelligent automatic methods to identify or verify the identification of a person based on physiological characteristics. In the past two decades, the matter of identifying of palms is the extensive research field of machine's vision and pattern recognition. One of the extensive applications for recognition of palm lines is the field of security and verification. In controlling the high population areas, such as airports, railway stations, subway and ect, this method is more effective than other methods of surveillance. So several photographs have taken from the lines in the palms of people and the device must be able to identify these people at different times, in different orientations of light and ect. Current methods of identifying palms uses of four types of palm's features: texture, lines, appearance and orientation. According to the extracted features are divided into five categories; Texture based methods, Line based methods, Appearance based methods, Orientation based methods, Multi feature based method. Combination methods in the features level are divided into four categories: series combination, parallel combination, weight combination, and core based combination. In order to identification by a person's palm lines, the lines must be properly extracted. One-way to define these lines are using different methods of edge detection. Sobel edge detection and Morphological operation is used in. In this thesis, for the complete extraction of useful features from an appearance, a simple but powerful method an uniform local binary pattern to identification of palms are expressed because this method can pull out all the useful information of an appearance.