A New Features Extracted for Recognizing a Hand Geometry Using BPNN
Firas M. Al-Fiky, Zainab Salih Ageed · 2014
The biometrics plays a vital role in person recognition, in this method a 66 features has been calculated and determined for the right hand, the method has two main phases, the first contain the data collection and preproecessing, while the second contain the training and testing of an artificial neural network. The proposed method suggested the BPNN for training with one input layer, one hidden layer, and one output layer. The recognition rate RR for the neural network after testing and using proposed features was clearly shows an enhancement in results through the comparison between the previous works and the proposed method. —————————— a —————————— 1 I NTRODUCTION he human identification system using image prepro- cessing has been used in many application and fields of security. Finger print, face and iris recognition systems are developed and implemented in the banking access, military, customs monitoring system, police recording system for crimi- nal and security system of special objects(1). As a method of biometric recognition, hand-shape recognition has always been recognized as an effective means of personal identification. At present, the international biometric product based on hand- shape shares 25% of the total sale of biometric recognition product, which is only less than fingerprint recognition. As the hand shape images are easily acquired, meanwhile, comparing with other biometric feature's collection, the means of which is easily accepted. It has a greater potential and prospect for de- veloping and researching (2). A human body posses several physiological characteris- tics that can serve as biometric features. Also a human being develops several unique behavioral traits which can also serve as biometric features. The various physiological characteristics that are generally used are face, iris, fingerprints, palm-prints, hand geometry and voice. Face is the biometric primarily used by human beings to recognize each other. This makes it an ob- vious choice for biometric. The difficulty however is in the fact that the biometric system has to rival the complexity of the hu- man brain. Fingerprints have also been used for quite some time now and have established its value as a biometric. The challenge now is to develop more advanced systems which can process partial prints and speed up the matching process. The behavioral characteristics include signature, handwriting analy- sis, voice, keystroke pattern and gait. Signatures and handwrit- ing have been used extensively as biometrics. However; they have been used only as an offline biometrics, i.e. no data is col- lected during the process of signing or writing. Automated bi- ometric systems vastly improve the accuracy of the operation by including data obtained during these processes (3).