This paper focuses on the application of object identification and recognition techniques in the service of homeland security. Online face images are captured, preprocessed and applied to a feature extraction unit. Then, a linear prediction code (LPC) process is executed. The linear prediction coefficients represent the principle features to an image classifier where these features are compared with those of preprocessed ones. The minimum Euclidian distance recognizer is used. In this work a small data base of the faces of different people is built up. The assumed attributes are personal data and the features of the facial image. The recognition accuracy is enhanced and is of the order of 80%. The recognition accuracy can be improved using an adaptive recursive LPC.
This paper focuses on the application of object identification and recognition techniques in the service of homeland security. Online face images are captured, preprocessed and applied to a feature extraction unit. Then, a linear prediction code (LPC) process is executed. The linear prediction co...
مادة فرعية
Omani Journal of Applied Sciences (OJAS)