engleski

Biometric identification systems in function of transport safety by applying CVQ compression method

Generally, fingerprint, face and iris recognition occupy a large place in biometric identification, and they are among the most popular biometric methods. Popularity of biometric identification is growing in all areas of transport, such as maritime, air and road transport. All these biometric methods operate following the same principle: creation of unique digital templates that are extracted from data such as iris, face and fmgerprint images. Specially designed algorithms are used for creation of unique digital templates. Future tendencies and politics suggest usage of real fingerprint, face and iris images instead of templates. Biometric identification systems in transport often require storage and transmission of images, and therefore images should be compressed in order to save resources. This paper explores and analyses effects of Classified Vector Quantization (CVQ) compression method against fingerprint, iris and face images. Grayscale images from public databases are used in the experiment. Results show that CVQ is not equally effective against all three types of images at the same bit rate level because some types have different image properties in terms of image edges and abrupt transitions between pixel values. Advantages over ordinary Vector Quantization (VQ) are pointed as well.

biometrics ; identification ; CVQ ; compression ; transport ; safety

nije evidentirano