Image preprocessing algorithm for palm print based on Matlab

Abstract

In our modern world, an increasingly number of business activities need to use

identification system computerized. E-Commerce applications such as e-banking,

building access demand fast, real time and accurate personal identification.

Traditional token based method such as a physical key, ID card or knowledge based

method, such as a password is time-consuming and inefficient. They can be lost,

stolen, or forgotten. To overcome the shortcomings of the traditional methods and

meet today’s security requirements; Biometrics-based personal identification is

emerging as a powerful means for recognizing a person’s identity. It plays an

important role in today’s society.

The purpose of biometric based palm print identification is to examine human’s palm

print by achieving palm print lines from its region of interest. Although the system of

biometric based palm print identification becomes well-established, the qualities of

palm print images are quite low due to incomplete image pre-processing. The

enhancements for these images are necessary before extracting palm print regions of

interest.

The existing image processing methods provided for biometric based palm print

identification is watershed and noise-deduction function. Before these, some image

enhancement functions need to be applied in order to have better observing to pictures.

These functions include Laplacian Sharpening, averaging filter, and power law

transformation etc. The purpose of these functions is to have high-contrast images

with clearer edges and palm print lines. These important processing will make some

improvement in the following identifications.