| Summary: | 碩士 === 國立臺灣科技大學 === 資訊工程系 === 98 === The technology has become indispensable to modern science. Technology products are always full of around us. More and more products of technology have come out .Such as PDA, Smart phone, notebook also debit cards, electronic wallets and Internet banking... brings convenient to our life . On the other hand it also brings security concerns. In order to solve these problems, the recognition methods related to biometric one after another.
The vein recognition is the most popular in recent years. It uses the infrared LED irradiate fingers or palms to visualize the presentation of the vein biological characteristics .There are palm vein, finger vein, dorsal vein and wrist and some other. But in general, palm vein and the finger vein is the mainstream.
This paper will use fewer the characteristics of fingers to do identify work. Firstly, we use Gaussian smoothing to do image enhancement. Secondly, we do the binaries according to the threshold value, then we thinning the image. In order to extract feature points (bifurcate points). Then calculate the distance between these feature points as a basis for matching. This method can resist a slight rotation and translation. Finally, we use the similarity of the vein shape to do the second matching. Make up the insufficiency that only uses the distance matching method. The experimental results confirm the method has the higher Identification rate, does not have the invasion and low cost than other papers.
|
|---|
