Human Hand Vein Image Acquisition And Matching Of Blood Vessels

The back of the body has a very rich vein, each person’s hand vein are not the same,it is difficult to be changed through surgery and other means, with high security anduniqueness, he back of the hand vein recognition vein recognition areas of the humanbody can be regarded as the most fiery development. Vein recognition technology is oftenlinked to the infrared, since the use of human blood is capable of absorbing infrared lightof the extracted vein image characteristics, when a large amount of hemoglobin absorbsinfrared light, the vein will be highlighted, which just need a better response to infraredlight image sensor and it can extract the vein image. So to identify vein must design ahigh accuracy vein recognition instruments first, this instrument should have betterinfrared light source so as to adequately highlight the veins, there must be a betterresponse to infrared light image sensor, extracting a vein image in order to prepare more.Human hand vein recognition, the infrared light shines on a uniform body back first, andextract the vein image image sensor later, the computer and the extracted image somepreprocessing then, such as filtering, binarization, thinning repair or so, to mention onlyrepresentative characteristic value can be taken out of this image, and the image and thefeature values are stored in the computer. When the time required to identify matching,the real image and the extracted feature value of the vein, the image and the featurevalue and the previously stored comparison or matching similarity, similarity or contrastaccording to the image matching, the situation determines whether this person or theidentity of it is the same person.Vein recognition algorithms described in this article includes three parts: hand veinimage acquisition; hand vein image preprocessing and hand vein pattern recognition.Some of the major sharpening the image preprocessing Wiener filtering, image, medianfiltering, image skeleton extraction and refinement after refinement or repair and othercomponents. And image recognition algorithm has a good translation, rotation, stretchinginvariance.