On Recognition Algorithm Of Thenar Palmprint And Identification System Design

As can be able to diagnosis disease quickly and painlessly, accuratly and early, palm medical diagnose has been generally recognized. According to the different texture, color and other information of palm, Chinese medicine experts can give diagnosis, this allows patients to advance the prevention or the right remedy. The clinical practice of medical experts found that, thenar palmprint roughness has some relationship with asthma allergic diseases. According to the probability or severity of the illness, or accordance with the roughness of the thenar palmprint, the thenar palmprint can be divided into four grades.Palm medical diagnose are highly influenced by the experience of the doctor, which is subjectively. Therefore, the purpose of this article is design the algorithm and system which would automatically divide the thenar palmprint into four classifications. Then make expert knowledge be standardization, unionization and objectify.This article mainly introduces the thenar palmprint preprocessing and classification algorithm and the thenar palmprint identification system design, so as to achieve the purpose of classification.First, considering the optical principle, we choose the ring light, industrial cameras and fixed focus lens to design and produce a palmprint acquisition system, it can reduce the interference outside and get stable palmprint image; Then use the positioning technology ROI to get the region of thenar palmprint; Introduce the intermediaries algorithm, and do image filtering with intermediary filtering method to the acquired thenar palmprint, and has achieved good results. Next, extract eight GLCM characteristics of thenar palmprint and analysis the advantages; Finally, choose hierarchical support vector machines based on binary tree, and make the extracted feature vectors as input vector to train classifiers in order to do multi-classification to thenar palmprint.Based on the algorithm above, we use VC++as development tools, MFC as framework, have developed the quantitative identification system of thenar palmprint, it can help doctors make a diagnosis.