Research On Thenar Palmprint Quantitative Recognition Algorithm And Aided Diagnosis System Design

Palm diagnosis is one of widely used diagnostic methods in Tradition Chinese Medicine(TCM), according to years of clinical experience medical experts found that Asthma or Allergic diseases patients have rough thenar palmprint. After medical experts'long-term clinical study, the thenar palmprint is divided into four levels based on their roughness:Ⅰ,Ⅱwas negative, andⅢ,Ⅳwas positive, using computer technology to research thenar palmprint is a field of creative work. Based upon the foundation of thenar palmprint classification, this dissertation mainly studies the approach to thenar palmprint quantitative and classification, the research includes the followings:Image preprocessing of thenar palmprint, using the method of positioning axes to locate and separate the thenar palmprint, median filter is used to remove the noise of the thenar palmprint. The combination of histogram and high-frequncy filtering is used to enhanced the image texture of the thenar palmprint.Using gray level coocurrence matrix to extract texture features of the thenar palmprint, enhanced thenar palmprint, we extract four directions gray level co-ocurrence matrixs. Based on the gray level co-ocurrence, we extract the UNI,CON,COR,ENT,VAR,IDM,SA,DE eight characters for classification.Discuss the support vector machine method , which is applied to the thenar palmprint's GLCM classification, first, introduce the basic theory of SVM, then through the research and analysis the accuracy of classification with different characteristics of features, and ultimately determine the final characteristic of features which has the maximum classification accuracy of the thenar palmprint.The design of thenar palmprint quantitative and recognition system is discussed and analyzed, and based on this system, the thenar palmprint database is established. Through combine the medical knowledge, analysis the application of thenar palmprint in Aided diagnosis system. At last simply analysis the case-based reasoning expert system, which is used in thenar palmprint aided diagnosis system.