CAJ | 학술논문

首先将掌纹图像经过定位分割获得感兴趣的大鱼际区域，再经中值滤波去噪，采用高频强调滤波和直方图均衡化相结合的方法增强图像，得到纹理清晰的大鱼际掌纹图像。然后采用灰度共生矩阵方法提取利于二分类的图像特征。最后基于支持向量机的方法实现大鱼际掌纹阴阳两类的识别。仿真实验结果表明，所提出的图像处理与识别算法对大鱼际掌纹阴性与阳性的分类具有较高的效率和准确率。
수선장장문도상경과정위분할획득감흥취적대어제구역，재경중치려파거조，채용고빈강조려파화직방도균형화상결합적방법증강도상，득도문리청석적대어제장문도상。연후채용회도공생구진방법제취리우이분류적도상특정。최후기우지지향량궤적방법실현대어제장문음양량류적식별。방진실험결과표명，소제출적도상처리여식별산법대대어제장문음성여양성적분류구유교고적효솔화준학솔。
At the beginning palm print is located and segmented to get thenar region that arouses interest.Then after median value noise filtration,images are enhanced by high-frequency emphasis filtration and histogram equalization combination method to obtain clear texture thenar palm print image.Next grey level co-occurrence matrix method is used to extract image features bias to binary classification.At last thenar palm print both Yin and Yang classes are recognized by SVMmethod.Simulation experiment result illustrates that the proposed image processing and recognition algorithm acquires higher efficiency and accuracy with thenar palm print Yin and Yang classifications.