基于Harris角点检测和聚类算法的掌纹图像ROI提取方法  

ROI extraction method of palmprint based on the Harris point detection and clustering algorithm

作　　者：陈典典 程培培 马军山[1] CHEN Diandian;CHENG Peipei;MA Junshan(School of Optical-Eleclrical and Compuler Engineering,University of Shanghai for Science and Technology,Shanghai 200093,China)

机构地区：[1]上海理工大学光电信息与计算机工程学院,上海200093

出　　处：《光学仪器》2018年第5期27-34,共8页Optical Instruments

摘　　要：为了实现从掌纹图像中提取出稳定的感兴趣的区域(ROI)图像的目标,设计了一种新的掌纹ROI提取方法。使用Harris角点检测算法对二值化后的掌纹图像进行角点提取,收集相应区域的角点,利用聚类算法得到角点簇的中心点坐标,从中寻找关键点建立坐标系,提取ROI。分别利用该方法在不同掌纹数据库中进行了ROI提取实验。实验结果表明,该方法对不同的掌纹数据库图像均能保持很好的提取效果,成功率均达到了99%以上。In order to achieve the goal of extracting a stable region of interest （ROD image from palmprint images, a new method for extracting palmprint ROI was propose＆ Harris point detection algorithm is used to extract the angular point d the palmprint image after binarization and collect the angular points in the corresponding region. Clustering algorihm is used to obtain the central coordinates of the angular cluster. The key points are found in the cluster center to establish coordinate system to extract the ROI. The ROI extraction experiment was carried out in the different palmprint database. The results show that the algorithm can effectively extract the ROI in the different palmprint database, with the accuracy rate of more than 99%.

关 键 词：掌纹识别 HARRIS算法 聚类算法 感兴趣的区域 

分 类 号：TP391.4[自动化与计算机技术—计算机应用技术]