基于误差扩散和均值漂移的DSA特征点选择

Error diffusion and mean shift based feature points selection algorithm of DSA images

数字减影血管造影(DSA)技术在血管疾病的诊断和治疗中起着重要的作用。在DSA图像配准算法中,控制点的精度、数量以及它们在图像上分布的情况在很大程度上决定了几何校正和配准的精度。控制点通常采用规则的网格点,然而更加有效的方法是利用图像本身的特征提取控制点。提出了一种基于误差扩散(Error Diffusion)和均值漂移(Mean Shift)的DSA特征点选择算法。该算法引入误差扩散是为了利用图像本身的特征自适应地选择控制点。在边缘和纹理复杂区域取较多控制点(网格密集),相反在平坦区域取较少控制点。引入均值漂移的目的是为了减少控制点的数量。实验结果表明,该算法可以自适应地把控制点安排到血管等关键图像特征上,并可以根据需要对控制点的数量进行优化。

In DSA image registration algorithm,the precision of the control points as well as their number and the distribution in image determine the accuracy of geometric correction and registration.Control points usually adopt the grid points; however,a more effective method is to extract control points adaptively according to the image feature.A control point＇s selection algorithm of DSA images is proposed based on error diffusion and means shift algorithms.The paper introduces error diffusion algorithm to take advantage of the characteristics of the image to adaptively select control points.In the edge and texture complex area,more control points（grid intensive）will be selected.On the contrary in the flat area less control points will be selected.Introduction of the mean shift is to reduce the number of control points.Experimental results show that proposed algorithm can adaptively put the control points to blood vessels and other key image characteristics,and can optimize the number of control points according to practical needs.