基于DBSCAN的图像分割算法研究

图像分割能将一幅数字图像分成多个的不同区域,是计算机视觉的主要研究领域之一。文中在系统性研究DBSCAN算法的基础上，提出了一种改进型DBSCAN图像分割方法。该方法首先计算图像中每个像素点的局部密度，然后通过寻找局部密度峰值点来确定核心点，同时将邻域内的像素点加入同一簇中，来处理不同密度区域和噪声点的影响。实验结果表明，该算法对参数敏感度较低，能有效处理不同密度区域和噪声点，相比标准DBSCAN图像分割方法，其在聚类正确率、精度和效率等方面的表现更优秀。

Research on Image Segmentation Algorithm Based on DBSCAN

Image segmentation can divide a digital image into multiple different regions, which is one of the main research fields of computer vision. Based on the systematic study of the DBSCAN algorithm, this paper proposes an improved DBSCAN image segmentation method. The method first calculates the local density of each pixel in the image, and then determines the core point by looking for the local density peak point, and adds the pixels in the neighborhood to the same cluster to deal with the influence of different density regions and noise points. The experimental results show that the algorithm has low sensitivity to parameters and can effectively deal with different density regions and noise points. Compared with the standard DBSCAN image segmentation method, its performance is better in terms of clustering accuracy, accuracy and efficiency.