DBSCAN（Density-Based Spatial Clustering of Applications with Noise，具有噪声的基于密度的聚类方法）是一种基于密度的空间聚类算法。该算法将具有足够密度的区域划分为簇，并在具有噪声的空间数据库中发现任意形状的簇，它将簇定义为密度相连的点的最大集合。



该算法利用基于密度的聚类的概念，即要求聚类空间中的一定区域内所包含对象（点或其他空间对象）的数目不小于某一给定阈值。DBSCAN算法的显著优点是聚类速度快且能够有效处理噪声点和发现任意形状的空间聚类。但是由于它直接对整个数据库进行操作且进行聚类时使用了一个全局性的表征密度的参数，因此也具有两个比较明显的弱点：



（1）当数据量增大时，要求较大的内存支持I/O消耗也很大；



（2）当空间聚类的密度不均匀、聚类间距差相差很大时，聚类质量较差。-DBSCAN (Density-Based Spatial Clustering of Applications with Noise, with noise density-based clustering method) is the density based spatial clustering algorithm. The algorithm will have sufficient density region is divided into clusters, and discover clusters of arbitrary shape in spatial s with noise, the maximum density is defined as a collection it clusters connected points. The algorithm uses the concept of density-based clustering, which called for the number of clusters in space within a certain region containing the object (point or other space objects) is not less than a given threshold. DBSCAN significant advantages of clustering algorithm is fast and effective handling noises and found that spatial clustering of arbitrary shape. However, because it operates directly on the entire　and clustering when using a global parameter characterization density, it also has two obvious weaknesses: (1) When the amount of data increases, requiring larger memory supports I/O consumption i