The paper considers the problem of clustering pixels of a color raster image. The task is to compare the effectiveness of three different clustering methods: k-means, DBSCAN, agglomerative clustering.