Abstract: This paper introduces a method for seabed sediment classification.At first, texture features are extracted by gray level co-occurrence matrix based on backscatter image of multibeam data, then seabed sediment classification is carried out with supporting vector machine(SVM) using the backscatter combined with backscatter image.The result of classification by this method is better than the traditional method based on point data according to our practice.It will certainly make the regional marine geological survey more efficient and reliable.