Efficient Classification Method of Marine GIS Remote Sensing Database Under Dynamic Complex Network

Abstract

Zheng, Y.; Zhang, L.; Yang, B.; Zhang, G.; Liu, T., and Liu, S., 2020. Efficient classification method of marine GIS remote sensing database under dynamic complex network. In: Yang, D.F. and Wang, H. (eds.), Recent Advances in Marine Geology and Environmental Oceanography. Journal of Coastal Research, Special Issue No. 108, pp. 63–67. Coconut Creek (Florida), ISSN 0749-0208.In the traditional marine GIS database of remote sensing classification process, the classification accuracy is low, and the classification effect is not ideal. Therefore, an effective classification method based on relevant evidence for remote sensing database of marine GIS is proposed. The n-gram algorithm is used to calculate the GIS remote sensing data records, and the n-gram values representing the attributes of each record are obtained. The similarity of remote sensing data records is calculated, and the similar and repeated remote sensing data records are removed by sorting and merging. According to the theory of evidence and the relevant evidence synthesis method, the marine remote sensing data of GIS is applied to neural network classifier, and the classification is performed based on the results. The experimental results show that this method has the advantages of good performance, high classification accuracy, high recall rate, and high data-cleaning precision.