LAUSR

ABSTRACT Luo, D., 2019. Intelligent classification method for semi-supervised marine remote sensing images based on clustering kernel function. In: Hoang, A.T. and Aqeel Ashraf, M. (eds.), Research, Monitoring, and Engineering of Coastal, Port, and Marine Systems. Journal of Coastal Research, Special Issue No. 97, pp. 136–142. In order to improve the retrieval and recognition ability of marine remote sensing images, the method of data clustering and mining is used to optimize the classification of marine remote sensing images. The traditional ocean remote sensing image classification method adopts the image edge contour segmentation method and combines the edge contour feature extraction to realize the marine remote sensing image clustering retrieval. The classification accuracy of the marine remote sensing image is not high under the action of large disturbance. A classification algorithm of ocean remote sensing images based on clustering kernel function semi-supervised learning is proposed, and a remote monitoring transmission channel model of ocean remote sensing images under semi-supervised learning mode is constructed. Fuzzy fusion clustering and vector quantization coding are applied to the collected marine remote sensing images. The feature points are extracted and mined by using the feature detection algorithm of remote sensing spatial atlas, and the extracted feature points are taken as data input. The fuzzy kernel function clustering algorithm based on semi-supervised learning is used to extract the features of ocean remote sensing and classify marine remote sensing images. The simulation results show that the proposed algorithm is accurate for classification of marine remote sensing images in large marine remote sensing geographic information database, and has high accuracy for detecting feature points in marine remote sensing images, and the output peak signal-to-noise ratio (PSNR) is improved. The accuracy and robustness of ocean remote sensing image classification are improved.