Xing, Z. and Li, G., 2020. Research on intelligent recognition of coast image features in distributed system based on deep learning. 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. 26–31. Coconut Creek (Florida), ISSN 0749-0208.

The traditional content-based image recognition method has the disadvantages of low recognition efficiency and poor precision for coast images. Therefore, a massive image recognition method based on Hadoop is proposed in this paper, which implemented distributed computing for coast digital images. In this paper, the speeded up robust features (SURF) of images were acquired, and the SURF features of similar pictures were clustered via the k-means clustering algorithm. Finally, the image features were quantified by term frequency-inverse document frequency data mining technology, and the image was shaped according to the SURF features input by a user to achieve accurate identification of similar images. The analysis of results showed that the recognition means has superior efficiency and accuracy for identifying massive images, which greatly improved the recognition performance of the image and enhances the robustness of the system.