¹College of Automation Engineering, Qingdao University, Qingdao 266071, China
²College of Information Engineering, Qingdao University, Qingdao 266071, China
³School of Computer Engineering, Qingdao Technological University, Qingdao 266033, China

---

Abstract—In recent years, enteromorpha prolifera detection has received increasing attention. Supervised learning with remote sensing images can achieve satisfactory performances for green tide monitoring. However, data distributions between images obviously differ, and it would be too costly to label a massive amount of images for enteromorpha prolifera detection. Thus, this paper focuses on detecting enteromorpha prolifera using not only limited labelled data, but also a large amount of unlabelled data. We propose an effective semi-supervised clustering framework for enteromorpha prolifera detection, which can reduce the labelling cost and alleviate the overfitting problem. Experimental results prove the effectiveness and potential of our approach, with almost a 15% increase from baseline. In addition, the proposed approach can provide quantitative assessments for band data of moderate resolution imaging spectroradiometer (MODIS) images, and several often ignored bands, such as bands 5, 6, and 7, are shown to be useful for enteromorpha prolifera detection.

Index Terms—enteromorpha detection, semi-supervised clustering, remote sensing images

[PDF]

Cite: Shunyao Wu, Fengjing Shao, Ying Wang, Rencheng Sun, Jinlong Wang, "Enteromorpha Prolifera Detection with MODIS Image Using Semi-supervised Clustering," Journal of Computers vol. 9, no. 5, pp. 1259-1265, 2014.