Oil spill extraction by X-band marine radar using texture analysis and adaptive thresholding

ABSTRACT

Oil spills cause damage in ocean environment. It is important to identify the oil spills for further treatment. We propose an oil spill detection method based on X-band marine radar image using texture analysis. In this method, first, received radar image was denoised by erasing co-channel interference and small speckles. Then, texture analysis was used to indicate the location of oil spills. In the texture analysis, we proposed a texture index calculated by four texture features of grey level co-occurrence matrix (GLCM). The texture index in oil spill area is higher than it in other place, which can be used for extracting oil spill area roughly. In the end, precise extraction of an oil spill was carried out by an adaptive thresholding algorithm on the area selected by the proposed texture index. According to the X-band marine radar images sampled in the oil spill accident on Xingang Port of Dalian, China, an oil spill detected by the proposed texture analysis method is comparable to visual interpretation.

Additional information

Funding

This work was supported by the Fundamental Research Funds for the Central Universities [grant number 3132018150] and the Doctoral Scientific Research Foundation of Liaoning Province [grant number 201601062].