Classification of MODIS images combining surface temperature and texture features using the Support Vector Machine method for estimation of the extent of sea ice in the frozen Bohai Bay, China

Image classification of frozen areas and adjacent sea ice is important for monitoring the evolution of ocean freezing. This paper proposes a novel approach to the Moderate Resolution Imaging Spectroradiometer (MODIS) image classification and estimation of the extent of sea ice in frozen areas during recent global surface warming hiatus. We derived the texture feature (TF) and surface temperature (ST) from the MODIS image for classification and sea ice detection. We extracted MODIS TF by a grey-level co-occurrence matrix (GLCM), and retrieved the MODIS ST using a split-window method, and finally classified the image using a Support Vector Machine (SVM) convoluting the ST and TF methods. Results were compared and validated with those of conventional spectral-based supervised classification approaches. Results show that the overall accuracy and kappa coefficient (.) using the proposed method was much higher in comparison with those of the spectral-based maximum likelihood and SVM methods. The SVM fusion ST and TF method was effective and useful for MODIS 500 m image classification and sea ice mapping in frozen area. Combining ST and TF can improve sea ice extent estimation accuracy in the frozen Bohai Bay.