Estimation of category mixture rate for boundary area on remote sensing image due to fuzzy inference

The authors propose a fuzzy estimation method for the category mixture rate in the mixed pixel (mixel) of the satellite image. The distribution of the brightness lever for the training data is handled by a fuzzy set, and the category mixture rate in the mixel is estimated by fuzzy inference. The proposed method, therefore, takes into consideration the scattering of the training data in the membership function. The simulation result shows that the proposed method is a useful technique for estimating the category mixture rate of the mixel located on the boundary area between two different categories