Data redundancy can improve the performance of automatic feature extraction in remote sensing image if the concept of data fusion is applied in it. Extensions of classical set union and intersection do not lead to satisfactory results (the corresponding operators are either too indulgent or too severe), but fusing two operators that have opposite properties can obtain an obvious result. In this paper, context dependent fuzzy operator is used to achieve this compromise, and a new formation of weights is proposed. Based on the simulative result, the performance of this method is discussed.