Abstract: The paper proposes a new adaptive weighted average filtering algorithm based on the feature of local spatial pixels that removes the ‘salt & pepper’ noise.Firstly，all the pixels are separated into signal pixels an noise pixels according to a decision criterion，then，remove the noise pixels step by step using a new adaptive weighted average filtering algorithm that given in the following.The filtering weights are determined by a well-defined mathematical criterion，which is adaptive to the feature of neighboring pixels in the filtering window.The experiment results show that the algorithm can effectively remove the impulse noises and remains the image edge detail at the same time.Especially，the image is highly corrupted，it can provide more significant improvement than the conventional median filter.