An adaptive cloud theory filtering algorithm is proposed for salt and pepper noise removal,which divides the pixels of an image into signal part and noise part by means of the cloud maximum judgment criterion.The detected noise pixels are removed by using the adaptive cloud forecast,while the signal ones remain unchanged and are output directly.Experimental results indicate that:1) The proposed method can suppress salt and pepper noise and preserve image details effectively;and 2) The filtering effect of our method outperforms that of the traditional and the improved median filtering,especially under the condition of high-intensity noise.