ABSTRACT Recently, demand for digital image processing devices increases rapidly, more clear images have been required. But, in the process of digital image acquisition, processing and transmission, image degradation occurs due to various external reasons and researches about noise reduction are on the rise. Therefore, this study suggested the algorithm to process AWGN(additive white Gaussian noise) by separately processing as three levels according to the pixel distribution in the mask in order to remove AWGN(additive white Gaussian noise) which is added in the image. Regarding the processed results by applying Barbara images which were damaged by AWGN(), suggested algorithm showed the improvement by 2.87[dB], 2.95[dB], 2.88[dB], 1.52[dB], 1.49[dB], 1.58[dB] and 1.25[dB] respectively compared with the existing MF(×), A-TMF(×), AWMF(×), MF(×), A-TMF(×), AWMF(×), GF(×). 키워드 : AWGN, 열화 영상, 공간 가중치 필터, 적응 가중치 필터 Key word : AWGN, Corrupted image, Spatial weighted filter, Adaptive weighted filterJournal of the Korea Institute of Information andCommunication Engineering