@article{oai:nagano-nct.repo.nii.ac.jp:00000207,
 author = {宮崎, 敬},
 journal = {長野工業高等専門学校紀要},
 month = {Jun},
 note = {application, An impulsive noise, one of some noises included in a picture image, is a factor depressing image quality. In image processing, it is an important process to reduce such a noise in a picture image. The median filter and the improvement type median filter are effective against the elimination of this impulsive noise. But it is not sufficient in respect to the depreciation of image quality, or preservation of detailed edges. In this paper, the conditional median filter which uses a wavelet transformation and a fuzzy reasoning is proposed. This median filter is based on a impulsive noise being obtained as one of high frequency components, because the power difference of an impulsive noise compared to surrounding pixels is large. Finally, the method using the fixed threshold and the method using the fuzzy reasoning were compared to the conventional method, and then it turned out that the method using the fuzzy reasoning was most effective.},
 pages = {29--34},
 title = {インパルス性雑音除去へのウェーブレット変換の応用},
 volume = {36},
 year = {2002}
}