一种新的各向异性扩散红外噪声实时抑制算法_王娴雅.pdfVIP

29 1 Vol. 29, No. 1 2009 1 ACTA OPT ICA SINICA January , 2008 : 0253-2239( 2009) 01-0176-04 一种新的各向异性扩散红外噪声实时抑制算法 王娴雅 陈 钱 顾国华 ( , 210094) , , Perona-Malik( P-M ) , EVM- DM642 , : ; ; ; ; P-M N911.73 A doi: 10. 3788/ AO 0176 Rea-l TimeInfraredNoiseReduction Based on Anisotropic Diffusion Method Wang Xianya Chen Qian Gu Guohua (Ministerial Key Laboratory of JGM , Nanjin g Un iversity of Science an d echnology , Nanjing , J ian gsu 210094, China ) Abstract An infrared noise reduc ion algori hm based on aniso ropic par ial differen ial equa ion is proposed. By combining morphological opera ions wi h infrared imaging local fea ures, a new aniso ropic coefficien is es ablished. This coefficien ge s gradien opera or by morphological dila ing and eroding, improves he sensibili y of Perona-Malik (P-M) gradien , s reng hens diffusion in homogeneous area and weakens diffusion in he erogeneous area. The algori hm has run on he EVM-DM642 hardware pla form in real ime, and experimen s indica e he noise is smoo hed ou effec iv ely, and he image edge is also preserved be er . Key words image processing; infrared noise; aniso ropic diffusion; morphology ; Perona-Malik (P-M) equa ion 1 , [ 1, 2] , ( - 1/f ) ( [ 6] , , ) Catte Lions , , , , R, , , Perona-Malik ( P-M) [ 3, 4]