To detect weak targets in complex infrared background

this paper presents a new infrared image background suppression algorithm based on the 2rd generation Curvelet transform and ProbShrink. As the weak targets are described by high frequence of the 2rd generation Curvelet coefficients

we preserve these coefficients for later usage. Marginal statistics and joint satistics of the high frequence coefficents are used to find the positions of the target subband

then ProbShrink algorithm is used to suppress the background subband coeffients. Experiments show that our algorithm supress the infrared backgrounds well and preserve the weak targets.