Infrared image background suppression based on 2nd generation-Curvelet transform and ProbShrink algorithm

To improve the detecting performance for the weak targets in complex infrared background, a new infrared image background suppression algorithm based on the second generation Curvelet transform and ProbShrink algorithm is present. As the weak targets are described by the high frequence coefficients in the second generation Curvelet transform and the second generation Curvelet coefficients are sparse representations for edges, so that the parts of complex backgrounds can be subtracted by suppressing low frequency coefficients. The marginal statistics and joint statistics of the high frequence coefficents are used to find the positions of target subbands, then ProbShrink algorithm is used to effectively suppress the background subband coefficients. The experimental results show that the output SNR of complex cloud background with 4.16 dB input SNR can achieve 39.38 dB, which means presented algorithm can suppress infrared backgrounds and can preserve weak targets well.