Infrared and Visible Images Fusion Algorithm Based on Curvelet Transform

Aiming at the shortcoming of wavelet which can't efficiently capture image contour,this paper proposes an image fusion algorithm based on the second generation curvelet transform.It proposes a fusion rule containing weighted mean approximate coefficients and details coefficients combined with extended pixel-level multiresolution fusion framework,as well as contrast sensitivity band-pass function.It employs the subjective and objective evaluation to assess fusion results.Experiments indicate that the proposed approach is superior to wavelet in preserving source image edge contour and noise suppression,and fusion images better correspond with human visual perception characteristics.