Infrared and visible images fusion based on RPCA and NSCT

• RPCA is employed to obtain the sparse matrix of the source infrared and visible images.
• The sparse matrix of an image represents the salient information in it.
• The fusion scheme takes advantage of the translation invariance of NSCT.
• A novel infrared and visible image fusion algorithm is proposed.

Current infrared and visible images fusion algorithms cannot efficiently extract the object information in the infrared image while retaining the background information in visible image. To address this issue, we propose a new infrared and visible image fusion algorithm by taking advantage of robust principal component analysis (RPCA) and non-subsampled Contourlet transform (NSCT). Firstly, RPCA decomposition is performed on the infrared and visible images respectively to obtain their corresponding sparse matrixes, which can well represent the sparse feature of images. Secondly, the infrared and visible images are decomposed into low frequency sub-band and high-frequency sub-band coefficients by using NSCT. Subsequently, the sparse matrixes are used to guide the fusion rule of low frequency sub-band coefficients and high frequency sub-band coefficients. Experimental results demonstrate that our fusion algorithm can highlight the infrared objects as well as retain the background information in visible image.