One of the problems of surveillance systems such as militarily systems is low quality visible image in the environment with inadequate illumination or under bad weather conditions. One solution to this problem is using thermal vision technique. In this case, visual and thermal images are often fused for an improved image under different situations or environments. Indeed, with the availability of multisensor data in many fields, such as surveillance system, image fusion has emerged as a new and promising research area. Image fusion is a process of combining two or more source images from different modalities instruments into a single image with more information. There are many methods in image fusion which some of them have acceptable results. These approaches are usually based on transformation methods. A traditional scheme to realize Multi-resolution representation (MRI) is to apply 1-D filters separately in horizontal and vertical directions, commonly referred to as “separable” transform. In contrast, “non-separable” transform consist of 2-D filters and 2-D downsampling matrices which cannot be factorized into 1-D filter/downsampling pairs. Traditional wavelet transform (WT) is categorized as a separable transform, which is used to various applications such as image fusion, compression, noise removal, image edge enhancement, and feature extraction. Contours are abundant in natural images and cannot be categorized as either horizontal or vertical edges. The contourlet transform is used to extract curves which are not detected completely through other transforms like wavelet. The problem of the contourlet is its redundancy which is a bottleneck and decreases its efficiency. In recent year, a new wavelet based contourlet method has been proposed that do not have redundancy. But, the main disadvantage of this method (and contourlet) is the occurrence of artifacts. In this thesis we propose a new image fusion method to avoid this problem. We first introduce contourlet transform and show simulation results. Then we propose non-separable wavelet based contourlet method. Experimental results show that our method has better quality than conventional wavelet base contourlet method in both visually and objective assessments. Keywords: Image fusion, Wavelet, Contourlet, Non-Separable Wavelet, Objective Evaluation.