Intelligent and autonomous UAV’s navigation, which is based on the compliance of received images from UAVs with those from satellites, is one of the newest types of navigation that has received much attentions from researchers and industrialists of this area. This approach is effective, in terms of electronic warfare and efficiency, when high-quality images are available so that the effective features of images can be extracted. However, blurring is one of the main destructive factors leading to decrease the extraction rate and then weak satellites’ images adaptation. So, Image de-blurring has become a new challenging issue for researchers. In this paper, a new method is proposed for improving image quality, using super-resolution techniques based on Convolutional Neural Network (CNN) with non-linear multi-layer mapping, which plays an important role in de-blurring and removing noises from UAV images. The simulation results show that the proposed method has much better performance compared to the other benchmark techniques in term of peak signal to noise ratio (PSNR) so that the proposed method increases the aforementioned criteria about 5%.