Perceptual Quality Assessment of SAR Image Compression Based on Image Content Partition and Neural Network

Image quality assessment (IQA) is of fundamental importance for image compression applications. Traditional IQA measures used for Synthetic aperture radar (SAR) image compression do not consider the properties of Human visual system (HVS). Since human beings are the final users in most SAR image applications, the objective evaluation coordinate to human's perception is the most acceptable and practical IQA method. In this paper, we propose a novel objective approach based on image content partition and Neural network (NN) by introducing the HVS and SAR image characteristics. Experimental results demonstrate that the proposed metric correlates well with subjective quality of SAR image compression and outperforms those state-of-art objective models using Structural similarity index (SSIM), Singular value decomposition (SVD) and Visual information fidelity (VIF).