Distorted Image Quality Assessment Using Weighted Method of Wavelet Simulated CSF Property

Previous methods to evaluate image quality hardly can assess distorted image effectively. Their results usually cant match up to the human visual assessment. To assess distorted images in accuracy, this article proposes a W-MSSIM image quality assessment using weighted method of wavelet frequency band simulated CSF property. Since SSIM is simple and efficient, the proposed method adopts SSIM model to displace the formulas, which express the elements influencing human visual, and to simplify the process of human visual system simulation. CSF property can be weighted by wavelet coefficients of different frequency bands. Hence, the middle and high frequency information can be extracted from tested image and the reference image, respectively. Further, the information of two images is compared to overcome the limitation of SSIM, which cant effectively assess distorted image. The experimental test proves that the performance of proposed method is better than HVSNR and ESSIM.