Image Quality Evaluation Based on No-reference Method with Visual Component Detection
Li Zhu, Guoyou Wang · 2007
In this paper, we present a new image quality assessment that does not rely on reference images. Our model is to estimate human quality assessment by detecting visual components, then assessing quality by empirical model. We describe an instance of this model with visual component detection, and quality modeling is estimated using radial basis function (RBF) networks. Experiments with this prototype system show better results for human visual system(HVS).