An Image Quality Synthetic Evaluation Method Based On Image Space Multi-Channel Similar Decomposition

From the digital camera, satellite digital broadcast, cell phone video communication and so on, we can see that the researches in the field of visual information digitization has been used widely in People's Daily life, image quality assessment study has an important role. But the traditional image quality evaluation method has various defects, and can't adjust and satisfy the current rapid development of image technology needs already. Image quality assessment methods are mainly based on visual perception evaluation method and structured evaluation method, but the two kinds of method has a lot of inadequate. Visual perception evaluation method (HVS basic model) lack of the consider of space similar structure, and the space visual similarity analysis noticed structural analysis not considering the influence of multi-channel decomposition, therefore, must combining the visual perception space model and image space similarity decomposition.This article will space similar decomposition and brightness weighted method unifies, comprehensive consideration of the image space similarity and multi-channel decomposition characteristics and create a new image quality assessment algorithm. The main research contents and achievements:1. According to similar image space decomposition and synthetic methods and photometric and colorimetric theories, black and white image processing method will be expands to color image space, provided based on the three colors and based on color of two color image signal decomposition methods, and based on spatial vector synthesis methods established the design method of color space vector synthetic signal.2. Based on orthogonal transform quantitative method, design a kind of test signal compared to the subjective visual, and verifies the brightness for multichannel weighted extended network decompose the accuracy of model. And on this basis, according to colour visual space corresponding bandwidth experiment, make the weighted extended network method expands to brightness color space, realize the color detail image quality assessment method. 3. Through Combined with the image space similar decomposition and brightness weighted extended network organically, this paper presents an image quality synthetic evaluation method based on image space multi-channel similar decomposition. The method not only considering error image structure, but alse takes care of the influence of visual space dimension. And meanwhile make studied in color space.4. Established an experimental platform of image quality synthetic evaluation method based on image space multi-channel similar decomposition. Using the above three image quality assessment methods, Examine multiple types distortion, and relevant evaluation results are analyzed and compared.In this paper, a new algorithm for image quality evaluation calculation is based on the systematic theoretical analysis and a large number of experimental foundations. From the experiment results, the new algorithm is significantly better than the traditional method, and promote the result of subjective and objective Image quality assessment methods consistency.