A New Image Quality Approach Based on Decision Fusion

Image quality evaluation is becoming essential in many image processing problems. This paper proposes a new image quality evaluation approach based on decision fusion method of canonical correlation analysis (CCA). By exploring several diverse visual models, we constructed a comprehensive quality metric which can deal with complicated image distortion problem with increasing accuracy and robustness. Validation by comparing the proposed metric against other image quality metrics (IQMs) demonstrates that its fidelity prediction performs better across wide distortion range and types.