图像融合质量主客观评价相关性分析

Analysis of the correlation between objective and subjective evaluation of image fusion quality

建立图像融合质量评价标准对于在不同的应用场景中选择适当的融合算法,以及对现有融合算法的改进和研究新的融合算法都是至关重要的。本文在分析目前广泛使用的图像融合质量主客观评价方法的基础上,对图像融合质量进行了主客观评价并对主客观评价结果进行了相关性分析,发现结合了人类视觉特性的客观评价方法与主观分析结果较为一致,充分说明了建立新型的、结合主观因素与客观因素的图像融合评价准则的必要性。

Establishment of image fusion quality evaluation standard is important for the evaluation of image fusion.It can help to select appropriate algorithm in practical applications and improve the existing algorithm performance and develop new algorithms.Based on the analysis of existing objective and subjective evaluation methods of image fusion quality,several fusion methods are evaluated by different objective and subjective evaluation methods and the correlation between the objective and subjective quality metric are analyzed.The experimental results show that the objective quality metric based on HVS is more consistent with subjective results.These results give a well support to the necessary of establishing new evaluation method with combined objective and subjective factors.