关键词: 融合图像质量评价,视觉信息保真度,高斯尺度混合模型,人眼视觉系统 中图法分类号TP391.4文献标识码A

Abstract: Objective quality assessment of fusion image is of fundamental necessary in real-time digital image fusion system．A fusion image quality objective assessment method based on visual information fidelity (VIF) was proposed considering the assessment results of traditional objective assessment methods are disagree with the human subjective assessment results．Assuming that the image sources meet Gaussian scale mixture model,regarding image fusion proces-sing as image signal distortion channel,and considering human visual noise characteristics,VIF digital image fusion objective quality assessment model was established,and an image fusion quality evaluation index－FVIF was derived also．The experient indicated that the method can evaluate fusion image quality comprehensively and is better than other methods．The results of FVIF method are consistent with the human subjective assessment results.

Key words: Fusion image quality assessment,Visual information fidelity(VIF),Gaussian scale mixture(GSM),Human visual system(HVS)