视觉评价夜视彩色融合图像质量的实验研究

STUDY ON PERCEPTUAL EVALUATION OF FUSED IMAGE QUALITY FOR COLOR NIGHT VISION

提出夜视彩色融合图像质量评价的三个基本指标是“目标探测性”、“细节”和“色彩”,并且通过视觉评价实验,研究了“感知质量”与这三个指标的相关性.对十个场景的微光与红外图像采用四种融合算法产生的40幅假彩色图像进行评价实验,实验结果得出:融合图像都有好目标探测性,感知质量与目标探测性的相关性很小;感知质量与细节的相关系数是0.89,感知质量与色彩的相关系数是0.75.说明在满足目标探测性要求条件下,细节是视觉感知融合图像质量的最主要评价指标,而且色彩成为夜视彩色图像质量的一个重要指标.

With the development and the application of visible-IR image fusion techniques in color vision night, evaluating the performance of image-fused algorithms is becoming an important aspect. In this study, three basic visual evaluation factors, target detection, details and colorfulness, were presented for evaluating perceptual quality of color fused images, and the degree of correlation between the perceptual quality and the three evaluation factors was investigated. Visual evaluation experiment on the color images fused visible and IR images of ten scenes by four fusion algorithms was conducted. The experimental results show that the fuesd images have good performances in target detection, the correlation coefficient between the perceptual qulity and details is 0.89, and the correlation coefficient between the perceptual qulity and colorfulness is 0.75. It indicates that details are the main factor and the colorfulness plays an important role when targets are beyong the detected level.