子空间辨识彩色融合图像质量主观评价实验

Subjective Evaluation for Quality Assessment of Color Fusion Images Using Subspace Identification Method

为了发掘图像质量单一评价指标对融合图像目标探测性的影响,采集典型场景图像,开展彩色融合图像主观评价实验。考虑相邻图像的质量评价结果对当前图像评价结果的影响,提出采用子空间辨识方法建立基于目标探测的融合图像综合质量高阶预测模型。通过对实验结果的分析与处理,证实该方法的可行性和有效性,获得了较好的实验结果和实验教学效果,为进一步实现基于目标探测的双波段彩色融合图像质量客观评价提供实验基础与理论依据。

In the paper, the subjective evaluation experiments are carried out to find out that the single evaluation index affects the object detection probability of the fusion image. Considering the effects of the neighbors′ evaluation results on current the subjective evaluation, we present higher-order fusion images comprehensive quality evaluation models of different scenarios for the object detection based on the subspace identification method. Data analyzing and processing are carried out, and satisfactory experimental results and teaching effectiveness are obtained. In addition the experiments provide the theoretical foundation and experiment evidence for comprehensive quality of color fusion images of on the detection probability of the object.