在现有图像质量评价方法相关原理基础上,提出了一种基于图像边缘信息的拼接质量评价新方法。该方法针对图像拼接结果的特点,先对待评价图像进行边缘提取,然后利用拼接前后图像的边缘轮廓信息,综合图像像素误差信息和结构信息,根据其均值和方差等统计信息与影响图像拼接质量的主要因素(拼接错位和亮度突变)之间的关系,对拼接图像进行评价。该评价方法得出的评价结果更加符合人眼视觉对图像拼接质量的主观评价感受,较准确地反映了拼接图像的真实质量和所使用图像拼接算法的性能。
On the basis of the theory of existing methods for image quality assessment, a novel method for quality as- sessment of mosaicked image based on the information of image edge was presented. The method was in accordance with the features of mosaicked image. Firstly, the edges of image were detected. Then according to the information of image edges, the information of pixel error and structure was considered. The mean value and variance of the difference of edge map were related with the misplacement and the brightness mutation, which influenced the quality of mosaicked image. Lastly, the evaluate procedure was done based on the relationship. The new method was more consistent with the subjec- tive feeling to the mosaicked image quality. This method accurately reflects the real quality of mosaicked image and the performance of the algorithm of image mosaic.