Euclidean Quality Assessment for Binary Images

An objective quality assessment algorithm special for binary images is proposed. In the algorithm Euclidean distance between noise and signal points measures the noise energy and the Euler number change of images describes the structural effects caused by noise. Furthermore assessment on image quality is calculated in quantity in PSNR form. Generally image assessments could be classified into two categories: subjective and objective assessment. The latter is judged by the correlation coefficient with subjective quality measurement Mean Opinion Score (MOS). The experiments show that the results of the algorithm are highly correlative with subjective MOS and the algorithm is more simple and computational saving than traditional objective assessments