模糊图像质量的无参评价
首发时间:2013-02-13
摘要:随着信息技术的高速发展,数字图像已经成为信息传递的主要方式。在数字图像信号的采集过程中,往往会因为各种原因导致数字图像模糊。例如,采样率过低,对焦不准等。在数字图像的处理过程中,也会由于各种操作导致图像模糊,例如均值滤波等。所以,关于模糊图像的视觉质量评估是一个很重要的评价图像处理方法是否合理以及采样是否充分的方式。在这篇文章中,我们提出一种模糊图像的质量评估算法。该算法在使用自然图像幅度谱的斜率描述自然图像的模糊度的基础上,通过加入了人眼对对比度的感知模型,克服了自然图像幅度谱的斜率对图像对比度感知不敏感的缺点。实验结果表明该算法所得出的客观评价结果与主观评价值之间有良好的一致性、准确性和单调性。
For information in English, please click here
Blurred Image Quality No-reference Assessment
Abstract:With the rapid development of information technology, digital images have become the main way of information transmission. In acquisition and processing of digital images, images are blurred by a variety of reasons, e.g. the low sample rate, out-of-focus, mean filtering and so on. Therefore, the blurred image visual quality assessment is very important to evaluate the effective of image processing algorithm and the reasonable of the sample rate. In this article, a blurred image quality assessment algorithm is proposed. The basis of this algorithm is that the slope of the natural image amplitude spectrum can describe the blur degree of the natural image. In order to overcome the shortcoming of the slope of the natural image amplitude spectrum algorithm, a contrast perception model is added to the slope of the natural image amplitude spectrum algorithm. Experimental results between the objective and the subjective assessment show a good consistency, accuracy and monotonicity.
Keywords: image quality assessment the slope of the magnitude spectral the contrast perception
论文图表:
引用
No.****
同行评议
共计0人参与
勘误表
模糊图像质量的无参评价
评论
全部评论0/1000