一种图像清晰度评价方法

图像清晰度评价在图像分析和识别中具有重要的意义.对于同一内容的图像,清晰图像的边缘相对模糊图像具有较大的灰度变化率的特征.提出了一种以图像边缘灰度变化率为指标的图像清晰度评价方法,即以每一行灰度值连续下降间隔像素点数最多的边缘灰度变化率代表这一行的灰度变化率,计算出图像所有行灰度变化率均值为评价指标.分析了清晰图像与模糊图像某行边缘灰度变化规律,设计了这种新的图像清晰度评价方法的实现流程.以红外图像和一般的可见光和运动模糊图像为对象,进行了该方法与相邻像素灰度方差法和方差法对比研究.实验结果表明:该方法能更好地满足单调性和无偏性,具有更高的灵敏度.

Image clarity-evaluation-function method

Clarity-evaluation-function is very important in image processing and recognition. It is found that the clarity image has obviously high grey gradient than the blurred image while the object images are same. A new clarity-evaluation-function of grey gradient was proposed. The edge grey gradient,whose spaced pixle numbers of continue grey decline in a line were most,was regarded as the grey gradient in the line. The average change of grey gradient in all line in the whole image was counted as the clarity-eva...