欢迎访问《图学学报》 分享到:

图学学报

• 视觉与图像 • 上一篇    下一篇

基于梯度边缘最大值的图像清晰度评价

  

  • 出版日期:2016-04-28 发布日期:2016-05-20

Sharpness Assessment for Remote Sensing Image Based on Maximum Gradient

  • Online:2016-04-28 Published:2016-05-20

摘要: 光电遥感相机获取的遥感图像传回地面后都会出现一定降质模糊现象,通过复原可
以得到较高质量的遥感图像。得到复原图像后,需要一些方法来评价复原图像质量提高的程度。
为此提出了一种遥感图像清晰度的评价方法,通过提取图像的梯度,找出梯度最大的区域,计算
出该区域的像元个数,并以此作为图像的清晰度参数,计算复原图像清晰度与降质图像清晰度之
间的相对误差,得到清晰度提升率。通过设计实验来验证该方法的有效性。结果表明,该方法可
以准确评估复原图像的清晰度提升率,对弱振铃波纹有较好地抑制效果,在信噪比大于22 dB 时,
评价结果不受噪声影响。

关键词: 图像质量评价, 图像清晰度, 遥感图像, 梯度, 信噪比

Abstract: The image acquired by the remote sensing camera will appear to be a certain fuzzy
phenomenon after transmitted it to the ground. It will get high quality image after restoration, and it is
necessary to use some methods to assess the quality of recovery image. This paper proposes a sharpness
assessment for remote sensing image, which picks up gradient from images, calculates the number of
the pixels on the region of maximum gradient and the data is used as sharpness parameter. We get the
upgrade rate after comparing the sharpness parameters between the fuzzy image and the recovered
image. We demonstrate the validity of this new method through some experiments. The experimental
results show that the method can accurately assess the sharpness upgrade rate of the restored image, and
get better results in the inhibiting effect of weak ringing ripples and the anti-noise performance when
the signal noise ratio over 22 dB.

Key words: image quality assessment, image resolution, remote sensing image, gradient, signal noise
ratio