Fig 1.
A gray-scale image is pre-processed with local contrast normalization and then a number of image patches are randomly cropped for CNN training, validation and final testing.
Fig 2.
It consists of three layers, the input layer, the hidden layer and the output layer.
Fig 3.
A semantic description of GRNN.
It consists of four layers, the input layer, the pattern layer, the summation layer and the output layer.
Fig 4.
Example of Gaussian blurring images in four databases.
Fig 5.
CNN prediction performance with Ni or Pn changes.
Table 1.
CNN performance with regard to kernel number and kernel size.
Fig 6.
GRNN (left) and SVR (right) respectively perform when the spread parameter σ and the cost function c changes based on learned CNN features.
Fig 7.
One trained kernel visualized by using “monarch.bmp”.
After convolutional filtering with the trained kernel, edge structures is hard to notice in heavily blurred images (y11), while fine structures can be seen in relatively high-quality images (y96).
Table 2.
Performance evaluation with PLCC on Gaussian blurring images.
Table 3.
Performance evaluation of SROCC on Gaussian blurring images.
Fig 8.
The time spent on score prediction of image sharpness.
Several algorithms show promise in real-time image sharpness estimation.
Fig 9.
Effect of color information on our CNN.
Compared to gray-scale input, color image input positively enhances our network’s prediction metrics.