Research Article
An Objective Evaluation Method for Image Restoration
Table 1
Details of the experiments conducted.
| | Experiment | Images | Blur | Noise | De-convolution methods |
| | Exp 1a | single chip image | Box blur ( pixels) | No noise | Wiener [16, 17], CLS, Inverse | | Exp 1b | single Lena image | Box blur ( pixels) | Poisson noise | Wiener, CLS, Inverse | | Exp 2a | Two Rice images | Gaussian blur | Gaussian of variance (in dB) (image 1), (image 2) | Wiener, CLS, inverse filter, Total variation [18], Lucy [19] | | Exp 2b | 20 Mountain images | Each blurred by Motion blurs of different trajectory. | Gaussian noise | Wiener | | Exp 2c | Single Cameraman image | Box blur ( pixels) | Gaussian noise | CLS [9] with lambda 0.1 to 80 in increments of 10 |
|
|
