Document image analysis: issues, comparison of methods and remaining problems
Artificial Intelligence Review
| Hi-index | 0.00 |
In order to eliminate noise and preserve the edge and texture detail in image, this article proposes a method of image denoise based on soft-threshold and edge enhancement. To enhance edge, first we use the canny edge detection operator to detect the feature of the image, then we preprocess the edge image with a flat operator, next we use stationary wavelet transform to process the processed edge image and noisy image, then we add these wavelet coefficients at the corresponding level, so we can use the soft threshold denoise technology to eliminate the noise. Since the feature is enhanced before soft threshold denoise, the feature can be preserved well. From the result we can see that comparing to the wiener filter, this denoise method has visually pleasing image and the PSNR is improved. Comparing the traditional soft threshold denoise technology, this method has 1-2dB improvement in PSNR.