[WEEK 4] Image Colorization Results With Filters
This week, we tried a different filter. This filter is a bilateral filter. In this post W will explain how the bilateral filter affects our results.
Bilateral filter is higly effective in noise removal while keeping edges sharp. Last weeks, we tried to use gaussian blur and unsharpen. In gausian blur we lost edge information and some results not work well. In unsharpen filter, we got edge information but for some images, colorization did not work well. Gaussian filter takes the a neighbourhood around the pixel and find its gaussian weighted average. It does not consider whether pixel is an edge pixel or not. So it blurs the edges also. Bilateral filter also takes a gaussian filter, but it makes sure only those pixels with similar intensity to central pixel is considered for blurring.
If we compare normal output and bilateral output, there are no huge differences. In small detail Bilateral lost information, but if we consider big detail it gives more beatiful images.
Next weeks we consider other filter and try to get more good results. See you next week.
