An Image Repair Method Based on Grid Extraction
Ling Xia Liu, Qiang Song, Yu Feng Shen · Advanced materials research · 2012
The repair of digital images generally needs people to manually mark the areas under repair. This paper puts forward a new method, which can automatically mark the areas under repair in the images --- mesh obstacles. First mesh obstacle can be treated as grid, via this obvious feature that grid is formed by two groups of regular cross line in the space, and then through the calculation to extract the mesh obstacles covering on the images, so this don't need people to do manually mark. In order to guarantee the grid lines can completely cover the images, the extracted grid lines can be done proper inflation. Experiments prove this method is effective and much easier to achieve.