区域的深度计算及其在图象渐变处理中的应用

The Computation of Region Depth and Its Application to Image Feathering

为了解决局部图象处理容易产生边界效应的问题 ,可以区域中任意一点的深度值为基础 ,来得到权函数的数学表达式 ,然后根据权函数对图象上给定区域范围内的象素进行处理 ,使区域中离区域边界较近的象素的灰度值改正量小一些 ,而使区域中离区域边界较远的象素的灰度值改正量大一些 ,从而实现区域图象的渐变处理 .另外 ,为了克服迭代算法速度慢的缺点 ,还提出了一种快速标记多边形区域内部点的深度值算法 .实验结果表明 ,所提出的渐变处理方法 ,在克服区域图象增强处理时容易产生的边界效应方面是十分有效的 ,可以获得满意的效果 .

In order to resolve the problem of image feathering in the region, the formula of weight function is presented based on the depth information of the region and every pixel in the region is processed according to the weight. As a result, the closer the pixel is to the region boundary, the less the correction of pixel value is, and the farther the pixel is to the region boundary, the more the correction of pixel value is. This method of image feathering is also tested using real images. In order to overcome the time consuming disadvantage of iterative algorithm for computation of region depth, a fast algorithm is also proposed and tested using convex, concave and cross polygon region. The experiment shows that the method proposed above is very effective and can acquire good results in image feathering of the region.