图像缩放的分片连续算法

Patch-wise continuous approach to image zooming.

在图像处理中 ,传统的放大或缩小变换不能保证图像内各物体之间边界的清晰 ,这往往不能适合许多场合的应用 .为此 ,提出了一种新的图像缩放方法 .该方法对源图像建立了分片连续数学模型 .应用本方法得到的图像 ,边界分明、层面清晰、色彩丰富 ,忠实地反映了原始图像的面貌 .

Image zooming is an important operation for image processing and visualization, but one main shortcoming of some existing zooming methods is that the object borders may be blurred after transformation. A new image zooming method based on patch-wise continuous surface fitting is presented. With this approach a large varieties of gray images can be zoomed efficiently and keep the clear borders of source images clearly.