一种图像尺寸调整的纹理感知突出边缘模型

摘要

图像尺寸调整的目的是将一幅给定图像调整为任意指定大小而不产生视觉扭曲，为此，须定义可靠的图像重要性图。利用总变差正则化空间的高阶统计提出一种新的图像重要性图，称之为纹理感知突出边缘图，基于纹理感知突出边缘图利用重要性扩散方法对图像进行尺寸调整。数值实验结果显示，在图像背景杂乱且受到噪声干扰的情况下，与目前流行的其他方法相比，文中方法能获得更好的图像尺寸调整效果。%Image resizing aims at adapting a given image to an arbitrary size without severe visual distortions .To do this ,it is essential to define a reliable Image Importance Image (IIM ) .By using the Higher Order Statistics (HOS ) of the Total Variation (TV ) regularization space .A novel IIM is introduced ,which is called Texture-Perceived Salient Edge (TPSE) map .The numerical experiments show that the proposed method is more effective for image resizing than the popular methods in the cluttered background and in the presence of noise as well .