NIF-based seam carving for image resizing

作者：Guo, Dongyan; Ding, Jundi; Tang, Jinhui^*; Xu, Min; Zhao, Chunxia

来源：Multimedia Systems, 2015, 21(6): 603-613.

DOI：10.1007/s00530-014-0425-6

• Image resizing
• Seam carving
• Neighborhood inhomogeneity factor
• Content-aware image manipulation

In this paper, improving the existing seam carving technique, we propose a novel image-resizing method which considers both texture and color information of the original image. Neighborhood inhomogeneity factor (NIF), which describes the image inhomogeneity, is used to compute the saliency map of an image in the Lab space. Although a saliency map well preserves the uniformity of important areas within an image, it lacks details of the image. Hence, we propose to combine a gradient map of an image with the saliency map to generate a final energy map for seam carving. Extensive experiments and comparisons are conducted to evaluate the performance of the proposed NIF-based seam carving method. Experimental results demonstrate that the proposed method preserves the original details and scale of important areas and achieves robust resizing results with reducing distortion, even when the background shows a high degree of resemblance in appearance to the contained objects .

• 出版日期2015-11
• 单位南京理工大学