Fast seam carving with strip constraints

In this paper, we present a fast seam carving approach with strip constraints. To reduce the artifacts caused by seam carving, the seam carving is constrained in several non-overlapped strips, whose objective size is unequal. The whole resizing scheme involves the following steps. Firstly, we divide the source image into several strips of equal space and calculate the importance of each strip by averaging the saliency values. Secondly, we distribute the size of the target image into all strips by their importance with the constraints of the objective function. Thirdly, with the objective size, we could determine the number of removed seams for each strip. Then the fast seam carving scheme is used to resize the strip to the objective size. The experimental results show that suitable number of grids could avoid removing the seams adjacently and could improve the image quality visually.

[1]  Ariel Shamir,et al.  Improved seam carving for video retargeting , 2008, SIGGRAPH 2008.

[2]  Christof Koch,et al.  A Model of Saliency-Based Visual Attention for Rapid Scene Analysis , 2009 .

[3]  Ariel Shamir,et al.  Seam Carving for Content-Aware Image Resizing , 2007, ACM Trans. Graph..

[4]  Hua Huang,et al.  Real-time content-aware image resizing , 2009, Science in China Series F: Information Sciences.

[5]  Olga Sorkine-Hornung,et al.  Visual media retargeting , 2009, SIGGRAPH ASIA Courses.

[6]  Xing Xie,et al.  A visual attention model for adapting images on small displays , 2003, Multimedia Systems.

[7]  Lihi Zelnik-Manor,et al.  Context-Aware Saliency Detection , 2012, IEEE Trans. Pattern Anal. Mach. Intell..

[8]  Yasuyuki Matsushita,et al.  Image retargeting using importance diffusion , 2009, 2009 16th IEEE International Conference on Image Processing (ICIP).

[9]  Ariel Shamir,et al.  Cropping Scaling Seam carving Warping Multi-operator , 2009 .