Journal of System Simulation ›› 2016, Vol. 28 ›› Issue (9): 2169-2175.

Previous Articles     Next Articles

Image Resizing Method Based on Gradient and Visual Saliency

Cai Xingquan1, Buni Honghao1, Qiao Yuqing1,2, Liu Jinghua1,3, Li Fengxia4   

  1. 1. School of Computer Science, North China University of Technology, Beijing 100144, China;
    2. Department of New Media, Kennesaw State University, Atlanta 30060, USA;
    3. Xingtang Communications Technology Co. Ltd, Beijing 100191, China;
    4. Beijing Laboratory of Intelligent Information Technology, Beijing Institute of Technology, Beijing 100081, China
  • Received:2016-05-30 Revised:2016-07-14 Online:2016-09-08 Published:2020-08-14

Abstract: In order to delete the image region without obvious features, and not damage the main features of the target for mobile applications, the image resizing method was provided based on gradient and visual saliency. The original image was grayed, and the gradient energy matrix from the gradient of neighborhood pixels was obtained. The complex energy matrix was obtained by weighted sum of visual saliency energy and image gradient energy matrix. The cumulative energy matrix and minimal energy trajectory matrix were calculated, and the minimal energy seam of image was got. The image resizing was completed by deleting the minimal energy seam in horizontal vertical interspersed sequence. The experiments results show that the complex energy matrix combined the gradient energy with visual saliency energy can recognize target object better, and this method discards the background without obvious features and preserves target object features.

Key words: image resizing, visual saliency, gradient energy, minimal energy seam

CLC Number: