摘要
针对Seam Carving图像缩放方法中能量函数的局限性和动态规划算法的复杂性,提出了一种基于图像内容的快速图像缩放方法。结合基于视觉注意力模型的显著度图,完善图像内容的能量描述;同时,提出一种基于图像内容的粗、细能量线自适应机制,与层次加速动态规划一起,加速缩放效率。仿真实验结果表明,该算法在有效保护图像内容的同时,降低了算法的计算复杂度,减少了图像缩放时间。
This paper proposed Seam Carving for image scaling method of energy function of the limitations and the complexity of dynamic programming algorithm which was based on image content fast image scaling method. It incorporated visual attention model based saliency map to make the energy description of image content perfect. Moreover,proposed a content based adaptive seam mechanism to improve the resizing efficiency with the technique of hierarchical acceleration dynamic programming. Simulation results demonstrate that the proposed approach can effectively protect important image content and lower the computational complexity,and thus decrease the time consumption for image resizing.
出处
《计算机应用研究》
CSCD
北大核心
2010年第9期3594-3597,共4页
Application Research of Computers
基金
国家自然科学基金资助项目(60702065)
图像处理与图像通信江苏省重点实验室开放课题基金资助项目(ZK207006)
新型显示技术及应用集成教育部重点实验室(上海大学)开放课题基金资助项目(P200801)
关键词
显著度图
连续度
层次加速动态规划
自适应能量线
saliency map
continuous degree
hierarchically accelerated dynamic programming （ HADP）
adaptive seam
