随着基于显著性的视觉注意计算框架的演化和机器视觉应用的不断增长,显著物体提取成为基于显著性的视觉注意研究领域的重要研究方向。文中介绍了显著物体提取算法的最新研究成果。首先给出了显著物体提取的关键问题。然后根据不同的提取框架对现有的显著物体提取算法进行了分类和分析。并在一个包含1 000幅图像的公开的显著物体数据集上对不同的显著物体提取算法进行了评测。最后总结并展望了显著物体提取算法下一步发展方向。
With the evolution of saliency - based visual attention computation framework and the increase of computer vision applications, salient object extraction becomes an appealing topic in saliency -based visual attention research area. In this survey, the paper introduces recent development of salient object extraction algorithms. First, it provides key issues of salient object extraction, then categorizes and analyzes existing salient object extraction algorithms according to different extraction frameworks. And it uses a public salient object dataset including 1000 images to evaluate different salient object extraction algorithms. Finally, the paper states the conclusion and discusses the future work on salient object extraction alg rithms.