基于SLIC超像素和贝叶斯框架的显著性区域检测

Salient Region Detection Based on SLIC Superpixel and Bayesian Framework

提出一种基于SLIC超像素和贝叶斯框架的显著性区域检测算法.首先,在图像的预处理阶段,为了降低计算的复杂度,采用SLIC算法来提取给定图像的超像素.然后,在每一个尺度之下,考虑以下三个准则:区域对比度、完整性以及中心偏差,进而结合贝叶斯框架再进行后续显著性检测;之后通过加权求和以及归一化操作后计算得到最终的显著性图.最后,由一个滤波器来更进一步来提高检测效果以便于优化最终的显著性图.在MSRA10K基准数据库上与当前比较流行的几种方法来进行相关定性和定量的比较,实验结果表明,本文所提算法的性能均高于当前比较流行的方法.

In view of the problem of bottom-up algorithm, an effective SLIC superpixel-based and Bayesian framework-based salient region detection algorithm is proposed. Firstly, in order to reduce the computation complexity, the SLIC algorithm is used to extract the super pixels of a given image in the image preprocessing stage. Next, at each scale, three essential cues such as local contrast, integrity and center bias are considered within the Bayesian framework. Then, the saliency maps are computed by weighted summation and nor- malization. Finally, the final saliency map is optimized by a guided filter in order to further improve the detection results. A series of relevant qualitative and quantitative experiments are done on a MSRA10K dataset containing 1,000 test images with ground truths, and results show that the performance of the proposed saliency algorithm is better than the most prevailing current methods.