基于马尔可夫随机场的显著区域检测

Visual Salient Objects Detection Based on MRF

显著区域检测被应用在计算机视觉处理的各个方面,然而大目标和复杂背景中显著对象检测存在检测内容缺失和误检的问题。提出一种基于粗略显著区域的马尔可夫随机场(MRF)模型检测方法。首先,应用Harris角点检测定位粗略显著区域,去除边缘附近角点做凸包操作,然后基于粗略区域的先验概率,应用马尔可夫随机场模型检测图像显著区域。在图像测试库的实验结果显示,提出的检测方法提高了检测的准确度和完整度。

Salieny detection is utilized in many vision tasks in recent years. But parts of objects missing and detect false are still big issue when the object of image is huge or the background is complex. In this paper, we propose a new computational saliency dectection model which is implemented with a coarse to fine strategy under MRF frame- work. Firstly, saliency points based on Harris are applied to get a coarse location of the saliency region. And then, based on the rough region, we compute a prior map for the MRF to achieve the final sailent objects detection. Experimental results on a dataset show that the detected section based on the model is more accurate and complete.