高鲁棒性图像显著区域提取及描述方法

Detection and Description of Image Salient Regions

摘要: 针对现有的图像显著区域提取及描述方法中存在的计算复杂、鲁棒性差等问题,提出了一种新的显著区域提取和描述方法.首先,通过基于连通区域的方法提取图像中的显著区域,然后,采用不变矩描述子对区域的特征进行描述.实验结果表明,该方法能有效地提取和描述图像中的显著区域,且对平移、旋转和缩放变换具有很高的鲁棒性.

Abstract: To solve the problems of complexity and low robustness that exist in present image salient region detection and description methods, a new salient region detection and description method was proposed. First, salient regions were detected using a method based on connected regions. Then, an invariant moment descriptor was used to descript the detected regions. Experimental results show its effectiveness and robustness to translation, rotation and scaling.