Object bank (OB) representation is a novel image representation for high-level visual tasks,which encodes semantic and spatial information of the objects within an image.However,the poor precision of the object detectors in OB method influences the extraction effect of high-level image feature.In order to solve this problem,a new OB method improved by Graph-Based Visual Saliency (GBVS) is proposed.Firstly,GBVS saliency model is utilized to process the image and detect the saliency regions and extract better high-level feature.The experiments results show that the proposed method performs better in classification and increases the classification accuracy of 4％.%Object Bank (OB)是一种基于物体的高层语义图像特征表示,该方法所提取的高层特征具有较丰富的场景语义.然而,该方法所采用的物体检测器检测物体的准确率并不高,从而影响高层特征的提取效果.针对OB方法中物体检测器准确率较低的缺点,提出一种基于Graph-Based Visual Saliency (GBVS)显著性分析算法改进的OB方法.先通过GBVS方法对图像进行显著性处理,计算图像中的显著性区域,然后结合OB方法中的物体检测器对显著区域进行检测,提取更具有场景语义的高层特征.实验结果表明,该方法突出了具有显著性的物体,提高了OB方法中目标检测器的准确率,在OB方法的基础上提取出了更具有显著性的图像特征,并在分类准确率上比OB方法提高了4％.
展开▼
