利用多尺度语义模型的复杂图像目标自动提取方法
Automatic Object Extraction Using Hierarchical Semantic Model
-
摘要: 针对现有方法对复杂图像目标自动提取性能欠佳等不足,提出了一种新的利用多尺度语义模型的目标自动提取方法。首先,采用多尺度分割得到的图像块作为目标提取的候选区域;然后,利用语义模型获取目标的语义分布信息;最后,目标提取阶段统计各个图像块的语义相关函数,并通过最大化该函数确定出目标。实验结果表明,此方法能够准确、有效地提取出目标,精度高,用户工作量少。Abstract: A new hierarchical semantic model method is proposed to overcome the disadvantages existing in most of the relative methods.Firstly,a multi-scale segmentation is employed to obtain the candidate object regions.Secondly,semantic model is utilized to get objects semantic distribution information.Lastly,in the extraction processing stage,a correlative function based on the semantic model and the candidate object regions is built up and maximized to extract the objects.Experimental results demonstrate the precision,robustness,and effectiveness of the proposed method.
下载:
