基于语义标签的高铁接触网图像目标检测研究

Research on Object Detection Method of High-Speed Railway Catenary Image Based on Semantic Label

基于深度学习的目标检测模型近些年快速发展,可以准确检测出物体在图像中的位置。但是多数检测模型一般不考虑物体之间的相对位置关系,而图像中物体间的相对关系包含有效高层语义信息,有利于对特殊的指代个体进行目标检测。通过提出基于语义标签与谓词字典构建图像中物体之间的相对位置关系的方法,对高层语义信息进行建模,实现物体感知任务与语义认知任务的协同,使图像目标检测模型具有推测相对位置语义的能力。论文通过对京津段高铁接触网图像的目标检测进行实验,实验结果表明上述方法的可行性。

The object detection model based on deep learning has developed rapidly in recent years,and can accurately detect the position of an object in an image.However,most detection models generally do not consider the relative positional relationship between objects,and the relative relationship between objects in the image contains effective high-level semantic information,which is conducive to target detection for special reference individuals.This paper proposes a method of constructing the relative position relation between objects in the image based on semantic tag and predicate dictionary to model the high-level semantic information.Therefore,the coordination of object perception task and semantic cognitive task is achieved and the image object detection model has the ability to speculate the relative position semantics.In this paper,an experiment was carried out on the object detection of the catenary image of high-speed railway in Beijing-Tianjin section,and the experimental results show that the above method is feasible.