基于组件树和霍夫森林的场景文字检测识别

Text detection and recognition in natural scenes based on component tree and Hough forest

自然场景中的文字检测与识别是图像理解中的重要部分,在大部分的系统设计中,检测和识别被看成是孤立的两部分进行处理,本文提出使用多类霍夫森林建立一个统一的检测识别框架。同时为了解决霍夫森林在类别增多时识别率下降,以及在尺度多变的情况下定位偏移的问题,文中提出利用组件树提取出具有层级的连通域,同时针对文字本身的特征建立分类器。通过级联该分类器,提取出文本的候选位置并确定目标的尺度大小,为后级精细的定位和识别奠定基础。实验结果显示该方案在检测和识别方面均与目前最优的方案具有竞争性。

Text detection and recognition in natural scenes play an important role in image understanding. In most of current system design, detection and recognition are isolated and processed separately. A unified framework for detection and recognition based on multi-class Hough forest is proposed. In order to improve the performance when the quantity of classes increases, as well as improve accuracy with uncertain scale, component tree is used for extracting connected component with hierarchy, while a set of features based on text characteristics is extracted and feed to a classifier. With the help of the classifier, the scale of the target is determined and all candidate texts are located, which build the foundation of subsequent stage for fine positioning and recognition. Experiments show that the scheme is competitive with current optimal solutions in both detection and recognition.