计算机应用 ›› 2019, Vol. 39 ›› Issue (9): 2535-2540.DOI: 10.11772/j.issn.1001-9081.2019030462

• 人工智能 • 上一篇    下一篇

基于RCF的精细边缘检测模型

景年昭, 杨维   

  1. 北京交通大学 电子信息工程学院, 北京 100044
  • 收稿日期:2019-03-20 修回日期:2019-04-11 出版日期:2019-09-10 发布日期:2019-04-19
  • 通讯作者: 景年昭
  • 作者简介:景年昭(1994-),男,山东济南人,硕士研究生,主要研究方向:计算机视觉、信息处理;杨维(1964-),男,北京人,教授,博士生导师,博士,主要研究方向:无线通信、信息处理。
  • 基金资助:

    国家重点研发计划项目(2016YFC0801800)。

Fine edge detection model based on RCF

JING Nianzhao, YANG Wei   

  1. School of Electronic and Information Engineering, Beijing Jiaotong University, Beijing 100044, China
  • Received:2019-03-20 Revised:2019-04-11 Online:2019-09-10 Published:2019-04-19
  • Supported by:

    This work is partially supported by the National Key Research and Development Program of China (2016YFC0801800).

摘要:

针对目前基于深度学习的边缘检测技术生成的边缘粗糙及模糊等问题,提出一种基于更丰富特征的边缘检测(RCF)模型的端到端的精细边缘检测模型。该模型以RCF模型为基础,在主干网络中引入"注意力"机制,采用SE模块提取图像边缘特征,并且去掉主干网络部分下采样,避免细节信息过度丢失,使用扩张卷积技术增大模型感受野,并利用残差结构将不同尺度的边缘图进行融合。对伯克利分割数据集(BSDS500)进行增强,使用一种多步骤的训练方式在BSDS500和PASCAL VOC Context数据集上进行训练,并用BSDS500进行测试实验。实验结果表明,该模型将全局最佳(ODS)和单图最佳(OIS)指标分别提高到了0.817和0.838,在不影响实时性的前提下可以输出更精细的边缘,同时还具有较好的鲁棒性。

关键词: 边缘检测, 更丰富的卷积特征检测, 深度学习, 扩张卷积, 注意力机制

Abstract:

Aiming at the roughness and blur of edges generated by edge detection technology based on deep learning, an end-to-end fine edge detection model based on RCF (Richer Convolutional Features for edge detection) was proposed. In this model based on RCF model, attention mechanism was introduced in the backbone network, Squeeze-and-Excitation (SE) module was used to extract image edge features. In order to avoid excessive loss of detail information, two subsampling in the backbone network were removed. In order to increase the receptive field of the model, dilation convolution was used in the backbone. A residual module was used to fuse the edge images in different scales. The model was trained on the Berkeley Segmentation Data Set (BSDS500)and PASCAL VOC Context dataset by a multi-step training approach and was tested on the BSDS500. The experimental results show that the model improves the ODS (Optimal Dataset Scale) and OIS (Optimal Image Scale) to 0.817 and 0.838 respectively, and it not only generates finer edges without affecting real-time performance but also has better robustness.

Key words: edge detection, Richer Convolutional Features for edge detection (RCF), deep learning, dilation convolution, attention mechanism

中图分类号: