郑州大学学报（理学版）2019，Vol.51Issue(4)：16-22,7.DOI:10.13705/j.issn.1671-6841.2018299
航拍图像中绝缘子串的识别与分割方法研究
Research on Recognition and Segmentation of Insulator Strings in Aerial Images
摘要
Abstract
In order to realize the identification and segmentation of insulator strings in aerial images, a deep learning algorithm was proposed by using Faster R-CNN and FCN as the core framework to form an image detection platform. With the combination of the Faster R-CNN and the Resnet-101, the regions of interest were classified, and then Bounding Box regression and coordinate correction were made in order to achieve insulator strings detection. The FCN was fine-tunned on, to adapt to the new-insulator dataset, which could realize the segmentation of the insulator strings in complex background. Compared with the existing methods, the method presented not only could realize the identification and segmentation of insu-lator strings in different illumination conditions, different shooting angles and complex background inter-ference, but also could have short processing time, high precision and strong robustness.关键词
绝缘子串/Faster R-CNN/FCN/深度学习/绝缘子串识别/语义分割Key words
insulator strings/Faster R-CNN/FCN/deep learning/insulator strings detection/semantic segmentation分类
信息技术与安全科学引用本文复制引用
高金峰,吕易航..航拍图像中绝缘子串的识别与分割方法研究[J].郑州大学学报（理学版）,2019,51(4):16-22,7.基金项目
国家自然科学基金项目(61876169). （61876169）
