摘要
针对高压杆塔的安全性易受鸟巢等异物影响的情况,提出基于经典深度学习理论的Fast R-CNN算法,实现对异物的快速检测,降低安全风险。该算法的基本思路是,通过Selective Search法提取杆塔图像候选区域,并基于CaffeNet网络模型优化参数,经过多次迭代和样本训练,最后智能检测出杆塔图像中的鸟巢并定位目标区域。
In order to reduce the damage caused by the bird's nest to the safety of high-voltage transmission lines,a Fast R-CNN algorithm is proposed based on the classic deep learning,which can realize fast and efficient foreign object detection.The selective search method algorithm is used to extract the candidate area of the tower image,and the parameters are optimized based on the CaffeNet network model.Finally,the bird's nest foreign matters in the tower image are intelligently detected,and the target area with significant results is located after different iterations and training samples adjustment.
作者
师飘
张超
郑祥明
SHI Piao;ZHANG Chao;ZHENG Xiangming(Department of Electronics and Information Engineering,Bozhou University,Bozhou Anhui 236800,China)
出处
《重庆科技学院学报（自然科学版）》
CAS
2020年第2期83-87,共5页
Journal of Chongqing University of Science and Technology：Natural Sciences Edition
基金
安徽省高校优秀青年人才支持计划项目(GXYQ2017109)
安徽省高校自然科学研究项目“基于敏捷开发的高校实习实训APP建设”(KJ2017A703)
亳州学院“嵌入式系统开发与应用”创客实验室项目(2017CKSY02)。
