基于卷积神经网络的输电杆塔缺陷智能识别系统研究与应用

Research and Application of Transmission Tower Defect Intelligent Recognition System Based on Convolution Neural Network

针对当前电力系统输电线路的巡检现状,本文设计了一套贴合实际生产应用的输电缺陷智能识别系统,实现了数据快速分类、图像智能识别、识别结果人工审核、报告生成等功能,并且对两阶段的检测关键技术展开了研究。经过实践应用发现,该系统能够有效提高巡视数据分析效率,为输电线路实现智能化提供有效的支撑。

Aiming at the current inspection status of power system transmission lines,this paper designs a set of transmission defect intelligent identification system suitable for actual production applications,which realizes the functions of rapid data classification,image intelligent identification,manual review of identification results,report generation,etc.The key technology of two-stage detection has been studied.Through practical application,it is found that the system can effectively improve the efficiency of patrol data analysis and provide effective support for the realization of intelligent transmission lines.