arXiv Computer Vision· Sampath Balaji Masabathula (Eastern Power Distribution Company of Andhra Pradesh Limited, Andhra University College of Engineering, Visakhapatnam, India), Srija Vanapalli (Eastern Power Distribution Company of Andhra Pradesh Limited, Andhra University College of Engineering, Visakhapatnam, India), Gangadhar Kanchi (Eastern Power Distribution Company of Andhra Pradesh Limited), Sreenivasa Rao V. A. N (Eastern Power Distribution Company of Andhra Pradesh Limited), Vijaya Lalitha V (Eastern Power Distribution Company of Andhra Pradesh Limited), Sasank Chilamkurthy (Von Neumann AI)·· 1 天前AI 评分32
基于智能手机的配电基础设施现场 AI 巡检系统
Smartphone-Based, Field-Deployed AI Inspection System for Power Distribution Infrastructure
AI 导读
一套基于智能手机的 AI 巡检系统在 APEPDCL 五条配电馈线试点部署，识别出 1222 处缺陷，其中 84 处在研究期内完成整改。该系统结合深度学习、单目深度估计与几何分析，可评估电杆倾斜、横担与顶夹倾斜、电压等级、植被净空风险及配电变压器对地净空。研究同时公开了现场采集数据集 PD-Defect，但现有观测不足以证明该系统带来可靠性提升。
来源：arXiv Computer Vision · arxiv.org