Photovoltaic equipment segmentation using YOLOv8 and RepViT from a UAV perspective

doi:10.13474/j.cnki.11-2246.2026.0915

Abstract

Abstract: [Purposes] To address the challenges of small-scale and densely distributed rooftop photovoltaic (PV)modules in high-altitude UAV imagery,this study proposes an efficient,automated method for precise PV identification and area estimation to support regional-scale monitoring of distributed PV assets.[Methods] An improved instance segmentation model is developed based on YOLOv8-seg by replacing the original backbone with a lightweight RepViT architecture,utilizing structural re-parameterization to enhance global feature modeling and representation efficiency.Furthermore,a multi-stage feature fusion strategy is employed in the neck network to strengthen the perception of small-scale modules,and a mask-based physical area estimation model is constructed by integrating ground sampling distance(GSD).[Findings] Experiments conducted on UAV datasets from Zengcheng district,Guangzhou,show that the improved model achieves a Box mAP₅₀ of 95.4% and a Mask mAP₅₀ of 94.5%,representing precision increases of 3.4% and 2.9% respectively compared to the baseline.In large-scale practical tests,the model successfully achieved accurate multi-scale photovoltaic area calculation spanning three orders of magnitude, with errors meeting the requirements of engineering applications.[Conclusions] The proposed method effectively reduces parameter count while significantly improving segmentation accuracy and inference speed,providing a high-precision,automated technical solution for the census and fine-grained management of distributed rooftop solar resources．

Key words: UAV remote sensing, rooftop photovoltaic, instance segmentation, YOLOv8, RepViT, area estimation

CLC Number:

P237

Zhong Wenping, Zhang Xinchang, Huang Yu. Photovoltaic equipment segmentation using YOLOv8 and RepViT from a UAV perspective[J]. Bulletin of Surveying and Mapping, 2026, 0(9): 101-107.

Add to citation manager EndNote|Reference Manager|ProCite|BibTeX|RefWorks

URL: http://tb.chinasmp.com/EN/10.13474/j.cnki.11-2246.2026.0915

http://tb.chinasmp.com/EN/Y2026/V0/I9/101

References

[1] Mallapaty S.How China could be carbon neutral by mid-century[J].Nature, 2020, 586(7830): 482-483.
[2] Liu Zhu, Deng Zhu, He Gang, et al.Challenges and opportunities for carbon neutrality in China[J].Nature Reviews Earth & Environment, 2022, 3(2): 141-155.
[3] Li Jianglong, Huang Jiashun.The expansion of China's solar energy:challenges and policy options[J].Renewable and Sustainable Energy Reviews, 2020, 132: 110002.
[4] Wang Mudan, Mao Xianqiang, Gao Yubing, et al.Potential of carbon emission reduction and financial feasibility of urban rooftop photovoltaic power generation in Beijing[J].Journal of Cleaner Production, 2018, 203: 1119-1131.
[5] Hijjawi U, Lakshminarayana S, Xu T H, et al.A review of automated solar photovoltaic defect detection systems:approaches, challenges, and future orientations[J].Solar Energy, 2023, 266: 112186.
[6] 赵永强, 饶元, 董世鹏, 等.深度学习目标检测方法综述[J].中国图象图形学报, 2020, 25(4): 629-654.
[7] Su Jinya, Yi Dewei, Su Baofeng, et al.Aerial visual perception in smart farming: field study of wheat yellow rust monitoring[J].IEEE Transactions on Industrial Informatics, 2021, 17(3): 2242-2249.
[8] 魏文豪, 衷诚, 郭健辰, 等.基于改进YOLOv11的光伏场区施工进度检测[J].测绘通报, 2025(7): 132-137.
[9] 赵效祖, 苟长龙, 杨扬.基于CBAM增强的轻量级遥感影像语义分割方法[J].测绘通报, 2025(10): 36-42.
[10] Wang Ao, Chen Hui, Lin Zijia, et al.Rep ViT: revisiting mobile CNN from ViT perspective[C]//Proceedings of 2024 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Seattle: IEEE, 2024: 15909-15920.
[11] Ding Xiaohan, Zhang Xiangyu, Ma Ningning, et al.RepVGG: making VGG-style ConvNets great again[C]//Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).Nashville: IEEE, 2021: 13728-13737.
[12] Lin T Y, Dollar P, Girshick R, et al.Feature pyramid networks for object detection[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition(CVPR).Honolulu:IEEE, 2017: 936-944.
[13] Wang C Y, Mark Liao H Y, Wu Y H, et al.CSPNet: a new backbone that can enhance learning capability of CNN[C]//Proceedings of 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).Seattle: IEEE, 2020: 1571-1580.

[1]	Zhang Huiran, Bai Zihan, Gao Jianwei, Zhang Yu, Wu Hui. Individual extraction of pole-like objects for combining mobile vehicle-borne imagery and point clouds [J]. Bulletin of Surveying and Mapping, 2026, 0(7): 124-129.
[2]	Qiu Jun, Ou Weilin, Zhang Yunsheng, Li Yuanzhi. Application of FPN-YOLOv9 with channel-spatial attention fusion for river pollution detection [J]. Bulletin of Surveying and Mapping, 2026, 0(7): 142-148.
[3]	TENG Min, ZHANG Bo, XU Jiawei, LIN Cong, SHEN Yu, CHU Zhengwei. Vehicle object detection approach in drone imagery based on improved YOLOv8s [J]. Bulletin of Surveying and Mapping, 2026, 0(2): 68-73,80.
[4]	YIN Junjie, GUAN Daiwanjing, LI Hao, ZHANG Xiaoyang, MA Yujie, XING Hanfa. UAV-based object recognition dataset for coastal sewage outfalls [J]. Bulletin of Surveying and Mapping, 2025, 0(8): 112-117.
[5]	SUN Jiyuan, JI Song, GAO Ding, LI Kai, ZHANG Ruiying. SFR-YOLO: small target detection algorithm for UAV imagery based on improved YOLOv8 [J]. Bulletin of Surveying and Mapping, 2025, 0(7): 32-39.
[6]	XU Jinghai, LIU Yang, QI Mengxuan, JING Haoran. Optimized layout of rooftop photovoltaic panels driven by UAV real-scene 3D model [J]. Bulletin of Surveying and Mapping, 2025, 0(12): 1-6.
[7]	WANG Chao, FU Qiang, CUI Zhifang, TANG Tian. Remote sensing feature identification of river channel by fusing multimodal data [J]. Bulletin of Surveying and Mapping, 2025, 0(12): 178-183.
[8]	SUN Bin, ZHANG Ruiling, REN Chenfeng, LIN Yunhao, SUN Chao, LIU Yihan, LIU Mengjie, YUAN Debao, XU Zhihua. An instance segmentation method for mining-induced ground fissures in shallow coal seams using UAV imagery [J]. Bulletin of Surveying and Mapping, 2025, 0(11): 78-83.
[9]	LAN Guiwen, XU Zirui, REN Xinyue, ZHONG Zhan, GUO Ruidong, FAN Donglin. An improved algorithm for detecting components of power transmission lines from aerial inspection images [J]. Bulletin of Surveying and Mapping, 2024, 0(9): 38-43,49.
[10]	GUO Baoyun, YAO Yukai, LI Cailin, WANG Yue, SUN Na, LU Yihui. Application of improved 3D-BoNet to segmentation and 3D reconstruction of point cloud instances [J]. Bulletin of Surveying and Mapping, 2024, 0(6): 30-35.
[11]	CHEN Guangliang, XIE Yunguang, LIU Yuqi, CAI Yuezhen, CHEN Dingan. A screening method of redundant aerial photographs for oblique photogrammetry based on optical geometry and building feature recognition [J]. Bulletin of Surveying and Mapping, 2024, 0(4): 140-144.
[12]	LI Yuting, YUAN Zhenchao, ZHANG Li. Traffic signs detection algorithm based on improved YOLOv8 [J]. Bulletin of Surveying and Mapping, 2024, 0(12): 106-110.
[13]	CHEN Zongjie, YANG Wunian, ZHANG Wanlin, LIU Mingxin, DONG Xiujun, HE Xianding. Evaluation of forest flammability based on inversion data of vegetation canopy water content [J]. Bulletin of Surveying and Mapping, 2023, 0(4): 99-105.
[14]	SONG Weidong, BI Chunyang, ZHAO Fengshu. Improved Mask R-CNN highway disease detection algorithm [J]. Bulletin of Surveying and Mapping, 2023, 0(1): 58-64.
[15]	SHI Yun, SHI Longlong, NIU Minjie, ZHAO Kan. Multi-task automatic identification of loess landslide based on one-stage instance segmentation network [J]. Bulletin of Surveying and Mapping, 2022, 0(4): 26-31.