پیوندهای مفید
اخبار و اعلانات
آمار
| تعداد نشریات | 127 |
| تعداد شمارهها | 7,140 |
| تعداد مقالات | 76,848 |
| تعداد مشاهده مقاله | 154,510,444 |
| تعداد دریافت فایل اصل مقاله | 116,556,259 |
A YOLO-Based Approach for Small-Object Defect Detection in High-Voltage Transmission Towers | ||
| Earth Observation and Geomatics Engineering | ||
| دوره 9، شماره 1، شهریور 2025، صفحه 117-125 اصل مقاله (688.01 K) | ||
| نوع مقاله: Original Article | ||
| شناسه دیجیتال (DOI): 10.22059/eoge.2026.409788.1204 | ||
| نویسندگان | ||
| mohammad shahriari Naeini؛ Ali Abzal* ؛ Jamal Asgari؛ Hossein Bagheri | ||
| University of Isfahan | ||
| چکیده | ||
| Reliable transmission of electrical energy through high-voltage power towers requires continuous and accurate inspection of their components and potential defects. This research aims to facilitate the transition from traditional, labor-intensive inspection methods to automated inspection systems. Existing image datasets often fail to comprehensively cover all components and defects of power towers, and certain critical parts are overlooked. For example, due to their small size, essential elements such as split pins and bolts have not received sufficient attention. In this study, images captured by drones from a 400 kV transmission line were utilized. The goal was to move toward automated inspection by annotating key components in two classes—intact and damaged—and training object detection models from the YOLO family on the generated dataset. The performance of the trained models was evaluated. The best model achieved a precision of 0.754 for classification and a mean Intersection over Union (mIoU) of 0.740 for object localization. | ||
| کلیدواژهها | ||
| Object Detection؛ YOLO؛ High-Voltage Power Towers؛ Deep Learning؛ Automated Inspection | ||
|
آمار تعداد مشاهده مقاله: 49 تعداد دریافت فایل اصل مقاله: 26 |
||
سامانه مدیریت نشریات علمی. قدرت گرفته از سیناوب