# Integrating Frequency-Spatial Features for Energy-Efficient OPGW Target Recognition in UAV-Assisted Mobile Monitoring

## TL;DR

This paper introduces OPGW-DETR, an energy-efficient system for detecting OPGW cables during UAV inspections, improving accuracy and communication reliability in power grid monitoring.

## Contribution

The novel OPGW-DETR model integrates MC-GAP and a hybrid gating mechanism for low-power, real-time detection of OPGW cables in UAV inspections.

## Key findings

- The S-scale OPGW-DETR model achieves a 3.9% improvement in average precision (AP) over the baseline.
- The model shows a 2.5% improvement in AP50, reducing misidentification risks during UAV inspections.
- OPGW-DETR enables reliable detection under UAV battery and bandwidth constraints, ensuring uninterrupted grid monitoring.

## Abstract

Optical Fiber Composite Overhead Ground Wire (OPGW) cables serve dual functions in power systems, lightning protection and critical communication infrastructure for real-time grid monitoring. Accurate OPGW identification during UAV inspections is essential to prevent miscuts and maintain power-communication functionality. However, detecting small, twisted OPGW segments among visually similar ground wires is challenging, particularly given the computational and energy constraints of edge-based UAV platforms. We propose OPGW-DETR, a lightweight detector based on the D-FINE framework, optimized for low-power operation to enable reliable detection. The model incorporates two key innovations: multi-scale convolutional global average pooling (MC-GAP), which fuses spatial features across multiple receptive fields and integrates spectrally motivated features for enhanced fine-grained representation, and a hybrid gating mechanism that dynamically balances global and spatial features while preserving original information through residual connections. By enabling real-time inference with minimal energy consumption, OPGW-DETR addresses UAV battery and bandwidth limitations while ensuring continuous detection capability. Evaluated on a custom OPGW dataset, the S-scale model achieves 3.9% improvement in average precision (AP) and 2.5% improvement in AP50 over the baseline. By mitigating misidentification risks, these gains improve communication reliability. As a result, uninterrupted grid monitoring becomes feasible in low-power UAV inspection scenarios, where accurate detection is essential to ensure communication integrity and safeguard the power grid.