28 March 2025 in IEEE Transactions on Instrumentation and Measurement
In the process of utilizing thermal infrared sensors to identify photovoltaic (PV) hot-spot faults, due to the diverse shapes of hot-spot faults and environmental interferences, fault characteristics cannot be effectively expressed, which poses a challenge for traditional detection networks in achieving accurate detection. Consequently, an adaptive PV hot-spot fault detection network based on region perception and cross-channel feature aggregation is proposed. First, to reduce the interference of pseudo-highlight features in complex backgrounds, a Lite-UNet segmentation network is designed to remove background redundant information and enable the detection network to concentrate on the region of the PV panel. Second, to fully capture the geometric deformations and weak edge characteristics of hot-spot faults, a C2f DCN module is designed, which enhances the feature extraction capabilities by adaptively adjusting the receptive field. Subsequently, to address the problem of feature coupling often encountered in extracting dense hot-spot fault features, a dense object visual enhancement (DOVE) module was proposed. Through the dynamic aggregation of cross-space features, the feature information of different channels is integrated to improve the detection accuracy of hot-spot faults. Additionally, to further enhance the effective fusion of multi-scale features, an adaptive scale converter module (ASCM) detection head is designed. Finally, seven traditional detection methods are chosen for comparison in order to confirm the benefits of the proposed method. According to the experimental results, the suggested method can detect hot-spot faults with an accuracy of up to 86.5% in complicated inspection environments, which is an increase of 5.6% compared with the basis of original network.
5 June 2025 in Measurement Science and Technology
In coal mine belt conveying systems, foreign objects falling may cause safety risks, such as conveyor belt tearing or blockage. To address these challenges in complex underground environments, a foreign object detection method for coal conveyor belts based on brightness self-balancing and multi-scale feature fusion called BS-MFFNet is introduced. First, to mitigate the impact of poor lighting and coal dust interference, an environmental perception dynamic enhancer is designed, thus improving image quality. Second, to enhance the detection accuracy of irregular targets, deformable convolutions are integrated into the backbone network. Subsequently, to further enhance feature representation capabilities, a multi-scale feature fusion network is developed, integrating both global and local information to provide richer target features. Lastly, an adaptive kernel-aware attention module is incorporated into the detection head to enhance target representation in complex, low-illumination backgrounds. Experimental results show that BS-MFFNet outperforms eight classical algorithms, achieving an average accuracy of 97.4%, an score of 0.93, and a detection speed of 74.6 fps in environments with weak lighting and high coal dust concentrations.
2 September 2025 in IEEE Transactions on Instrumentation and Measurement
Insulator parallel gaps, as critical lightning protection components in high-voltage transmission lines, are prone to faults such as short circuits and excessive spacing, which compromise line safety and power supply reliability. However, detecting insulator parallel gaps from unmanned aerial vehicle (UAV) captured images is challenged by complex backgrounds, structural distortion, and occlusion. Thus, a hierarchical detection method incorporating lightweight localization and attention mechanism based diagnosis (KLSD-HDet) is proposed, which uses a localization network to capture the faults and then conducts fault diagnosis. First, to accurately capture target objects, a lightweight fault localization network (KDeFus-LNet) is designed, with nonlinear feature extraction and target localization capabilities in complex backgrounds enhanced. Second, leveraging the spatial geometric information from three-dimensional fault models, a multi-view fault image generation method is developed to compensate for the missing feature representations of partial viewpoints in real-world datasets. Then, a cross-space learning and multi-scale residual-based fault diagnosis network (S-CR2-DNet) is proposed to improve multi-view fault representation understanding and diagnostic accuracy. Finally, knowledge distillation is employed to lightweight S-CR2-DNet, enhancing its practical applicability. Extensive experiments validate that KLSD-HDet outperforms state-of-the-art methods, achieving 94.49% detection precision, improved by 6.65% compared to SOTAs.
17 March 2026 in Measurement Science and Technology
To address the challenge of multi-scale fault targets that compromise detection accuracy in complex environments, this paper proposes MGBFNet, a Multi-Scale Fault Detection Network for Transmission Lines based on Gated Mechanism and Bidirectional Feature Fusion. First, to address the issue of low accuracy in multi-scale object detection under complex environmental interference, gated multi-order aggregation module (GMAM) is proposed, which enhances multi-scale feature perception through parallel multi-order convolutions and adaptive gated aggregation mechanisms. Secondly, to address the problem of target occlusion arising from shooting angle changes, a bidirectional information exchange and fusion module (BIEFM) is desiged. Through hierarchical bidirectional information exchange between dual-stream networks, complementary global structural features are dynamically extracted, thereby enhancing feature fusion performance. Thirdly, to alleviate the problem of low accuracy caused by unbalanced sample numbers, the Slide loss function is incorporated to dynamically balance the weights of simple and difficult samples. Finally, to rigorously evaluate the effectiveness of the proposed approach, a dataset encompassing fault samples from twelve distinct categories was constructed, and extensive comparative experiments were performed against ten well-established detection methods. According to the experimental results, the suggested method can detect multi-scale fault targets on power transmission lines with an accuracy of up to 95.4\% in complex environments, which is an increase of 4.6\% compared with YOLOv8 and 4.1\% compared with YOLOv11.
1 January 2026 in IEEE Transactions on Geoscience and Remote Sensing
1 January 2026 in IEEE Transactions on Instrumentation and Measurement