A multi-scale UAV aerial image object detection model MS-YOLOv7 based on YOLOv7 was proposed to address the issues of a large number of objects and a high proportion of small objects that commonly exist in the Unmanned Aerial Vehicle (UAV) aerial image. The new network is developed with a multiple detection head and a CBAM convolutional attention module to extract features at different scales. To solve the problem of high-density object detection, a YOLOv7 network architecture combined with the Swin Transformer units is proposed, and a new pyramidal pooling module, SPPFS is incorporated into the network. Finally, we incorporate the SoftNMS and the Mish activation function to improve the network’s ability to identify overlapping and occlusion objects. Various experiments on the open-source dataset VisDrone2019 reveal that our new model brings a significant performance boost compared to other state-of-the-art (SOTA) models. Compared with the YOLOv7 object detection algorithm of the baseline network, the $\mathrm{mAP}0.5$ of MS-YOLOv7 increased by 6.0%, the $\mathrm{mAP}0.95$ increased by 4.9%. Ablation experiments show that the designed modules can improve detection accuracy and visually display the detection effect in different scenarios. This experiment demonstrates the applicability of the MS-YOLOv7 for UAV aerial photograph object detection.

Image denoising has been a knotty issue in the computer vision field, although the developing deep learning technology has brought remarkable improvements in image denoising. Denoising networks based on deep learning technology still face some problems, such as in their accuracy and robustness. This paper constructs a robust denoising network based on a generative adversarial network (GAN). Since the neural network has the phenomena of gradient dispersion and feature disappearance, the global residual is added to the autoencoder in the generator network, to extract and learn the features of the input image, so as to ensure the stability of the network. On this basis, we proposed an optimization algorithm (OA), to train and optimize the mean and variance of noise on each node of the generator. Then the robustness of the denoising network was improved through back propagation. Experimental results showed that the model’s denoising effect is remarkable. The accuracy of the proposed model was over 99% in the MNIST data set and over 90% in the CIFAR10 data set. The peak signal to noise ratio (PSNR) and structural similarity (SSIM) values of the proposed model were better than the state-of-the-art models in the BDS500 data set. Moreover, an anti-interference test of the model showed that the defense capacities of both the fast gradient sign method (FGSM) and project gradient descent (PGD) attacks were significantly improved, with PSNR and SSIM values decreased by less than 2%.

Using deep learning methods to detect potential safety hazards in transmission lines is the mainstream method for power grid security monitoring. However, the existing model is too complex to adapt to edge device deployment and real-time detection. Therefore, an edge–real-time transmission line safety hazard detection method (ETLSH-YOLO) was proposed to reduce the model’s complexity and improve the model’s robustness. Firstly, a re-parameterized Ghost efficient layer aggregation network (RepGhostCSPELAN) was designed to effectively fuse the feature information of different layers while enhancing the model’s expression ability and reducing the number of model parameters and floating-point operations. Then, a spatial channel decoupled downsampling block (CSDovn) was designed to reduce computational redundancy and improve the computational efficiency of the model. Then, coordinate attention (CA) was added in the process of multi-scale feature fusion to suppress the interference of complex background and improve the global perception ability of the model object. Finally, the Mish activation function was used to improve the network’s training speed, convergence, and generalization ability. The experimental results show that the mAP50 of this model improved by 1.73% compared with the baseline model, and the number of parameters and floating-point operations were reduced by 33.96% and 22.22%, respectively. This model lays the foundation for solving the dilemma of edge device deployment.

Ice-covered transmission lines seriously affect the normal operation of the power transmission system. Resonance deicing based on different ice thicknesses is an effective method to solve the issue of ice-covered transmission lines. In order to obtain accurate ice thickness of transmission lines, this paper designs an ice thickness of transmission line recognition model based on Global Micro Strip Awareness Net (GMSA-Net) and proposes a Mixed Strip Convolution Module (MSCM) and a global micro awareness module (GMAM). The MSCM adapts to the shape of ice-covered transmission lines by using strip convolutions with different receptive fields, improving the encoder’s ability to extract ice-covered features; the GMAM perceives through both global and micro parts, mining the connections between semantic information. Finally, the ice thickness of the generated segmented image is calculated using the method of regional pixel statistics. Experiments are conducted on the dataset of ice-covered transmission lines. The mean Intersection over Union (mIoU) of image segmentation reaches 96.4%, the balanced F-Score (F1-Score) is 98.1%, and the identification error of ice thickness is within 3.8%. Experimental results prove that this method can accurately identify the ice thickness of transmission lines, providing a control basis for the application of resonant deicing engineering.

The recognition of ice thickness on transmission lines serves as a prerequisite for controlling de-icing robots to carry out precise de-icing operations. To address the issue that existing edge computing terminals fail to meet the demands of ice thickness recognition algorithms, this paper introduces an Efficient Edge Computing Network (EECNet) specifically designed for identifying ice thickness on transmission lines. Firstly, pruning is applied to the Efficient Neural Network (ENet), removing redundant components within the encoder to decrease both the computational complexity and the number of parameters in the model. Secondly, a Dilated Asymmetric Bottleneck Module (DABM) is proposed. By integrating different types of convolutions, this module effectively strengthens the model’s capability to extract features from ice-covered transmission lines. Then, an Efficient Partial Conv Module (EPCM) is designed, introducing an adaptive partial convolution selection mechanism that innovatively combines attention mechanisms with partial convolutions. This design enhances the model’s ability to select important feature channels. The method involves segmenting ice-covered images to obtain iced regions and then calculating the ice thickness using the iced area and known cable parameters. Experimental validation on an ice-covered transmission line dataset shows that EECNet achieves a segmentation accuracy of 92.7% in terms of the Mean Intersection over Union (mIoU) and an F1-Score of 96.2%, with an ice thickness recognition error below 3.4%. Compared to ENet, the model’s parameter count is reduced by 41.7%, and the detection speed on OrangePi 5 Pro is improved by 27.3%. After INT8 quantization, the detection speed is increased by 26.3%. These results demonstrate that EECNet not only enhances the recognition speed on edge equipment but also maintains high-precision ice thickness recognition.

Ice-covered transmission lines seriously affect the stable operation of the power system. Deploying a recognition network for measuring the ice thickness on transmission lines within a deicing robot, and controlling the robot to perform resonant deicing, is an effective solution. In order to solve the problem that the existing recognition network is not suitable for an edge device, an ice-thickness recognition network for transmission lines based on efficient dynamic perception (EDPNet) is proposed. Firstly, a lightweight multidimensional recombination convolution (LMRC) is designed to split the ordinary convolution for lightweight design and extract feature information of different scales for reorganization. Then, a lightweight deep fusion module (LDFM) is designed, which combines the attention mechanism with different features to enhance the information interaction between the encoder and decoder. Then, a new dynamic loss function is adopted in the training process to guide the model to perform refined detection of ice-covered boundaries. Finally, we count the ice pixels and calculate the ice thickness. The model is deployed on an OrangePi5 Plus edge computing board. Compared with the baseline model, the maximum ice-thickness detection error is 4.2%, the model parameters are reduced by 86.1%, and the detection speed is increased by 74.6%. Experimental results show that EDPNet can efficiently complete the task of identifying ice-covered transmission lines and has certain engineering application value.