RESUMEN
Fire incidents near power transmission lines pose significant safety hazards to the regular operation of the power system. Therefore, achieving fast and accurate smoke detection around power transmission lines is crucial. Due to the complexity and variability of smoke scenarios, existing smoke detection models suffer from low detection accuracy and slow detection speed. This paper proposes an improved model for smoke detection in high-voltage power transmission lines based on the improved YOLOv7-tiny. First, we construct a dataset for smoke detection in high-voltage power transmission lines. Due to the limited number of real samples, we employ a particle system to randomly generate smoke and composite it into randomly selected real scenes, effectively expanding the dataset with high quality. Next, we introduce multiple parameter-free attention modules into the YOLOv7-tiny model and replace regular convolutions in the Neck of the model with Spd-Conv (Space-to-depth Conv) to improve detection accuracy and speed. Finally, we utilize the synthesized smoke dataset as the source domain for model transfer learning. We pre-train the improved model and fine-tune it on a dataset consisting of real scenarios. Experimental results demonstrate that the proposed improved YOLOv7-tiny model achieves a 2.61% increase in mean Average Precision (mAP) for smoke detection on power transmission lines compared to the original model. The precision is improved by 2.26%, and the recall is improved by 7.25%. Compared to other object detection models, the smoke detection proposed in this paper achieves high detection accuracy and speed. Our model also improved detection accuracy on the already publicly available wildfire smoke dataset Figlib (Fire Ignition Library).
RESUMEN
Accurate cloud detection is an important step to improve the utilization rate of remote sensing (RS). However, existing cloud detection algorithms have difficulty in identifying edge clouds and broken clouds. Therefore, based on the channel data of the Himawari-8 satellite, this work proposes a method that combines the feature enhancement module with the Gaussian mixture model (GMM). First, statistical analysis using the probability density functions (PDFs) of spectral data from clouds and underlying surface pixels was conducted, selecting cluster features suitable for daytime and nighttime. Then, in this work, the Laplacian operator is introduced to enhance the spectral features of cloud edges and broken clouds. Additionally, enhanced spectral features are input into the debugged GMM model for cloud detection. Validation against visual interpretation shows promising consistency, with the proposed algorithm outperforming other methods such as RF, KNN and GMM in accuracy metrics, demonstrating its potential for high-precision cloud detection in RS images.
RESUMEN
With the widespread adoption of service-oriented architectures (SOA), services with the same functionality but the different Quality of Service (QoS) are proliferating, which is challenging the ability of users to build high-quality services. It is often costly for users to evaluate the QoS of all feasible services; therefore, it is necessary to investigate QoS prediction algorithms to help users find services that meet their needs. In this paper, we propose a QoS prediction algorithm called the MFDK model, which is able to fill in historical sparse QoS values by a non-negative matrix decomposition algorithm and predict future QoS values by a deep neural network. In addition, this model uses a Kalman filter algorithm to correct the model prediction values with real-time QoS observations to reduce its prediction error. Through extensive simulation experiments on the WS-DREAM dataset, we analytically validate that the MFDK model has better prediction accuracy compared to the baseline model, and it can maintain good prediction results under different tensor densities and observation densities. We further demonstrate the rationality of our proposed model and its prediction performance through model ablation experiments and parameter tuning experiments.
