The space environment has become highly congested due to the increasing debris, seriously threatening safety of orbiting spacecraft. Space-based situational awareness, as a comprehensive capability threat knowledge, analysis, and decision-making, is significant importance ensure security maintain normal order. Various awareness systems have been designed launched. Data acquisition, target recognition, monitoring constituting key technologies make major contributions, various advanced...

Addressing the challenge that existing deep learning models face in accurately segmenting metal corrosion boundaries and small areas. In this paper, a SegFormer detection method based on parallel extraction of edge features is proposed. Firstly, to solve boundary ambiguity problem images, an edge-feature module (EEM) introduced construct spatial branch network assist model extracting shallow details information from images. Secondly, mitigate loss target feature during reconstruction...

Introduction The increasing integration of renewable energy sources, such as wind and solar, into power grids introduces significant challenges due to their inherent variability unpredictability. Traditional fossil-fuel-based systems are ill-equipped maintain stability cost-effectiveness in this evolving landscape. Methods This study presents a novel framework that integrates robust optimization with online learning dynamically manage uncertainties generation. Robust ensures system...

Document Processing Systems (DPSs) support office workers to manage information. classification is a major function of DPSs. By analyzing document’s layout and conceptual structures, we present in this paper sample-based approach document classification. We represent structure by an ordered labeled tree through procedure known as nested segmentation the set attribute type pairs. The similarities between be classified sample documents are determined previously developed approximate matching...

To improve the performance of acoustic logging tool in detecting three-dimensional formation, larger and more complicated transducer arrays have been used, which will greatly increase difficulty fault diagnosis during assembly maintenance. As a result, traditional passive diagnostic methods become inefficient, very skilled assemblers maintainers are required. In this study, fault-diagnosis requirement for at different levels has analyzed from perspective designer. An intelligent system...

We present a novel approach to register the visible and infrared power grid images in this paper. This method is based on coarse-to-fine structure for registration. First, we obtain feature points of multi-modality experience. Second, similarity geometric transformation model employed coarsely. At last, fine-grained built correct deviation coarse The experiment section shows good performance our proposed images.

As learning-based models become predominant in image processing field, especially classification and object detection, quality assessment (IQA) has been a hot issue since the of train images is great importance fro model performance. This paper proposes novel blind IQA evolved from VGG-16[1], state-of-the-art deep convolutional neural network(CNN). It modifies network to apply it regression task by keeping only one neuron last fully-connected layer, output which considered as metric after...

In order to ensure the safety of worksites staff, solve problem that worksite monitoring video transmission occupies many network resources, has high time delay and cannot detect staff not wearing helmets in real time, a method for detecting images based on cloud-edge cooperation is designed. The YOLOv3 algorithm used training helmet detection model. Using feature model can reduce processing latency, tasks such as optimization, which are computationally intensive have low real-time...

Transmitter source is a significant part of remote-exploration acoustic logging tool. Its power and directional-radiation characteristics are key factors that affect the tool performance. Plasma features high wide band. Thus, it expected to significantly improve exploration depth while maintaining resolution in logging. However, cannot generate directional field, which seriously undermines its advantages as source. Therefore, exploring suitable method for radiation plasma necessary. In this...

Uers organize networks and exchanges with the help of social networks, clasiffy friend to so-called circle friends.Construction such a will take user great deal time.This paper presents machine learning methods, he mutual networking between friends is as points clustering problem on user's personal network.Through relational model multi-point created by overlap we can analyze measure similarity user-specific information find hierarchical nested circles.In this paper, obtain real data from...

Owing to the limitations of real-time transmission, reliable downhole data storage and fast ground reading have become key technologies in developing tools for acoustic logging while drilling (LWD). In order improve reliability system conditions high temperature, intensive shake periodic power supply, improvements were made terms hardware software. hardware, we integrated acquisition control module into one circuit board, reduce complexity process, by adopting controller combination digital...

As learning-based models become predominant in image processing field, especially classification and object detection, quality assessment (IQA) has been a hot issue since the of train images is great importance fro model performance. This paper proposes novel blind IQA evolved from VGG-16 [1], state-of-the-art deep convolutional neural network(CNN). It modifies network to apply it regression task by keeping only one neuron last fully-connected layer, output which considered as metric after...

Cloud cover hinders the usability of optical remote sensing imagery. Existing cloud detection methods either require hand-crafted features or utilize deep networks. Generally, networks perform better than features. However, for need massive and expensive pixel-level annotation labels. To alleviate that, this paper proposes a weakly supervised learning-based method using only block-level labels, with new global convolutional pooling operation local pruning strategy to improve performance. For...

At present, target detection based on deep learning has become a trend. The large model in high accuracy, but with the huge network depth and width, making them difficult to deploy embedded systems limited hardware resources. To address this limitation, Firstly, we build feature extraction yolov5, CBAM (Convolutional Block Attention Module) attention structure are used improve accuracy. Finally, force iterative channel-level pruning guide Sparse training of BatchNormalization (BN) layers....

In order to better guarantee the operation effect of substation equipment,a remote fault diagnosis method equipment based on imagerecognition technology is proposed. Combined with image recognition tech-nology, running tracked and collected,the information characteristics are deeply excavated,and area accurately judged. Remotepositioning has been carried out realize accurate detection substa-tion fault. Finally, through experiment, diagnosismethod technologyis in actual application process...