To address the problems of low network accuracy, slow speed, and a large number model parameters in printed circuit board (PCB) defect detection, an improved detection algorithm PCB surface defects based on YOLOv5 is proposed, named PCB-YOLO, this paper. Based K-means++ algorithm, more suitable anchors for dataset are obtained, small target layer added to make PCB-YOLO pay attention information. Swin transformer embedded into backbone network, united mechanism constructed reduce interference...

At present, UAV aerial photography has a good prospect in agricultural production, disaster response, and other aspects. The application of UAVs can greatly improve work efficiency decision-making accuracy. However, owing to inherent features such as wide field view large differences the target scale images, this lead existing detection algorithms missing small targets or causing incorrect detections. To solve these problems, paper proposes algorithm for based on improved YOLOv5s. Firstly,...

Semantic segmentation plays an indispensable role in automatic analysis of remote sensing image data. However, the abundant semantic information and irregular shape patterns images are difficult to utilize, making it hard segment only using convolution single-scale feature maps. To achieve better performance, a multiscale pyramid decoder (MFPD) is proposed fuse features extracted by vision transformer (ViT). The employs novel 2-D-to-3-D transform method obtain maps that contain rich context...

Unmanned Aerial Vehicles (UAVs), due to their flexible deployment, are used as a three-dimensional space assistant tool for Vehicular Edge Computing (VEC) cover moving vehicles. However, existing work generally assumes relatively uniform vehicle distribution while the actual road conditions vary over time. The time-varying location of vehicles and congestion in peak hours pose challenges vehicular edge computing. First, traffic can lead imbalanced UAVs load, that is covering congested areas...

This study developed a computer-aided detection (CAD) scheme for pulmonary embolism (PE) and investigated several approaches to improve CAD performance. In the study, 20 computed tomography examinations with various lung diseases were selected, which include 44 verified PE lesions. The proposed consists of five basic steps: 1) segmentation; 2) candidate extraction using an intensity mask tobogganing region growing; 3) feature extraction; 4) false-positive (FP) reduction artificial neural...

The excessive discharge of phosphorus-containing wastewater contributes to eutrophication, posing a serious threat aquatic ecosystems. Therefore, methods such as electrocoagulation should be utilized remove phosphorus from prior discharging it into water body. In this study, we aimed determine the effectiveness in treating simulated under different parameters, including anode material (aluminum, iron, and magnesium), electrode distance (ED) (1, 2.5, 4.5 cm), pH (3, 6, 9), current density...

Smoking behavior in public places and fire bans seriously threatens the safety of people's lives property. In order to ensure safety, this paper proposes a smoking detection method based on YOLOv5 algorithm image processing. Aiming at small target cigarettes, uses K-means adding layer improve algorithm, realizes improvement accuracy. On self-made data set, false rate is 0%, AP 92.3%, which 6.7% higher than that YOLOv5s algorithm.

Artificial neural networks (ANN) have been widely used in computer-assisted diagnosis (CAD) schemes as a classification tool to identify abnormalities digitized mammograms. Because of certain limitations ANNs, some investigators argue that Bayesian belief network (BBN) may exhibit higher performance. In this study we compared the performance an ANN and BBN same CAD scheme. The common databases genetic algorithm (GA) were optimize both networks. experimental results demonstrated using GA...

Surface defects, which often occur during the production of aluminum profiles, can directly affect quality and should be monitored in real time. This paper proposes an effective, lightweight detection method for profiles to realize real-time surface defect with ensured accuracy. Based on YOLOv5s framework, a network model is designed by adding attention mechanism depth-separable convolution aluminum. The improves limitations framework regarding its accuracy speed. backbone GCANet built based...

The ETER701 trial demonstrated that benmelstobart combined with anlotinib and etoposide-carboplatin (EC) significantly extends survival in patients extensive-stage small cell lung cancer (ES-SCLC), setting a new record for median overall survival. In contrast, plus EC only prolongs progression-free However, there is currently no evidence evaluating the cost-effectiveness of these regimens as first-line treatments. Therefore, this study assesses three treatment options from perspective...

Vehicle re-identification (re-id) is an important issue in the transportation and vehicle tracking area. The existing FastReID framework has made significant improvements data preprocessing, model structure parameter configuration of re-id. However, did not consider directional differences between images, which makes it difficult to balance image pairs with different during calculating similarity vehicles. To make a further improvement based on FastReID, this paper proposes Orientation...

AbstractUsing stochastic resonance (SR) mechanism, the output signal can be enhanced by adding noise to nonlinear system. Therefore, an image denoising algorithm based on adaptive bi-dimensional (ABSR) is proposed in this paper. Firstly, sampled as a signal, and dynamic system model constructed. The peak signal-to-noise ratio (PSNR) structural similarity (SSIM) of are used double evaluation system, optimal parameters automatically obtained adjusting using reverse positioning method. Compared...

Intelligent Building spaces are critical in creating smarter living spaces, which composed of interconnected Internet Things (IoT) devices. The growing dependence on these advanced, connected environments has led to several operational challenges related privacy, security, and scalability due their distributed architectures the complex interplay among numerous IoT systems. Recently, Blockchain Technology emerged as a solution for addressing privacy security intelligent buildings. However,...

It is difficult to extract the lane lines accurately in current auxiliary driving system due complexity of environment. In this paper, a new detection method which provides an improved accuracy proposed. Firstly, deep learning network Unet adopted get potential lines. Secondly, Canny edge and Hough transform are used fit vanishing point. Thirdly, position point segment region interest (ROI). Finally, slope relationship between front back frames video select The experimental results show...