Unstructured Road Extraction in UAV Images based on Lightweight Model

There is no unified planning standard for unstructured roads,and the morphological structures of these roads are complex and varied.It is important to maintain a balance between accuracy and speed for unstructured road extraction models.Unstructured road extraction algorithms based on deep learning have problems such as high model complexity,high computational cost,and the inability to adapt to current edge computing devices.Therefore,it is best to use lightweight network models.Considering the need for lightweight models and the characteristics of unstructured roads with different pattern shapes,such as blocks and strips,a TMB(Triple Multi-Block)feature extraction module is proposed,and the overall structure of the TMBNet network is described.The TMB module was compared with SS-nbt,Non-bottleneck-1D,and other modules via experiments.The feasibility and effectiveness of the TMB module design were proven through experiments and visualizations.The comparison experiment,using multiple convolution kernel categories,proved that the TMB module can improve the segmentation accuracy of the network.The comparison with different semantic segmentation networks demonstrates that the TMBNet network has advantages in terms of unstructured road extraction.

Ultra-lightweight CNN design based on neural architecture search and knowledge distillation: A novel method to build the automatic recognition model of space target ISAR images

In this paper,a novel method of ultra-lightweight convolution neural network(CNN)design based on neural architecture search(NAS)and knowledge distillation(KD)is proposed.It can realize the automatic construction of the space target inverse synthetic aperture radar(ISAR)image recognition model with ultra-lightweight and high accuracy.This method introduces the NAS method into the radar image recognition for the first time,which solves the time-consuming and labor-consuming problems in the artificial design of the space target ISAR image automatic recognition model(STIIARM).On this basis,the NAS model’s knowledge is transferred to the student model with lower computational complexity by the flow of the solution procedure(FSP)distillation method.Thus,the decline of recognition accuracy caused by the direct compression of model structural parameters can be effectively avoided,and the ultralightweight STIIARM can be obtained.In the method,the Inverted Linear Bottleneck(ILB)and Inverted Residual Block(IRB)are firstly taken as each block’s basic structure in CNN.And the expansion ratio,output filter size,number of IRBs,and convolution kernel size are set as the search parameters to construct a hierarchical decomposition search space.Then,the recognition accuracy and computational complexity are taken as the objective function and constraint conditions,respectively,and the global optimization model of the CNN architecture search is established.Next,the simulated annealing(SA)algorithm is used as the search strategy to search out the lightweight and high accuracy STIIARM directly.After that,based on the three principles of similar block structure,the same corresponding channel number,and the minimum computational complexity,the more lightweight student model is designed,and the FSP matrix pairing between the NAS model and student model is completed.Finally,by minimizing the loss between the FSP matrix pairs of the NAS model and student model,the student model’s weight adjustment is completed.Thus the ultra-lightweight and high accuracy STIIARM is obtained.The proposed method’s effectiveness is verified by the simulation experiments on the ISAR image dataset of five types of space targets.

High performance“non-local”generic face reconstruction model using the lightweight Speckle-Transformer(SpT)UNet

Significant progress has been made in computational imaging(CI),in which deep convolutional neural networks(CNNs)have demonstrated that sparse speckle patterns can be reconstructed.However,due to the limited“local”kernel size of the convolutional operator,for the spatially dense patterns,such as the generic face images,the performance of CNNs is limited.Here,we propose a“non-local”model,termed the Speckle-Transformer(SpT)UNet,for speckle feature extraction of generic face images.It is worth noting that the lightweight SpT UNet reveals a high efficiency and strong comparative performance with Pearson Correlation Coefficient(PCC),and structural similarity measure(SSIM)exceeding 0.989,and 0.950,respectively.

Survey of lightweighting methods of huge 3D models for online Web3D visualization

Background With the rapid development of Web3D technologies, the online Web3D visualization, particularly for complex models or scenes, has been in a great demand. Owing to the major conflict between the Web3D system load and resource consumption in the processing of these huge models, the huge 3D model lightweighting methods for online Web3D visualization are reviewed in this paper. Methods By observing the geometry redundancy introduced by man-made operations in the modeling procedure, several categories of light-weighting related work that aim at reducing the amount of data and resource consumption are elaborated for Web3D visualization. Results By comparing perspectives, the characteristics of each method are summarized, and among the reviewed methods, the geometric redundancy removal that achieves the lightweight goal by detecting and removing the repeated components is an appropriate method for current online Web3D visualization. Meanwhile, the learning algorithm, still in improvement period at present, is our expected future research topic. Conclusions Various aspects should be considered in an efficient lightweight method for online Web3D visualization, such as characteristics of original data, combination or extension of existing methods, scheduling strategy, cache man-agement, and rendering mechanism. Meanwhile, innovation methods, particularly the learning algorithm, are worth exploring.

Brittleness Generation Mechanism and Failure Model of High Strength Lightweight Aggregate Concrete

The brittleness generation mechanism of high strength lightweight aggregate con-crete(HSLWAC) was presented, and it was indicated that lightweight aggregate was the vulnerable spot, initiating brittleness. Based on the analysis of the brittleness failure by the load-deflection curve, the brittleness presented by HSLWAC was more prominent compared with ordinary lightweight aggregate concrete of the same strength grade. The model of brittleness failure was also established.

Cephalopods Classification Using Fine Tuned Lightweight Transfer Learning Models

Cephalopods identification is a formidable task that involves hand inspection and close observation by a malacologist.Manual observation and iden-tification take time and are always contingent on the involvement of experts.A system is proposed to alleviate this challenge that uses transfer learning techni-ques to classify the cephalopods automatically.In the proposed method,only the Lightweight pre-trained networks are chosen to enable IoT in the task of cephalopod recognition.First,the efficiency of the chosen models is determined by evaluating their performance and comparing thefindings.Second,the models arefine-tuned by adding dense layers and tweaking hyperparameters to improve the classification of accuracy.The models also employ a well-tuned Rectified Adam optimizer to increase the accuracy rates.Third,Adam with Gradient Cen-tralisation(RAdamGC)is proposed and used infine-tuned models to reduce the training time.The framework enables an Internet of Things(IoT)or embedded device to perform the classification tasks by embedding a suitable lightweight pre-trained network.Thefine-tuned models,MobileNetV2,InceptionV3,and NASNet Mobile have achieved a classification accuracy of 89.74%,87.12%,and 89.74%,respectively.Thefindings have indicated that thefine-tuned models can classify different kinds of cephalopods.The results have also demonstrated that there is a significant reduction in the training time with RAdamGC.

Simulation of Fracture Process of Lightweight Aggregate Concrete Based on Digital Image Processing Technology

The mechanical properties and failure mechanism of lightweight aggregate concrete(LWAC)is a hot topic in the engineering field,and the relationship between its microstructure and macroscopic mechanical properties is also a frontier research topic in the academic field.In this study,the image processing technology is used to establish a micro-structure model of lightweight aggregate concrete.Through the information extraction and processing of the section image of actual light aggregate concrete specimens,the mesostructural model of light aggregate concrete with real aggregate characteristics is established.The numerical simulation of uniaxial tensile test,uniaxial compression test and three-point bending test of lightweight aggregate concrete are carried out using a new finite element method-the base force element method respectively.Firstly,the image processing technology is used to produce beam specimens,uniaxial compression specimens and uniaxial tensile specimens of light aggregate concrete,which can better simulate the aggregate shape and random distribution of real light aggregate concrete.Secondly,the three-point bending test is numerically simulated.Thirdly,the uniaxial compression specimen generated by image processing technology is numerically simulated.Fourth,the uniaxial tensile specimen generated by image processing technology is numerically simulated.The mechanical behavior and damage mode of the specimen during loading were analyzed.The results of numerical simulation are compared and analyzed with those of relevant experiments.The feasibility and correctness of the micromodel established in this study for analyzing the micromechanics of lightweight aggregate concrete materials are verified.Image processing technology has a broad application prospect in the field of concrete mesoscopic damage analysis.

Research on Track Fastener Service Status Detection Based on Improved Yolov4 Model

As an important part of railway lines, the healthy service status of track fasteners was very important to ensure the safety of trains. The application of deep learning algorithms was becoming an important method to realize its state detection. However, there was often a deficiency that the detection accuracy and calculation speed of model were difficult to balance, when the traditional deep learning model is used to detect the service state of track fasteners. Targeting this issue, an improved Yolov4 model for detecting the service status of track fasteners was proposed. Firstly, the Mixup data augmentation technology was introduced into Yolov4 model to enhance the generalization ability of model. Secondly, the MobileNet-V2 lightweight network was employed in lieu of the CSPDarknet53 network as the backbone, thereby reducing the number of algorithm parameters and improving the model’s computational efficiency. Finally, the SE attention mechanism was incorporated to boost the importance of rail fastener identification by emphasizing relevant image features, ensuring that the network’s focus was primarily on the fasteners being inspected. The algorithm achieved both high precision and high speed operation of the rail fastener service state detection, while realizing the lightweight of model. The experimental results revealed that, the MAP value of the rail fastener service state detection algorithm based on the improved Yolov4 model reaches 83.2%, which is 2.83% higher than that of the traditional Yolov4 model, and the calculation speed was improved by 67.39%. Compared with the traditional Yolov4 model, the proposed method achieved the collaborative optimization of detection accuracy and calculation speed.

光固化快速成型的轻质AGARD-B模型气动特性实验研究

传统金属制风洞模型质量大,设计加工周期长、成本高,为此笔者提出了采用一种光固化快速成型技术,设计加工轻质风洞模型的方法,并对比分析了引起光固化快速成型的轻质和金属AGARD-B模型气动特性间差异的原因。结果表明,在跨声速范围,轻质模型同金属模型的气动特性基本吻合,光固化快速成型技术的高速风洞模型设计方法基本可行。

A Novel Deep Neural Network Compression Model for Airport Object Detection

A novel deep neural network compression model for airport object detection has been presented.This novel model aims at disadvantages of deep neural network,i.e.the complexity of the model and the great cost of calculation.According to the requirement of airport object detection,the model obtains temporal and spatial semantic rules from the uncompressed model.These spatial semantic rules are added to the model after parameter compression to assist the detection.The rules can improve the accuracy of the detection model in order to make up for the loss caused by parameter compression.The experiments show that the effect of the novel compression detection model is no worse than that of the uncompressed original model.Even some of the original model false detection can be eliminated through the prior knowledge.

Identification of Weather Phenomena Based on Lightweight Convolutional Neural Networks

Weather phenomenon recognition plays an important role in the field of meteorology.Nowadays,weather radars and weathers sensor have been widely used for weather recognition.However,given the high cost in deploying and maintaining the devices,it is difficult to apply them to intensive weather phenomenon recognition.Moreover,advanced machine learning models such as Convolutional Neural Networks(CNNs)have shown a lot of promise in meteorology,but these models also require intensive computation and large memory,which make it difficult to use them in reality.In practice,lightweight models are often used to solve such problems.However,lightweight models often result in significant performance losses.To this end,after taking a deep dive into a large number of lightweight models and summarizing their shortcomings,we propose a novel lightweight CNNs model which is constructed based on new building blocks.The experimental results show that the model proposed in this paper has comparable performance with the mainstream non-lightweight model while also saving 25 times of memory consumption.Such memory reduction is even better than that of existing lightweight models.

Preparation and properties of geopolymer-lightweight aggregate refractory concrete

Geopolymer-lightweight aggregate refractory concrete (GLARC) was prepared with geopolymer and lightweight aggregate. The mechanical property and heat-resistance (950 ℃) of GLARC were investigated. The effects of size of aggregate and mass ratio of geopolymer to aggregate on mechanical and thermal properties were also studied. The results show that the highest compressive strength of the heated refractory concrete is 43.3 MPa,and the strength loss is only 42%. The mechanical property and heat-resistance are influenced by the thickness of geopolymer covered with aggregate,which can be expressed as the quantity of geopolymer on per surface area of aggregate. In order to show the relationship between the thickness of geopolymer covered with aggregate and the thermal property of concrete,equal thickness model is presented,which provides a reference for the mix design of GLARC. For the haydite sand with size of 1.18-4.75 mm,the best amount of geopolymer per surface area of aggregate should be in the range of 0.300-0.500 mg/mm2.

BIM-based construction technologies for precast foamed lightweight concrete wallboards

To promote the visualisation and informatisation of the construction process of precast foamed lightweight concrete wallboards(PFLCWs),from the analysis of the construction requirements of PFLCWs,three key construction technologies based on building information modelling(BIM),namely,parameterised modelling for the PFLCW layout design,drawing generation to draw the PFLCW layout and quantity statistics for extracting PFLCW quantities,are proposed.Then,a reinforced concrete(RC)frame infilled with PFLCW is considered the test model to verify the feasibility of the aforementioned technologies.The results show that PFLCW layout design can be accomplished rapidly and visually using parameterised modelling technology.The PFLCW layout diagram can be generated directly using drawing generation technology.The proposed quantity statistics technology enables the automatic export of PFLCW bills of quantities.The built parameterised model helps construction workers rapidly and intuitively understand the specific layout details of PFLCWs.Moreover,the generated layout drawing and the bills of quantities based on the parameterised model can guide the production and on-site installation of PFLCWs.The research conclusions can serve as a practical guide and technical support for PFLCW engineering applications.

Prediction of three-dimensional elastic behavior of filament-wound composites based on the bridging model

This work provides a method to predict the three-dimensional equivalent elastic properties of the filament-wound composites based on the multi-scale homogenization principle.In the meso-scale,a representative volume element(RVE)is defined and the bridging model is adopted to establish a theoretical predictive model for its three-dimensional equivalent elastic constants.The results obtained through this method for the previous experimental model are compared with the ones gained respectively by experiments and classical laminate theory to verify the reliability of this model.In addition,the effects of some winding parameters,such as winding angle,on the equivalent elastic behavior of the filament-wound composites are analyzed.The rules gained can provide a theoretical reference for the optimum design of filament-wound composites.

Thermal Infrared Salient Human Detection Model Combined with Thermal Features in Airport Terminal

Target detection in low light background is one of the main tasks of night patrol robots for airport terminal.However,if some algorithms can run on a robot platform with limited computing resources,it is difficult for these algorithms to ensure the detection accuracy of human body in the airport terminal. A novel thermal infrared salient human detection model combined with thermal features called TFSHD is proposed. The TFSHD model is still based on U-Net,but the decoder module structure and model lightweight have been redesigned. In order to improve the detection accuracy of the algorithm in complex scenes,a fusion module composed of thermal branch and saliency branch is added to the decoder of the TFSHD model. Furthermore,a predictive loss function that is more sensitive to high temperature regions of the image is designed. Additionally,for the sake of reducing the computing resource requirements of the algorithm,a model lightweight scheme that includes simplifying the encoder network structure and controlling the number of decoder channels is adopted. The experimental results on four data sets show that the proposed method can not only ensure high detection accuracy and robustness of the algorithm,but also meet the needs of real-time detection of patrol robots with detection speed above 40 f/s.

Sensitivity Analysis of a Simplified Fire Dynamic Model

This paper discusses a method for performing a sensitivity analysis of parameters used in a simplified fire model for temperature estimates in the upper smoke layer during a fire. The results from the sensitivity analysis can be used when individual parameters affecting fire safety are assessed. If the variation of a single parameter is found to have a major impact on fire safety, it may be necessary to conservatively select this parameter in order to incorporate additional safety. We compare fire scenarios in rooms surrounded by lightweight as well as heavy walls in order to investigate which parameters are the most significant in each case. We apply the Sobol method, which is a quantitative method that gives the percentage of the total output variance that each parameter accounts for. The most important parameter is found to be the energy release rate that explains 92% of the uncertainty in the calculated results for the period before thermal penetration （te） has occurred. The analysis is also done for all combinations of two parameters in order to find the combination with the largest effect. The Sobol total for pairs had the highest value for the combination of energy release rate and area of opening, which explains 96% of the uncertainty. After thermal penetration, the energy release rate is still the most important parameter, but now only explains 49% of the variation. The second parameter is the thickness of the surface material, which explains 43%.

基于改进YOLOv5s的轻量级绝缘子缺失检测

针对现有绝缘子缺失检测模型计算复杂度高和小目标难以检测等问题,提出一种基于改进的YOLOv5s轻量级检测模型。首先,移除主干网络中的C3模块来减少模型的参数量。其次,在多尺度特征融合网络中引入卷积块注意力机制来提高复杂背景下模型的特征提取能力。同时,采用加权双向特征金字塔网络结构对特征进行双向跨尺度加权融合,提升网络在遮挡物、相似目标干扰下目标的检测性能。最后,选用SIoU损失函数提升网络的收敛速度和检测精度。实验结果表明,所提模型的平均精准率为96.8%,浮点运算数为2.8 GFLOPS,而原始YOLOv5s在保证97.4%的平均精准率下的浮点运算数为16.3 GFLOPS。相较于原始模型,所提模型对小目标、遮挡目标以及模糊等场景有着较强的鲁棒性,且在保证近似检测精度的同时极大减少了计算量。

IDD-YOLOv7:一种用于输电线路绝缘子多缺陷的轻量化检测方法

YOLO目标检测算法是当前基于图像的输电线路绝缘子缺陷检测的主流方法,然而现有模型复杂度较大,亟需合理有效的参数压缩方法作为前提条件,来为解决无人机边缘设备部署的困境问题奠定基础;同时,无人机航拍的绝缘子缺陷图像背景复杂、缺陷尺寸较小,容易出现误检、漏检等问题。为此,提出了一种用于输电线路绝缘子多缺陷检测的Insulator Defect Detection-YOLOv7(IDD-YOLOv7)模型,以降低模型复杂度,提高模型鲁棒性。首先,在多尺度特征融合的过程中加入坐标注意力(Coordinate Attention)机制,抑制复杂背景的干扰,提升模型对小目标的全局感知能力;之后,设计C3GhostNetV2模块,用于捕获不同空间像素之间的远程依赖性,在增强模型表达能力的同时降低模型的参数量和浮点运算量;最后,提出Focal-CIoU损失函数,提高模型高质量anchor的贡献,加快模型的收敛速度。实验结果表明,本文方法与基线模型相比mAP^(50)提升了3.8%,查准率和召回率分别提升了1.7%和7.6%,参数量和浮点运算量分别下降了18.3%和14.0%,绝缘子自爆、破损、闪络缺陷的AP^(50)分别提升了0.8%、4.5%、6.3%。

动车组应急蓄电池箱的多目标优化

针对动车组的应急蓄电池箱的安全问题和轻量化设计要求,综合多个优化目标对其进行优化分析。将主要部件的厚度作为设计变量,以应急蓄电池箱的总质量和和恶劣工况下的应力最小为优化目标,以其第1阶固有频率为约束函数,使用Box-Behnken设计方法获取样本数据。利用样本数据建立低阶多项式响应面模型,结合第三代非支配排序遗传算法(NSGA-Ⅲ)进行多目标优化。结果表明:相较于单一的响应面法或遗传算法,本文采用的响应面法与遗传算法结合的方式,使得优化后的参数更加合理,轻量化和安全性均得到了保障。

结合轻量化与多尺度融合的交通标志检测算法

交通标志检测在自动驾驶领域具有重要的应用价值,及时准确地检测交通目标对提高驾驶安全性和预防交通事故具有重要意义。针对交通标志尺寸小,易受遮挡,在复杂环境下容易出现漏检、错检等问题,在YOLOv8的结构基础上提出一种结合轻量化与多尺度融合的交通标志检测网络架构M-YOLO,构建M-YOLOs模型来应对高精度需求的检测任务,并调整网络深度得到更轻量化的M-YOLOn模型来解决不同环境下的检测需求。首先针对交通标志目标尺寸小、图像特征流失的问题,通过增加小目标检测层,保留更多的特征信息,提高网络对于小目标的特征学习能力。提出高效多尺度特征金字塔融合网络MPANet,将浅层特征图进行降维与跳跃连接,从而融合更多的图像特征信息。然后提出融合稀疏注意力和空间注意力的BRSA注意力模块,有效提取全局和局部的位置信息,减少复杂背景下对于关键信息的干扰。最后设计两种轻量高效的BBot模块和C2fGhost模块,以提高模型运算速度并减少参数量。实验结果表明,M-YOLO相较于YOLOv8,参数量降低约1/3。在TT100K数据集和GTSDB数据集上,M-YOLOs检测精度分别提升了9.7和2.1个百分点,M-YOLOn检测精度分别提升了14.5和2.6个百分点,在轻量化的同时具备更高的检测效果。M-YOLO架构解决了浅层特征图在特征提取过程中信息丢失的问题,并显著降低模型特征提取过程中冗余的计算开销,在实景采集的数据集上证实效果有效,表明在交通标志检测任务中具有应用价值。