基于You Only Look Once v2优化算法的车辆实时检测

针对基于You Only Look Once v2算法的目标检测存在精度低及稳健性差的问题,提出一种车辆目标实时检测的You Only Look Once v2优化算法;该算法以You Only Look Once v2算法为基础,通过增加网络深度,增强特征提取能力,同时,通过添加残差模块,解决网络深度增加带来的梯度消失或弥散问题;该方法将网络结构中低层特征与高层特征进行融合,提升对小目标车辆的检测精度。结果表明,通过在KITTI数据集上进行测试,优化后的算法在检测速度不变的情况下,提高了车辆目标检测精度,平均精度达到0.94,同时提升了小目标检测的准确性。

一种基于多尺度的目标检测锚点构造方法

目标检测是计算机视觉领域的研究热点和基础任务,其中基于锚点(Anchor)的目标检测已在众多领域得到广泛应用。当前锚点选取方法主要面临两个问题:基于特定数据集的先验取值尺寸固定、面对不同场景泛化能力弱。计算锚框的无监督K-means算法,受初始值影响较大,对目标尺寸较单一的数据集聚类产生的锚点差异较小,无法充分体现网络多尺度输出的特点。针对上述问题,本文提出一种基于多尺度的目标检测锚点构造方法(multi-scale-anchor,MSA),将聚类产生的锚点根据数据集本身的特性进行尺度的缩放和拉伸,优化的锚点即保留原数据集的特点也体现了模型多尺度的优势。另外,本方法应用在训练的预处理阶段,不增加模型推理时间。最后,选取单阶段主流算法YOLO(You Only Look Once),在多个不同场景的红外或工业场景数据集上进行丰富的实验。结果表明,多尺度锚点优化方法MSA能显著提高小样本场景的检测精度。

基于ID-YOLO的数字仪表检测方法

当前针对数字式仪表检测算法在边缘设备具有实时性差、泛化性差的问题,对此提出一种采用ID-YOLO(instrument detection-you only look once)模型的变电站数字仪表检测识别方法。所提算法以YOLOv5模型为基础,首先设计轻量骨干网络(light weight-YOLO, LW-YOLO)提取图像特征,降低网络参数,提高检测实时性;然后设计了一种双级路由注意力模块(bi-level routing attention moudle, BRAM),提高网络对小数点的检测精度以及网络的鲁棒性和泛化性;最后,引入损失函数α-IoU,通过设定不同的可调节参数α数值得到更准确的真实框与预测框的交并比计算,可以提高模型的检测精度。结果表明:相比于其他基于深度学习的数字仪表检测识别方法,所提方法在不同显示方式的数字仪表识别任务上具有更好的准确性和泛化性,而且可以在检测准确率领先的情况下,将模型在边缘设备上的检测速度从6.87帧/s提升至8.77帧/s,其实时性和检测精度均能够满足实际变电站智能数据采集、检测识别的工程需要。

基于深度学习YOLOX算法的混凝土构件裂缝智能化检测方法

针对现有混凝土构件裂缝人工检测操作不仅费时、费力,而且易出现错检、误检、漏检,以及部分位置难以开展检测的问题,提出一种基于深度学习YOLOX(You Only Look Once)算法的混凝土构件裂缝智能化检测方法;首先采集、整理包含各类混凝土构件的典型裂缝图像,并通过图像数据增强建立Pascal VOC数据集,然后基于Facebook公司开发的深度学习框架Pytorch,利用数据集训练YOLOX算法,并进行裂缝识别和验证;将训练完成后YOLOX算法移植至搭载安卓系统的手机端,进行现场实时检测操作。结果表明:在迭代次数为700时,混凝土构件裂缝识别精度可达88.84%,能有效筛分混凝土构件表面裂缝,并排除其他干扰项,证明了所提出的方法对裂缝具有较高的识别精度和广泛的适用性;经试验测试,移植至手机端的YOLOX算法能在提升便携性的同时保证高效、准确的检测效果,具有良好的应用前景。

Benchmarking YOLOv5 models for improved human detection in search and rescue missions

Drone or unmanned aerial vehicle(UAV)technology has undergone significant changes.The technology allows UAV to carry out a wide range of tasks with an increasing level of sophistication,since drones can cover a large area with cameras.Meanwhile,the increasing number of computer vision applications utilizing deep learning provides a unique insight into such applications.The primary target in UAV-based detection applications is humans,yet aerial recordings are not included in the massive datasets used to train object detectors,which makes it necessary to gather the model data from such platforms.You only look once(YOLO)version 4,RetinaNet,faster region-based convolutional neural network(R-CNN),and cascade R-CNN are several well-known detectors that have been studied in the past using a variety of datasets to replicate rescue scenes.Here,we used the search and rescue(SAR)dataset to train the you only look once version 5(YOLOv5)algorithm to validate its speed,accuracy,and low false detection rate.In comparison to YOLOv4 and R-CNN,the highest mean average accuracy of 96.9%is obtained by YOLOv5.For comparison,experimental findings utilizing the SAR and the human rescue imaging database on land(HERIDAL)datasets are presented.The results show that the YOLOv5-based approach is the most successful human detection model for SAR missions.

改进YOLO的遮挡行人检测仿真

针对已有的YOLO(you only look once)模型在行人目标检测中对遮挡及多尺度行人易造成漏检和精度较低的问题,提出改进YOLO行人检测算法YOLO-SSC-s(YOLO-spatial pyramid poolingshuffle attention-convolutional block attention module-simplified)。修改YOLO模型骨干网络,增强跨尺度特征提取能力;在3个YOLO层前的不同位置引入空间金字塔池化模块以及空间与通道、组特征2种注意力机制,加强对不同尺度行人的特征融合;为了缓解网络模型过于复杂而降低检测性能,提高模型训练效率,根据实际情况对网络结构进行简化。实验结果表明:与YOLOv3等检测模型相比,YOLO-SSC-s可有效提高遮挡情形下中、小行人目标的检测精度、速度,降低漏检率。

Intelligent identification of landslides in loess areas based on the improved YOLO algorithm:a case study of loess landslides in Baoji City

Loess landslides are one of the geological hazards prevalent in mountainous areas of Loess Plateau,seriously threatening people's lives and property safety.Accurate identification of landslides is a prerequisite for reducing the risk of landslide hazards.Traditional landslide interpretation methods often have the disadvantage of being laborious and difficult to use on a large scale compared with the recently developed deep learning-based landslide detection methods.In this study,we propose an improved deep learning model,landslide detectionyou only look once(LD-YOLO),based on the existing you only look once(YOLO)model for the intelligent identification of old and new landslides in loess areas.Specifically,remote sensing images of landslides in Baoji City,Shaanxi Province,China are acquired from the Google Earth Engine platform.The landslide images of Baoji City(excluding Qianyang County)are used to establish a loess landslide dataset for training the model.The landslide data of Qianyang County is used to verify the detection performance of the model.The focal and efficient IoU(Focal-EIoU)loss function and efficient channel attention(ECA)mechanism are incorporated into the 7th version of YOLO(YOLOv7)model to construct the LD-YOLO model,which makes it more suitable for the landslide detection task.The experiments yielded an improved LD-YOLO model with average precision of 92.05%,precision of 92.31%,recall of 90.28%,and F1-score of 91.28%for loess landslide detection.The landslides in Qianyang County were divided into two test sets,new landslides and old landslides,which were used to test the detection performance of LD-YOLO for both types of landslides.The results show that LD-YOLO detects old landslides with a detection precision of 82.75%and a recall of 80%.When detecting new landslides,the detection precision is 94.29%and the recall is 91.67%.It indicates that our proposed LD-YOLO model has strong detection performance for both new and old landslides in loess areas.Through a proposed solution that can realize the accurate detection of landslides in loess areas,this paper provides a valuable reference for the application of deep learning methods in landslide identification.

一种基于视觉识别的乒乓球捡球机设计与开发

针对目前乒乓球捡球机捡球机构不完善、乒乓球识别算法适应性差的问题,提出一种基于视觉识别的智能乒乓球捡球机.采用树莓派4B开发板作为控制单元,利用轻量化的you only look once(YOLO)v5s算法,对乒乓球进行识别;通过扇叶式集球机构,将乒乓球卷入收纳篮.实验结果表明:在乒乓球数小于150个的情况下,该捡球机的识别精确率与查全率均可达到95%以上,漏检率控制在7%以下.同时,集球机构结构简单、可靠、效率高,整体设计方案具有较好的实际应用价值.

An accurate detection algorithm for time backtracked projectile-induced water columns based on the improved YOLO network

During a sea firing training,the intelligent detection of projectile-induced water column targets in a firing video is the prerequisite for and critical to the automatic calculation of miss distance,while the correct and precise calculation of miss distance is directly affected by the accuracy,false alarm rate and time delay of detection.After analyzing the characteristics of projectile-induced water columns,an accurate detection algorithm for time backtracked projectile-induced water columns based on the improved you only look once(YOLO)network is put forward.The capability and accuracy of detecting projectileinduced water column targets with the conventional YOLO network are improved by optimizing the anchor box through K-means clustering and embedding the squeeze and excitation(SE)attention module.The detection area is limited by adopting a sea-sky line detection algorithm based on gray level co-occurrence matrix(GLCM),so as to effectively eliminate such disturbances as ocean waves and ship wakes,and lower the false alarm rate of projectile-induced water column detection.The improved algorithm increases the mAP50 of water column detection by 30.3%.On the basis of correct detection,a time backtracking algorithm is designed with mean shift to track images containing projectile-induced water column in reverse time sequence.It accurately detects a projectile-induced water column at the time of its initial appearance as well as its pixel position in images,and considerably reduces detection delay,so as to provide the support for the automatic,accurate,and real-time calculation of miss distance.

Face Mask and Social Distance Monitoring via Computer Vision and Deployable System Architecture

The coronavirus(COVID-19)is a lethal virus causing a rapidly infec-tious disease throughout the globe.Spreading awareness,taking preventive mea-sures,imposing strict restrictions on public gatherings,wearing facial masks,and maintaining safe social distancing have become crucial factors in keeping the virus at bay.Even though the world has spent a whole year preventing and curing the disease caused by the COVID-19 virus,the statistics show that the virus can cause an outbreak at any time on a large scale if thorough preventive measures are not maintained accordingly.Tofight the spread of this virus,technologically developed systems have become very useful.However,the implementation of an automatic,robust,continuous,and lightweight monitoring system that can be efficiently deployed on an embedded device still has not become prevalent in the mass community.This paper aims to develop an automatic system to simul-taneously detect social distance and face mask violation in real-time that has been deployed in an embedded system.A modified version of a convolutional neural network,the ResNet50 model,has been utilized to identify masked faces in peo-ple.You Only Look Once(YOLOv3)approach is applied for object detection and the DeepSORT technique is used to measure the social distance.The efficiency of the proposed model is tested on real-time video sequences taken from a video streaming source from an embedded system,Jetson Nano edge computing device,and smartphones,Android and iOS applications.Empirical results show that the implemented model can efficiently detect facial masks and social distance viola-tions with acceptable accuracy and precision scores.

Underwater Sea Cucumber Target Detection Based on Edge-Enhanced Scaling YOLOv4

Sea cucumber detection is widely recognized as the key to automatic culture.The underwater light environment is complex and easily obscured by mud,sand,reefs,and other underwater organisms.To date,research on sea cucumber detection has mostly concentrated on the distinction between prospective objects and the background.However,the key to proper distinction is the effective extraction of sea cucumber feature information.In this study,the edge-enhanced scaling You Only Look Once-v4(YOLOv4)(ESYv4)was proposed for sea cucumber detection.By emphasizing the target features in a way that reduced the impact of different hues and brightness values underwater on the misjudgment of sea cucumbers,a bidirectional cascade network(BDCN)was used to extract the overall edge greyscale image in the image and add up the original RGB image as the detected input.Meanwhile,the YOLOv4 model for backbone detection is scaled,and the number of parameters is reduced to 48%of the original number of parameters.Validation results of 783images indicated that the detection precision of positive sea cucumber samples reached 0.941.This improvement reflects that the algorithm is more effective to improve the edge feature information of the target.It thus contributes to the automatic multi-objective detection of underwater sea cucumbers.

Enhanced Deep Learning for Detecting Suspicious Fall Event in Video Data

Suspicious fall events are particularly significant hazards for the safety of patients and elders.Recently,suspicious fall event detection has become a robust research case in real-time monitoring.This paper aims to detect suspicious fall events during video monitoring of multiple people in different moving back-grounds in an indoor environment;it is further proposed to use a deep learning method known as Long Short Term Memory(LSTM)by introducing visual atten-tion-guided mechanism along with a bi-directional LSTM model.This method contributes essential information on the temporal and spatial locations of‘suspi-cious fall’events in learning the video frame in both forward and backward direc-tions.The effective“You only look once V4”(YOLO V4)–a real-time people detection system illustrates the detection of people in videos,followed by a track-ing module to get their trajectories.Convolutional Neural Network(CNN)fea-tures are extracted for each person tracked through bounding boxes.Subsequently,a visual attention-guided Bi-directional LSTM model is proposed for the final suspicious fall event detection.The proposed method is demonstrated using two different datasets to illustrate the efficiency.The proposed method is evaluated by comparing it with other state-of-the-art methods,showing that it achieves 96.9%accuracy,good performance,and robustness.Hence,it is accep-table to monitor and detect suspicious fall events.

Lira-YOLO: a lightweight model for ship detection in radar images

For the detection of marine ship objects in radar images, large-scale networks based on deep learning are difficult to be deployed on existing radar-equipped devices. This paper proposes a lightweight convolutional neural network, LiraNet, which combines the idea of dense connections, residual connections and group convolution, including stem blocks and extractor modules.The designed stem block uses a series of small convolutions to extract the input image features, and the extractor network adopts the designed two-way dense connection module, which further reduces the network operation complexity. Mounting LiraNet on the object detection framework Darknet, this paper proposes Lira-you only look once(Lira-YOLO), a lightweight model for ship detection in radar images, which can easily be deployed on the mobile devices. Lira-YOLO's prediction module uses a two-layer YOLO prediction layer and adds a residual module for better feature delivery. At the same time, in order to fully verify the performance of the model, mini-RD, a lightweight distance Doppler domain radar images dataset, is constructed. Experiments show that the network complexity of Lira-YOLO is low, being only 2.980 Bflops, and the parameter quantity is smaller, which is only 4.3 MB. The mean average precision(mAP) indicators on the mini-RD and SAR ship detection dataset(SSDD) reach 83.21% and 85.46%, respectively,which is comparable to the tiny-YOLOv3. Lira-YOLO has achieved a good detection accuracy with less memory and computational cost.

Manipulator-based autonomous inspections at road checkpoints:Application of faster YOLO for detecting large objects

With the increasing number of vehicles,manual security inspections are becoming more laborious at road checkpoints.To address it,a specialized Road Checkpoints Robot(RCRo)system is proposed,incorporated with enhanced You Only Look Once(YOLO)and a 6-degree-of-freedom(DOF)manipulator,for autonomous identity verification and vehicle inspection.The modified YOLO is characterized by large objects’sensitivity and faster detection speed,named“LF-YOLO”.The better sensitivity of large objects and the faster detection speed are achieved by means of the Dense module-based backbone network connecting two-scale detecting network,for object detection tasks,along with optimized anchor boxes and improved loss function.During the manipulator motion,Octree-aided motion control scheme is adopted for collision-free motion through Robot Operating System(ROS).The proposed LF-YOLO which utilizes continuous optimization strategy and residual technique provides a promising detector design,which has been found to be more effective during actual object detection,in terms of decreased average detection time by 68.25%and 60.60%,and increased average Intersection over Union(Io U)by 20.74%and6.79%compared to YOLOv3 and YOLOv4 through experiments.The comprehensive functional tests of RCRo system demonstrate the feasibility and competency of the multiple unmanned inspections in practice.

Weapons Detection for Security and Video Surveillance Using CNN and YOLO-V5s

In recent years,the number of Gun-related incidents has crossed over 250,000 per year and over 85%of the existing 1 billion firearms are in civilian hands,manual monitoring has not proven effective in detecting firearms.which is why an automated weapon detection system is needed.Various automated convolutional neural networks(CNN)weapon detection systems have been proposed in the past to generate good results.However,These techniques have high computation overhead and are slow to provide real-time detection which is essential for the weapon detection system.These models have a high rate of false negatives because they often fail to detect the guns due to the low quality and visibility issues of surveillance videos.This research work aims to minimize the rate of false negatives and false positives in weapon detection while keeping the speed of detection as a key parameter.The proposed framework is based on You Only Look Once(YOLO)and Area of Interest(AOI).Initially,themodels take pre-processed frames where the background is removed by the use of the Gaussian blur algorithm.The proposed architecture will be assessed through various performance parameters such as False Negative,False Positive,precision,recall rate,and F1 score.The results of this research work make it clear that due to YOLO-v5s high recall rate and speed of detection are achieved.Speed reached 0.010 s per frame compared to the 0.17 s of the Faster R-CNN.It is promising to be used in the field of security and weapon detection.

图像识别技术在矿用钢丝绳检测中的应用

为了解决架空乘人装置钢丝绳在长期使用中经常出现断丝、锈蚀、翘起等问题,根据图像识别技术,提出基于YOLO v3算法的智能钢丝绳检测方法,设计了一种探伤系统,能够实时探测钢丝绳损伤状况,并实现准确定位。YOLO v3利用Darknet-53作为特征提取网络,并引入了残差块,解决了因网络深度增加而引起的梯度消失问题。对比其它无损检测方法,该系统检测速度更快,达到66.8FPS,AP值达到91.03%,且大幅度提高了抗干扰能力,能够满足诸多环境下的生产要求。

Development of Mobile App to Support the Mobility of Visually Impaired People

In 2017,it was estimated that the number of persons of all ages visually affected would be two hundred and eighty-five million,of which thirty-nine million are blind.There are several innovative technical solutions available to facilitate the movement of these people.The next big challenge for technical people is to give cost-effective solutions.One of the challenges for people with visual impairments is navigating safely,recognizing obstacles,and moving freely between locations in unfamiliar environments.A new mobile application solution is developed,and the application can be installed in android mobile.The application will visualize the environment with portable cameras and persons with visual impairment directly to the environment.The designed system mainly uses the YOLO3 program to identify and locate the distance between the objects and the camera.Furthermore,it determines the direction of the object.Finally,the system will give the voice command to teach/inform the visually impaired people to navigate the environment.It is a novel work at the global level.The proposed approach is cost-effective and affordable to all strata of society.

Identification and Classification of Crowd Activities

The identification and classification of collective people’s activities are gaining momentum as significant themes in machine learning,with many potential applications emerging.The need for representation of collective human behavior is especially crucial in applications such as assessing security conditions and preventing crowd congestion.This paper investigates the capability of deep neural network(DNN)algorithms to achieve our carefully engineered pipeline for crowd analysis.It includes three principal stages that cover crowd analysis challenges.First,individual’s detection is represented using the You Only Look Once(YOLO)model for human detection and Kalman filter for multiple human tracking;Second,the density map and crowd counting of a certain location are generated using bounding boxes from a human detector;and Finally,in order to classify normal or abnormal crowds,individual activities are identified with pose estimation.The proposed system successfully achieves designing an effective collective representation of the crowd given the individuals in addition to introducing a significant change of crowd in terms of activities change.Experimental results onMOT20 and SDHA datasets demonstrate that the proposed system is robust and efficient.The framework achieves an improved performance of recognition and detection peoplewith a mean average precision of 99.0%,a real-time speed of 0.6ms non-maximumsuppression(NMS)per image for the SDHAdataset,and 95.3%mean average precision for MOT20 with 1.5ms NMS per image.

A Multi-Mode Public Transportation System Using Vehicular to Network Architecture

The number of accidents in the campus of Suranaree University of Technology(SUT)has increased due to increasing number of personal vehicles.In this paper,we focus on the development of public transportation system using Intelligent Transportation System(ITS)along with the limitation of personal vehicles using sharing economy model.The SUT Smart Transit is utilized as a major public transportation system,while MoreSai@SUT(electric motorcycle services)is a minor public transportation system in this work.They are called Multi-Mode Transportation system as a combination.Moreover,a Vehicle toNetwork(V2N)is used for developing theMulti-Mode Transportation system in the campus.Due to equipping vehicles with On Board Unit(OBU)and 4G LTE modules,the real time speed and locations are transmitted to the cloud.The data is then applied in the proposed mathematical model for the estimation of Estimated Time of Arrival(ETA).In terms of vehicle classifications and counts,we deployed CCTV cameras,and the recorded videos are analyzed by using You Only Look Once(YOLO)algorithm.The simulation and measurement results of SUT Smart Transit and MoreSai@SUT before the covid-19 pandemic are discussed.Contrary to the existing researches,the proposed system is implemented in the real environment.The final results unveil the attractiveness and satisfaction of users.Also,due to the proposed system,the CO_(2) gas gets reduced when Multi-Mode Transportation is implemented practically in the campus.

Power Plant Indicator Light Detection System Based on Improved YOLOv5

Electricity plays a vital role in daily life and economic development.The status of the indicator lights of the power plant needs to be checked regularly to ensure the normal supply of electricity.Aiming at the problem of a large amount of data and different sizes of indicator light detection,we propose an improved You Only Look Once vision 5(YOLOv5)power plant indicator light detection algorithm.The algorithm improves the feature extraction ability based on YOLOv5s.First,our algorithm enhances the ability of the network to perceive small objects by combining attention modules for multi-scale feature extraction.Second,we adjust the loss function to ensure the stability of the object frame during the regression process and improve the conver-gence accuracy.Finally,transfer learning is used to augment the dataset to improve the robustness of the algorithm.The experimental results show that the average accuracy of the proposed squeeze-and-excitation YOLOv5s(SE-YOLOv5s)algorithm is increased by 4.39%to 95.31%compared with the YOLOv5s algorithm.The proposed algorithm can better meet the engineering needs of power plant indicator light detection.