The advancement of autonomous driving heavily relies on the ability to accurate lane lines detection.As deep learning and computer vision technologies evolve,a variety of deep learning-based methods for lane line dete...The advancement of autonomous driving heavily relies on the ability to accurate lane lines detection.As deep learning and computer vision technologies evolve,a variety of deep learning-based methods for lane line detection have been proposed by researchers in the field.However,owing to the simple appearance of lane lines and the lack of distinctive features,it is easy for other objects with similar local appearances to interfere with the process of detecting lane lines.The precision of lane line detection is limited by the unpredictable quantity and diversity of lane lines.To address the aforementioned challenges,we propose a novel deep learning approach for lane line detection.This method leverages the Swin Transformer in conjunction with LaneNet(called ST-LaneNet).The experience results showed that the true positive detection rate can reach 97.53%for easy lanes and 96.83%for difficult lanes(such as scenes with severe occlusion and extreme lighting conditions),which can better accomplish the objective of detecting lane lines.In 1000 detection samples,the average detection accuracy can reach 97.83%,the average inference time per image can reach 17.8 ms,and the average number of frames per second can reach 64.8 Hz.The programming scripts and associated models for this project can be accessed openly at the following GitHub repository:https://github.com/Duane 711/Lane-line-detec tion-ST-LaneNet.展开更多
A new efftcient straight line detection algorithm, GPI （ Gray Projecting Integral） method is proposed. The gray values of a sub-window are projected onto a line, and sum the gray values which are projected onto one ...A new efftcient straight line detection algorithm, GPI （ Gray Projecting Integral） method is proposed. The gray values of a sub-window are projected onto a line, and sum the gray values which are projected onto one same point to shape a special vector, then rotate the projecting direction, obtain many such vectors corresponding to different projecting directions. The vectors can form a matrix, a GPI matrix of the sub-image. The problem of lines detection is converted into maxima or minima searching problem in the GPI matrix. Bused on the GPI matrix, the lines can be calculated. Different from traditional methods, the algorithm can detect the positions of lines accurately, quickly without previous edge detection, which costs less time, and avoids the error resulted from the poor threshold with traditional methods. This algorithm is useful and efftcient for numerous image understanding applications and robot visual navigation, especially for welded joint position detection in heavy noise.展开更多
Straight line detection is a fundamental problem in target recognition from remote sensing images since many man-made objects have straight boundaries.In this study,an integrated straight line detection method for rem...Straight line detection is a fundamental problem in target recognition from remote sensing images since many man-made objects have straight boundaries.In this study,an integrated straight line detection method for remote sensing images is proposed.In this method,the edge-based straight lines are extracted using a chain code tracing method and the phasebased straight lines are extracted using a phase grouping method.The two types of lines are combined using a rule-based feature fusion method by removing redundant line extraction.Since this method integrates the specialties of edge-and phase-based straight line detection methods,it can detect straight lines from remote sensing images with high correctness and robustness.展开更多
Accurate perception of lane line information is one of the basic requirements of unmanned driving technology, which is related to the localization of the vehicle and the determination of the forward direction. In this...Accurate perception of lane line information is one of the basic requirements of unmanned driving technology, which is related to the localization of the vehicle and the determination of the forward direction. In this paper, multi-level constraints are added to the lane line detection model PINet, which is used to improve the perception of lane lines. Predicted lane lines in the network are predicted to have real and imaginary attributes, which are used to enhance the perception of features around the lane lines, with pixel-level constraints on the lane lines;images are converted to bird’s-eye views, where the parallelism between lane lines is reconstructed, with lane line-level constraints on the predicted lane lines;and vanishing points are used to focus on the image hierarchy, with image-level constraints on the lane lines. The model proposed in this paper meets both accuracy (96.44%) and real-time (30 + FPS) requirements, has been tested on the highway on the ground, and has performed stably.展开更多
In smart driving for rail transit,a reliable obstacle detection system is an important guarantee for the safety of trains.Therein,the detection of the rail area directly affects the accuracy of the system to identify ...In smart driving for rail transit,a reliable obstacle detection system is an important guarantee for the safety of trains.Therein,the detection of the rail area directly affects the accuracy of the system to identify dangerous targets.Both the rail line and the lane are presented as thin line shapes in the image,but the rail scene is more complex,and the color of the rail line is more difficult to distinguish from the background.By comparison,there are already many deep learning-based lane detection algorithms,but there is a lack of public datasets and targeted deep learning detection algorithms for rail line detection.To address this,this paper constructs a rail image dataset RailwayLine and labels the rail line for the training and testing of models.This dataset contains rich rail images including single-rail,multi-rail,straight rail,curved rail,crossing rails,occlusion,blur,and different lighting conditions.To address the problem of the lack of deep learning-based rail line detection algorithms,we improve the CLRNet algorithm which has an excellent performance in lane detection,and propose the CLRNet-R algorithm for rail line detection.To address the problem of the rail line being thin and occupying fewer pixels in the image,making it difficult to distinguish from complex backgrounds,we introduce an attention mechanism to enhance global feature extraction ability and add a semantic segmentation head to enhance the features of the rail region by the binary probability of rail lines.To address the poor curve recognition performance and unsmooth output lines in the original CLRNet algorithm,we improve the weight allocation for line intersection-over-union calculation in the original framework and propose two loss functions based on local slopes to optimize the model’s local sampling point training constraints,improving the model’s fitting performance on curved rails and obtaining smooth and stable rail line detection results.Through experiments,this paper demonstrates that compared with other mainstream lane detection algorithms,the algorithm proposed in this paper has a better performance for rail line detection.展开更多
Recently,the development and application of lane line departure warning systems have been in the market.For any of the systems,the key part of lane line tracking,lane line identification,or lane line departure warning...Recently,the development and application of lane line departure warning systems have been in the market.For any of the systems,the key part of lane line tracking,lane line identification,or lane line departure warning is whether it can accurately and quickly detect lane lines.Since 1990 s,they have been studied and implemented for the situations defined by the good viewing conditions and the clear lane markings on road.After then,the accuracy for particular situations,the robustness for a wide range of scenarios,time efficiency and integration into higher-order tasks define visual lane line detection and tracking as a continuing research subject.At present,these kinds of lane marking line detection methods based on machine vision and image processing can be divided into two categories:the traditional image processing and semantic segmentation(includes deep learning)methods.The former mainly involves feature-based and model-based steps,and which can be classified into similarity-and discontinuity-based ones;and the model-based step includes different parametric straight line,curve or pattern models.The semantic segmentation includes different machine learning,neural network and deep learning methods,which is the new trend for the research and application of lane line departure warning systems.This paper describes and analyzes the lane line departure warning systems,image processing algorithms and semantic segmentation methods for lane line detection.展开更多
Effective features are essential for fault diagnosis.Due to the faint characteristics of a single line-to-ground(SLG)fault,fault line detection has become a challenge in resonant grounding distribution systems.This pa...Effective features are essential for fault diagnosis.Due to the faint characteristics of a single line-to-ground(SLG)fault,fault line detection has become a challenge in resonant grounding distribution systems.This paper proposes a novel fault line detection method using waveform fusion and one-dimensional convolutional neural networks(1-D CNN).After an SLG fault occurs,the first-half waves of zero-sequence currents are collected and superimposed with each other to achieve waveform fusion.The compelling feature of fused waveforms is extracted by 1-D CNN to determine whether the fused waveform source contains the fault line.Then,the 1-D CNN output is used to update the value of the counter in order to identify the fault line.Given the lack of fault data in existing distribution systems,the proposed method only needs a small quantity of data for model training and fault line detection.In addition,the proposed method owns fault-tolerant performance.Even if a few samples are misjudged,the fault line can still be detected correctly based on the full output results of 1-D CNN.Experimental results verified that the proposed method can work effectively under various fault conditions.展开更多
In industrial,the modern intelligent degree of board counting is relatively low.Some small manufacturers still count by hand,but for large factories,manual counting needs a lot of human resources,and the accuracy of c...In industrial,the modern intelligent degree of board counting is relatively low.Some small manufacturers still count by hand,but for large factories,manual counting needs a lot of human resources,and the accuracy of counting is low.With the rapid development of modern intelligence,image recognition is becoming more mature,some traditional algorithms keep emerging,such as Hough Transform,Fast Line Detector(FLD),Line Segment Detector(LSD)and other line detection algorithms,they have their own advantages and disadvantages,and are summarized and tested,compare which algorithm has higher accuracy and better effect in the field of board linear detection and counting.Finally,the operation mechanism,advantages and disadvantages are summarized,and the process and trend of the further optimization and development of the traditional algorithm line detection technology in the future are prospected,which provides some reference for the research in related fields.展开更多
A line laser with high power as the background light source for the design of a new photoelectric detection target is proposed in this paper, aiming to improve the detection ability of the traditional photoelectric de...A line laser with high power as the background light source for the design of a new photoelectric detection target is proposed in this paper, aiming to improve the detection ability of the traditional photoelectric detection target under low background illumination. The laser emitted pulse waveform function and the laser echo pulse response function were used to establish the mathematical model of the reflected echo power of projectile in the detection area and derive the calculation function of minimum detectable echo power in the line laser detection screen, according to information of the line laser emitted power, incident angle of projectile, duration time and detection distance of projectile passing through the line laser detection screen. Calculations and experimental results showed that the design method of line laser detection screen and calculation model of laser echo power are reasonable, and the detection ability of line laser detection screen is obviously higher than that of traditional photoelectric detection screen, especially in low background illumination;at the same time, the designed line laser detection screen was used to combine a six line laser detection screen intersection test system, based on live ammunition for shooting. The test system is stable and able to obtain the dynamic parameters of the flying projectile, verifying that the design of the line laser detection screen in new photoelectric detection target can be suitable for shooting range test applications.展开更多
In order to improve the performance of line spectrum detection,according to the feature that the underwater target radiated noise containing stable line spectrum,the differences of the phase difference between line sp...In order to improve the performance of line spectrum detection,according to the feature that the underwater target radiated noise containing stable line spectrum,the differences of the phase difference between line spectrum and background noise,a weighted line spectrum detection algorithm based on the phase variance is proposed in frequency domain.After phase difference alignment,the phase variance of line spectrum and the phase of background noise,respectively,are small and big in frequency domain,this method utilizes the weighted statistical algorithm to cumulate the frequency spectrum based on the phase variance,which can restrain the background noise disturbance,and enhance the signal to noise ratio（SNR）.The theory analysis and experimental results both verify that the proposed method can well enhance the energy of line spectrum,restrain the energy of background noise,and have better detection performance under lower SNR.展开更多
The noise's statistical characteristics are very important for signal detection.In this paper,the ambient noise statistical characteristics are investigated by using the recorded noise data in sea trials first,and...The noise's statistical characteristics are very important for signal detection.In this paper,the ambient noise statistical characteristics are investigated by using the recorded noise data in sea trials first,and the results show that the generalized Gaussian distribution is a suitable model for the ambient noise modeling.Thereafter,the optimal detector based on maximum likelihood ratio can be deduced,and the asymptotic detector is also derived under weak signal assumption.The detector's performance is verified by using numerical simulation,and the results showthat the optimal and asymptotic detectors outperform the conventional correlation-integration system due to accuracy modeling of ambient noise.展开更多
Transient Rayleigh wave detection is a high-precision nondestructive detection method.At present,it has been widely used in shallow exploration,but rarely used in tunnel lining quality detection.Through the tunnel lin...Transient Rayleigh wave detection is a high-precision nondestructive detection method.At present,it has been widely used in shallow exploration,but rarely used in tunnel lining quality detection.Through the tunnel lining physical model experiment,the layout defects of the double-layer reinforcement lining area were detected and the Rayleigh wave velocity profile and dispersion curve were analyzed after data process-ing,which finally verified the feasibility and accuracy of Rayleigh wave method in detecting the tunnel lining void area.The results show that the method is not affected by the reinforcement inside the lining,the shallow detection is less disturbed and the accuracy is higher,and the data will fluctuate slightly with the deepening of the detection depth.At the same time,this method responds quite accurately to the thickness of the concrete,allowing for the assessment of the tunnel lining’s lack of compactness.This method has high efficiency,good reliability,and simple data processing,and is suitable for nondestructive detection of internal defects of tun-nel lining structure.展开更多
Adapter ring is a commonly used component in non-cooperative satellites,which has high strength and is suitable to be recognized and grasped by the space manipulator.During proximity operations,this circle feature may...Adapter ring is a commonly used component in non-cooperative satellites,which has high strength and is suitable to be recognized and grasped by the space manipulator.During proximity operations,this circle feature may be occluded by the robot arm or limited field of view.Moreover,the captured images may be underexposed when there is not enough illumination.To address these problems,this paper presents a structured light vision system with three line lasers and a monocular camera.The lasers project lines onto the surface of the satellite,and six break points are formed along both sides of the adapter ring.A closed-form solution for real-time pose estimation is given using these break points.Then,a virtual structured light platform is constructed to simulate synthetic images of the target satellite.Compared with the predefined camera parameters and relative positions,the proposed method is demonstrated to be more effective,especially at a close distance.Besides,a physical space verification system is set up to prove the effectiveness and robustness of our method under different light conditions.Experimental results indicate that it is a practical and effective method for the pose measurement of on-orbit tasks.展开更多
The vanishing point detection technology helps automatic driving. In this paper, the straight lines on the road associated with the vanishing point are extracted efficiently by using the regional division and angle li...The vanishing point detection technology helps automatic driving. In this paper, the straight lines on the road associated with the vanishing point are extracted efficiently by using the regional division and angle limitation. And, the vanishing point is detected robustly by using the fast M-estimation method. Proposed method could detect straight-line features associated with vanishing point detection efficient on the road. And the vanishing point was detected exactly by the effect of the fast M-estimation method when the straight-line features not associated with vanishing point detection were detected. The processing time of the proposed method was faster than the camera flame rate （30 fps）. Thus, the proposed method is capable of real-time processing.展开更多
Line parameters play an important role in the control and management of distribution systems.Currently,phasor measurement unit(PMU)systems and supervisory control and data acquisition(SCADA)systems coexist in distribu...Line parameters play an important role in the control and management of distribution systems.Currently,phasor measurement unit(PMU)systems and supervisory control and data acquisition(SCADA)systems coexist in distribution systems.Unfortunately,SCADA and PMU measurements usually do not match each other,resulting in inaccurate detection and identification of line parameters based on measurements.To solve this problem,a data-driven method is proposed.SCADA measurements are taken as samples and PMU measurements as the population.A probability parameter identification index(PPII)is derived to detect the whole line parameter based on the probability density function(PDF)parameters of the measurements.For parameter identification,a power-loss PDF with the PMU time stamps and a power-loss chronological PDF are derived via kernel density estimation(KDE)and a conditional PDF.Then,the power-loss samples with the PMU time stamps and chronological correlations are generated by the two PDFs of the power loss via the Metropolis-Hastings(MH)algorithm.Finally,using the power-loss samples and PMU current measurements,the line parameters are identified using the total least squares(TLS)algorithm.Hardware simulations demonstrate the effectiveness of the proposed method for distribution network line parameter detection and identification.展开更多
Cooperative target identification is the prerequisite for the relative position and orientation measurement between the space robot arm and the to-be-arrested object. We propose an on- orbit real-time robust algorithm...Cooperative target identification is the prerequisite for the relative position and orientation measurement between the space robot arm and the to-be-arrested object. We propose an on- orbit real-time robust algorithm for cooperative target identification in complex background using the features of circle and lines. It first extracts only the interested edges in the target image using an adaptive threshold and refines them to about single-pixel-width with improved non-maximum suppression. Adapting a novel tracking approach, edge segments changing smoothly in tangential directions are obtained. With a small amount of calculation, large numbers of invalid edges are removed. From the few remained edges, valid circular arcs are extracted and reassembled to obtain circles according to a reliable criterion. Finally, the target is identified if there are certain numbers of straight lines whose relative positions with the circle match the known target pattern. Experiments demonstrate that the proposed algorithm accurately identifies the cooperative target within the range of 0.3 1.5 m under complex background at the speed of 8 frames per second, regardless of lighting condition and target attitude. The proposed algorithm is very suitable for real-time visual measurement of space robot arm because of its robustness and small memory requirement.展开更多
Detecting internal defects,particularly voids behind linings,is critical for ensuring the structural integrity of aging high-speed rail(HSR)tunnel networks.While ground-penetrating radar(GPR)is widely employed,systema...Detecting internal defects,particularly voids behind linings,is critical for ensuring the structural integrity of aging high-speed rail(HSR)tunnel networks.While ground-penetrating radar(GPR)is widely employed,systematic quantification of performance boundaries for air-coupled(A-CGPR)and ground-coupled(G-CGPR)systems within the complex electromagnetic environment of multilayer reinforced HSR tunnels remains limited.This study establishes physics-based quantitative performance limits for A-CGPR and G-CGPR through rigorously validated GPRMax finite-difference time-domain(FDTD)simulations and comprehensive field validation over a 300 m operational HSR tunnel section.Key performance metrics were quantified as functions of:(a)detection distance(A-CGPR:2.0–4.5 m;G-CGPR:≤0.1 m),(b)antenna frequency(A-CGPR:300 MHz;G-CGPR:400/900 MHz),(c)reinforcement configuration(unreinforced,single-layer,multilayer rebar),and(d)void geometry(axial length:0.1–1.0 m;radial depth:0.1–0.5 m).Key findings demonstrate:a.A-CGPR(300 MHz):Reliably detects axial voids≥0.3 m at distances≤3 m in minimally reinforced(single-layer rebar)linings(field R2=0.89).Performance degrades significantly at distances>3 m(>60%signal attenuation at 4.5 m)or under multilayer rebar interference,causing 25%–40%accuracy loss for voids<0.3 m.Optimal distance:2.0–2.5 m.b.G-CGPR(900 MHz):Achieves<5%size measurement error for axial voids≥0.1 m and radial voids≥0.2 m in unreinforced linings.Resolution degrades under multilayer reinforcement due to severe signal attenuation,increasing axial void detection error to 10%–20%for voids≥0.3 m and constraining radial size measurement.c.Synergistic Framework:A hybrid inspection protocol is proposed,integrating A-CGPR(20 km/h)for rapid large-area screening and targeted G-CGPR(3 km/h)for high-resolution verification of identified anomalies.This framework enhances NDT efficiency while reducing estimated lifecycle inspection costs by 34%compared to G-CGPR alone.This research provides the first physics-derived quantitative detection thresholds for A-CGPR and G-CGPR in multi-rebar HSR tunnels,validated through field-correlated simulations.Future work will focus on multi-frequency antenna arrays and deep learning algorithms to mitigate reinforcement interference.The established performance boundaries and hybrid framework offer critical guidance for optimizing tunnel lining inspection strategies in extensive HSR networks.展开更多
A new type of intelligent recolser controller installed on the outdoor rod is developed, which is mainly composed of microcontroller of Intel 87C196KC 20 and CPLD devices. This controller integrates all the functions ...A new type of intelligent recolser controller installed on the outdoor rod is developed, which is mainly composed of microcontroller of Intel 87C196KC 20 and CPLD devices. This controller integrates all the functions of measuring, controlling, protection, fault diagnosis, communication, remote controlled operation and self power devices with infra red remote control devices as a unit. The controller applies the distributed structure, field concentration line and intelligent technology to seal up the synthetic servomechanisms such as the microcomputer based protection and measuring devices in the second stage of the mini out door transformer substation, which are distributed on the outdoor circuit switches on the spot and formed as a whole. Therefore, this technology can transform a large number of ordinary homemade SF 6 circuit beaker and vacuum circuit breaker into intelligent circuit recloser, thus replacing the expensive imported automatic circuit recolser.展开更多
For the purpose of resolving the problem of performance deterioration introduced by inaccurate phase compensation in existing coherent averaging line spectrum detectors, a modified coherent detector is proposed. The t...For the purpose of resolving the problem of performance deterioration introduced by inaccurate phase compensation in existing coherent averaging line spectrum detectors, a modified coherent detector is proposed. The three point interpolation in frequency domain is applied to obtain accurate estimate of phase difference between segments when the segmented length is not an integral multiple of the signal period. Then the segmented data are multiplied by a complex coefficient to remove the phase difference and synchronize the phases of all the segments before coherent averaging. Theoretical analysis shows that there will be a gain of 3.9 dB at most by using the modified detector. The detection performance of the incoher- ent averaging power spectrum detector （AVGPR）, the phase coherent averaging detector, the modified coherent averaging detector are compared with each other by computer simulations. The results coincide basically with the theoretical analysis, which show the superiority of the modified detector to the former two detectors.展开更多
This paper proposes a novel airport detection method,which integrates the texture features and shape features of the airport.Eight texture features,such as the mean of the region,the deviation of the region,the smooth...This paper proposes a novel airport detection method,which integrates the texture features and shape features of the airport.Eight texture features,such as the mean of the region,the deviation of the region,the smoothness of the region,the skewness of a histogram,the uniformity of the region,the randomness of the region,the mean of the gradient image and the deviation of the gradient image,are used to represent the features of the region.In this method,first the long lines are detected and the regions where the lines locate are segmented.Second,support vector machine(SVM)based on Gaussian kernel is used as a classifier which discriminates the runway from other candidate regions.Experimental results show that the error rate of the proposed method is lower than those of conventional methods which detect airport only by the shape feature of runway.The detection accuracy of the proposed method is nearly ten times higher than that of Liu’s methods,and the method has favorable speed for a real-time system.展开更多
基金Supported by National Natural Science Foundation of China(Grant Nos.51605003,51575001)Natural Science Foundation of Anhui Higher Education Institutions of China(Grant No.KJ2020A0358)Young and Middle-Aged Top Talents Training Program of Anhui Polytechnic University of China.
文摘The advancement of autonomous driving heavily relies on the ability to accurate lane lines detection.As deep learning and computer vision technologies evolve,a variety of deep learning-based methods for lane line detection have been proposed by researchers in the field.However,owing to the simple appearance of lane lines and the lack of distinctive features,it is easy for other objects with similar local appearances to interfere with the process of detecting lane lines.The precision of lane line detection is limited by the unpredictable quantity and diversity of lane lines.To address the aforementioned challenges,we propose a novel deep learning approach for lane line detection.This method leverages the Swin Transformer in conjunction with LaneNet(called ST-LaneNet).The experience results showed that the true positive detection rate can reach 97.53%for easy lanes and 96.83%for difficult lanes(such as scenes with severe occlusion and extreme lighting conditions),which can better accomplish the objective of detecting lane lines.In 1000 detection samples,the average detection accuracy can reach 97.83%,the average inference time per image can reach 17.8 ms,and the average number of frames per second can reach 64.8 Hz.The programming scripts and associated models for this project can be accessed openly at the following GitHub repository:https://github.com/Duane 711/Lane-line-detec tion-ST-LaneNet.
基金This research was funded by Key Laboratory for Advanced Manufacturing by Materials Processing Technology, Ministry of Education, The Research Fund for the Doctoral Program of Higher Education （No. 20020003053）National Natural Science Foundation of China （ No. 50275083 ）.
文摘A new efftcient straight line detection algorithm, GPI （ Gray Projecting Integral） method is proposed. The gray values of a sub-window are projected onto a line, and sum the gray values which are projected onto one same point to shape a special vector, then rotate the projecting direction, obtain many such vectors corresponding to different projecting directions. The vectors can form a matrix, a GPI matrix of the sub-image. The problem of lines detection is converted into maxima or minima searching problem in the GPI matrix. Bused on the GPI matrix, the lines can be calculated. Different from traditional methods, the algorithm can detect the positions of lines accurately, quickly without previous edge detection, which costs less time, and avoids the error resulted from the poor threshold with traditional methods. This algorithm is useful and efftcient for numerous image understanding applications and robot visual navigation, especially for welded joint position detection in heavy noise.
基金This work is supported by the National Natural Science Foundation of China(No.41171321,40871189)the Natural Science Foundation of the Jiangsu Higher Education Institutions of China(No.11KJA420001)+1 种基金the Priority Academic Program Development of Jiangsu Higher Education Institutionsthe Qin Lan Project of Jiangsu,China.
文摘Straight line detection is a fundamental problem in target recognition from remote sensing images since many man-made objects have straight boundaries.In this study,an integrated straight line detection method for remote sensing images is proposed.In this method,the edge-based straight lines are extracted using a chain code tracing method and the phasebased straight lines are extracted using a phase grouping method.The two types of lines are combined using a rule-based feature fusion method by removing redundant line extraction.Since this method integrates the specialties of edge-and phase-based straight line detection methods,it can detect straight lines from remote sensing images with high correctness and robustness.
文摘Accurate perception of lane line information is one of the basic requirements of unmanned driving technology, which is related to the localization of the vehicle and the determination of the forward direction. In this paper, multi-level constraints are added to the lane line detection model PINet, which is used to improve the perception of lane lines. Predicted lane lines in the network are predicted to have real and imaginary attributes, which are used to enhance the perception of features around the lane lines, with pixel-level constraints on the lane lines;images are converted to bird’s-eye views, where the parallelism between lane lines is reconstructed, with lane line-level constraints on the predicted lane lines;and vanishing points are used to focus on the image hierarchy, with image-level constraints on the lane lines. The model proposed in this paper meets both accuracy (96.44%) and real-time (30 + FPS) requirements, has been tested on the highway on the ground, and has performed stably.
基金the Sichuan Science and Technology Program(No.2022YFS0557)the National Natural Science Foundation of China(No.61972271)。
文摘In smart driving for rail transit,a reliable obstacle detection system is an important guarantee for the safety of trains.Therein,the detection of the rail area directly affects the accuracy of the system to identify dangerous targets.Both the rail line and the lane are presented as thin line shapes in the image,but the rail scene is more complex,and the color of the rail line is more difficult to distinguish from the background.By comparison,there are already many deep learning-based lane detection algorithms,but there is a lack of public datasets and targeted deep learning detection algorithms for rail line detection.To address this,this paper constructs a rail image dataset RailwayLine and labels the rail line for the training and testing of models.This dataset contains rich rail images including single-rail,multi-rail,straight rail,curved rail,crossing rails,occlusion,blur,and different lighting conditions.To address the problem of the lack of deep learning-based rail line detection algorithms,we improve the CLRNet algorithm which has an excellent performance in lane detection,and propose the CLRNet-R algorithm for rail line detection.To address the problem of the rail line being thin and occupying fewer pixels in the image,making it difficult to distinguish from complex backgrounds,we introduce an attention mechanism to enhance global feature extraction ability and add a semantic segmentation head to enhance the features of the rail region by the binary probability of rail lines.To address the poor curve recognition performance and unsmooth output lines in the original CLRNet algorithm,we improve the weight allocation for line intersection-over-union calculation in the original framework and propose two loss functions based on local slopes to optimize the model’s local sampling point training constraints,improving the model’s fitting performance on curved rails and obtaining smooth and stable rail line detection results.Through experiments,this paper demonstrates that compared with other mainstream lane detection algorithms,the algorithm proposed in this paper has a better performance for rail line detection.
基金financially supported by the National Natural Science Foundation of China(grant No.61170147)the Scientific and Technological Project of Shaanxi Province in China(grant No.2019GY-038)。
文摘Recently,the development and application of lane line departure warning systems have been in the market.For any of the systems,the key part of lane line tracking,lane line identification,or lane line departure warning is whether it can accurately and quickly detect lane lines.Since 1990 s,they have been studied and implemented for the situations defined by the good viewing conditions and the clear lane markings on road.After then,the accuracy for particular situations,the robustness for a wide range of scenarios,time efficiency and integration into higher-order tasks define visual lane line detection and tracking as a continuing research subject.At present,these kinds of lane marking line detection methods based on machine vision and image processing can be divided into two categories:the traditional image processing and semantic segmentation(includes deep learning)methods.The former mainly involves feature-based and model-based steps,and which can be classified into similarity-and discontinuity-based ones;and the model-based step includes different parametric straight line,curve or pattern models.The semantic segmentation includes different machine learning,neural network and deep learning methods,which is the new trend for the research and application of lane line departure warning systems.This paper describes and analyzes the lane line departure warning systems,image processing algorithms and semantic segmentation methods for lane line detection.
基金supported by the National Natural Science Foundation of China through the Project of Research of Flexible and Adaptive Arc-Suppression Method for Single-Phase Grounding Fault in Distribution Networks(No.51677030).
文摘Effective features are essential for fault diagnosis.Due to the faint characteristics of a single line-to-ground(SLG)fault,fault line detection has become a challenge in resonant grounding distribution systems.This paper proposes a novel fault line detection method using waveform fusion and one-dimensional convolutional neural networks(1-D CNN).After an SLG fault occurs,the first-half waves of zero-sequence currents are collected and superimposed with each other to achieve waveform fusion.The compelling feature of fused waveforms is extracted by 1-D CNN to determine whether the fused waveform source contains the fault line.Then,the 1-D CNN output is used to update the value of the counter in order to identify the fault line.Given the lack of fault data in existing distribution systems,the proposed method only needs a small quantity of data for model training and fault line detection.In addition,the proposed method owns fault-tolerant performance.Even if a few samples are misjudged,the fault line can still be detected correctly based on the full output results of 1-D CNN.Experimental results verified that the proposed method can work effectively under various fault conditions.
基金supported by National Natural Science Foundation of China(No.61775139)Zhejiang Provincial Key R&D Plan(No.2020C02020).
文摘In industrial,the modern intelligent degree of board counting is relatively low.Some small manufacturers still count by hand,but for large factories,manual counting needs a lot of human resources,and the accuracy of counting is low.With the rapid development of modern intelligence,image recognition is becoming more mature,some traditional algorithms keep emerging,such as Hough Transform,Fast Line Detector(FLD),Line Segment Detector(LSD)and other line detection algorithms,they have their own advantages and disadvantages,and are summarized and tested,compare which algorithm has higher accuracy and better effect in the field of board linear detection and counting.Finally,the operation mechanism,advantages and disadvantages are summarized,and the process and trend of the further optimization and development of the traditional algorithm line detection technology in the future are prospected,which provides some reference for the research in related fields.
基金This work has been supported by Project of the National Natural Science Foundation of China(No.62073256,61773305)in part by the Key Science and Technology Program of Shaanxi Province(No.2020GY-125)Xi’an Science and Technology Innovation Talent Service Enterprise Project(No.2020KJRC0041).
文摘A line laser with high power as the background light source for the design of a new photoelectric detection target is proposed in this paper, aiming to improve the detection ability of the traditional photoelectric detection target under low background illumination. The laser emitted pulse waveform function and the laser echo pulse response function were used to establish the mathematical model of the reflected echo power of projectile in the detection area and derive the calculation function of minimum detectable echo power in the line laser detection screen, according to information of the line laser emitted power, incident angle of projectile, duration time and detection distance of projectile passing through the line laser detection screen. Calculations and experimental results showed that the design method of line laser detection screen and calculation model of laser echo power are reasonable, and the detection ability of line laser detection screen is obviously higher than that of traditional photoelectric detection screen, especially in low background illumination;at the same time, the designed line laser detection screen was used to combine a six line laser detection screen intersection test system, based on live ammunition for shooting. The test system is stable and able to obtain the dynamic parameters of the flying projectile, verifying that the design of the line laser detection screen in new photoelectric detection target can be suitable for shooting range test applications.
基金supported by the National Natural Science Foundation of China(61372180)the Young Talent Frontier Project of Institute of Acoustics of Chinese Academy of Sciences(Y454341261)
文摘In order to improve the performance of line spectrum detection,according to the feature that the underwater target radiated noise containing stable line spectrum,the differences of the phase difference between line spectrum and background noise,a weighted line spectrum detection algorithm based on the phase variance is proposed in frequency domain.After phase difference alignment,the phase variance of line spectrum and the phase of background noise,respectively,are small and big in frequency domain,this method utilizes the weighted statistical algorithm to cumulate the frequency spectrum based on the phase variance,which can restrain the background noise disturbance,and enhance the signal to noise ratio（SNR）.The theory analysis and experimental results both verify that the proposed method can well enhance the energy of line spectrum,restrain the energy of background noise,and have better detection performance under lower SNR.
基金Sponsored by the National Nature Science Foundation of China(11074308)China Postdoctoral Science Foundation(201003754)
文摘The noise's statistical characteristics are very important for signal detection.In this paper,the ambient noise statistical characteristics are investigated by using the recorded noise data in sea trials first,and the results show that the generalized Gaussian distribution is a suitable model for the ambient noise modeling.Thereafter,the optimal detector based on maximum likelihood ratio can be deduced,and the asymptotic detector is also derived under weak signal assumption.The detector's performance is verified by using numerical simulation,and the results showthat the optimal and asymptotic detectors outperform the conventional correlation-integration system due to accuracy modeling of ambient noise.
基金Supported by Project of Natural Science Foundation of Jilin Province(No.20220101172JC).
文摘Transient Rayleigh wave detection is a high-precision nondestructive detection method.At present,it has been widely used in shallow exploration,but rarely used in tunnel lining quality detection.Through the tunnel lining physical model experiment,the layout defects of the double-layer reinforcement lining area were detected and the Rayleigh wave velocity profile and dispersion curve were analyzed after data process-ing,which finally verified the feasibility and accuracy of Rayleigh wave method in detecting the tunnel lining void area.The results show that the method is not affected by the reinforcement inside the lining,the shallow detection is less disturbed and the accuracy is higher,and the data will fluctuate slightly with the deepening of the detection depth.At the same time,this method responds quite accurately to the thickness of the concrete,allowing for the assessment of the tunnel lining’s lack of compactness.This method has high efficiency,good reliability,and simple data processing,and is suitable for nondestructive detection of internal defects of tun-nel lining structure.
基金financial support provided by the Foundation for Innovative Research Groups of the National Natural Science Foundation of China (Nos. 51521003 and 61690210)
文摘Adapter ring is a commonly used component in non-cooperative satellites,which has high strength and is suitable to be recognized and grasped by the space manipulator.During proximity operations,this circle feature may be occluded by the robot arm or limited field of view.Moreover,the captured images may be underexposed when there is not enough illumination.To address these problems,this paper presents a structured light vision system with three line lasers and a monocular camera.The lasers project lines onto the surface of the satellite,and six break points are formed along both sides of the adapter ring.A closed-form solution for real-time pose estimation is given using these break points.Then,a virtual structured light platform is constructed to simulate synthetic images of the target satellite.Compared with the predefined camera parameters and relative positions,the proposed method is demonstrated to be more effective,especially at a close distance.Besides,a physical space verification system is set up to prove the effectiveness and robustness of our method under different light conditions.Experimental results indicate that it is a practical and effective method for the pose measurement of on-orbit tasks.
文摘The vanishing point detection technology helps automatic driving. In this paper, the straight lines on the road associated with the vanishing point are extracted efficiently by using the regional division and angle limitation. And, the vanishing point is detected robustly by using the fast M-estimation method. Proposed method could detect straight-line features associated with vanishing point detection efficient on the road. And the vanishing point was detected exactly by the effect of the fast M-estimation method when the straight-line features not associated with vanishing point detection were detected. The processing time of the proposed method was faster than the camera flame rate （30 fps）. Thus, the proposed method is capable of real-time processing.
基金supported by the National Key Research and Development Program under Grant 2017YFB0902900 and Grant 2017YFB0902902。
文摘Line parameters play an important role in the control and management of distribution systems.Currently,phasor measurement unit(PMU)systems and supervisory control and data acquisition(SCADA)systems coexist in distribution systems.Unfortunately,SCADA and PMU measurements usually do not match each other,resulting in inaccurate detection and identification of line parameters based on measurements.To solve this problem,a data-driven method is proposed.SCADA measurements are taken as samples and PMU measurements as the population.A probability parameter identification index(PPII)is derived to detect the whole line parameter based on the probability density function(PDF)parameters of the measurements.For parameter identification,a power-loss PDF with the PMU time stamps and a power-loss chronological PDF are derived via kernel density estimation(KDE)and a conditional PDF.Then,the power-loss samples with the PMU time stamps and chronological correlations are generated by the two PDFs of the power loss via the Metropolis-Hastings(MH)algorithm.Finally,using the power-loss samples and PMU current measurements,the line parameters are identified using the total least squares(TLS)algorithm.Hardware simulations demonstrate the effectiveness of the proposed method for distribution network line parameter detection and identification.
基金supported by the National Basic Research Program of China (No. 2013CB733103)
文摘Cooperative target identification is the prerequisite for the relative position and orientation measurement between the space robot arm and the to-be-arrested object. We propose an on- orbit real-time robust algorithm for cooperative target identification in complex background using the features of circle and lines. It first extracts only the interested edges in the target image using an adaptive threshold and refines them to about single-pixel-width with improved non-maximum suppression. Adapting a novel tracking approach, edge segments changing smoothly in tangential directions are obtained. With a small amount of calculation, large numbers of invalid edges are removed. From the few remained edges, valid circular arcs are extracted and reassembled to obtain circles according to a reliable criterion. Finally, the target is identified if there are certain numbers of straight lines whose relative positions with the circle match the known target pattern. Experiments demonstrate that the proposed algorithm accurately identifies the cooperative target within the range of 0.3 1.5 m under complex background at the speed of 8 frames per second, regardless of lighting condition and target attitude. The proposed algorithm is very suitable for real-time visual measurement of space robot arm because of its robustness and small memory requirement.
基金funded by the Key Project of Science&Technology Research ofChina Academy of Railway Sciences,grant number 2023YJ022.
文摘Detecting internal defects,particularly voids behind linings,is critical for ensuring the structural integrity of aging high-speed rail(HSR)tunnel networks.While ground-penetrating radar(GPR)is widely employed,systematic quantification of performance boundaries for air-coupled(A-CGPR)and ground-coupled(G-CGPR)systems within the complex electromagnetic environment of multilayer reinforced HSR tunnels remains limited.This study establishes physics-based quantitative performance limits for A-CGPR and G-CGPR through rigorously validated GPRMax finite-difference time-domain(FDTD)simulations and comprehensive field validation over a 300 m operational HSR tunnel section.Key performance metrics were quantified as functions of:(a)detection distance(A-CGPR:2.0–4.5 m;G-CGPR:≤0.1 m),(b)antenna frequency(A-CGPR:300 MHz;G-CGPR:400/900 MHz),(c)reinforcement configuration(unreinforced,single-layer,multilayer rebar),and(d)void geometry(axial length:0.1–1.0 m;radial depth:0.1–0.5 m).Key findings demonstrate:a.A-CGPR(300 MHz):Reliably detects axial voids≥0.3 m at distances≤3 m in minimally reinforced(single-layer rebar)linings(field R2=0.89).Performance degrades significantly at distances>3 m(>60%signal attenuation at 4.5 m)or under multilayer rebar interference,causing 25%–40%accuracy loss for voids<0.3 m.Optimal distance:2.0–2.5 m.b.G-CGPR(900 MHz):Achieves<5%size measurement error for axial voids≥0.1 m and radial voids≥0.2 m in unreinforced linings.Resolution degrades under multilayer reinforcement due to severe signal attenuation,increasing axial void detection error to 10%–20%for voids≥0.3 m and constraining radial size measurement.c.Synergistic Framework:A hybrid inspection protocol is proposed,integrating A-CGPR(20 km/h)for rapid large-area screening and targeted G-CGPR(3 km/h)for high-resolution verification of identified anomalies.This framework enhances NDT efficiency while reducing estimated lifecycle inspection costs by 34%compared to G-CGPR alone.This research provides the first physics-derived quantitative detection thresholds for A-CGPR and G-CGPR in multi-rebar HSR tunnels,validated through field-correlated simulations.Future work will focus on multi-frequency antenna arrays and deep learning algorithms to mitigate reinforcement interference.The established performance boundaries and hybrid framework offer critical guidance for optimizing tunnel lining inspection strategies in extensive HSR networks.
文摘A new type of intelligent recolser controller installed on the outdoor rod is developed, which is mainly composed of microcontroller of Intel 87C196KC 20 and CPLD devices. This controller integrates all the functions of measuring, controlling, protection, fault diagnosis, communication, remote controlled operation and self power devices with infra red remote control devices as a unit. The controller applies the distributed structure, field concentration line and intelligent technology to seal up the synthetic servomechanisms such as the microcomputer based protection and measuring devices in the second stage of the mini out door transformer substation, which are distributed on the outdoor circuit switches on the spot and formed as a whole. Therefore, this technology can transform a large number of ordinary homemade SF 6 circuit beaker and vacuum circuit breaker into intelligent circuit recloser, thus replacing the expensive imported automatic circuit recolser.
文摘For the purpose of resolving the problem of performance deterioration introduced by inaccurate phase compensation in existing coherent averaging line spectrum detectors, a modified coherent detector is proposed. The three point interpolation in frequency domain is applied to obtain accurate estimate of phase difference between segments when the segmented length is not an integral multiple of the signal period. Then the segmented data are multiplied by a complex coefficient to remove the phase difference and synchronize the phases of all the segments before coherent averaging. Theoretical analysis shows that there will be a gain of 3.9 dB at most by using the modified detector. The detection performance of the incoher- ent averaging power spectrum detector （AVGPR）, the phase coherent averaging detector, the modified coherent averaging detector are compared with each other by computer simulations. The results coincide basically with the theoretical analysis, which show the superiority of the modified detector to the former two detectors.
基金supported by the National Natural Science Foundation of China(Grant No.60175006).
文摘This paper proposes a novel airport detection method,which integrates the texture features and shape features of the airport.Eight texture features,such as the mean of the region,the deviation of the region,the smoothness of the region,the skewness of a histogram,the uniformity of the region,the randomness of the region,the mean of the gradient image and the deviation of the gradient image,are used to represent the features of the region.In this method,first the long lines are detected and the regions where the lines locate are segmented.Second,support vector machine(SVM)based on Gaussian kernel is used as a classifier which discriminates the runway from other candidate regions.Experimental results show that the error rate of the proposed method is lower than those of conventional methods which detect airport only by the shape feature of runway.The detection accuracy of the proposed method is nearly ten times higher than that of Liu’s methods,and the method has favorable speed for a real-time system.
