Detection of Surface Defects in Lightweight Insulators Using Improved YOLOv5
Yu Guo, Meiling Ma, and Dalin Li
Herein, an improved insulator defect-detection algorithm, YOLOv5, is proposed to overcome the shortcomings, including inconspicuous target features and poor detection of small targets when detecting trapped insulators using unmanned aerial vehicles, which cannot satisfy both detection speed and accuracy. First, ConvNeXt is applied to the YOLOv5 reference network to improve its ability to extract the features of obscure targets. Moreover, a coordinate attention mechanism is introduced into the reference network to improve its detection accuracy with respect to small targets in an image. Then, the improved model is pruned to eliminate its redundant channels, thus reducing the number of model parameters and making the model more lightweight. The experimental results show that the improved model achieves an average detection accuracy of 93.84% with respect to the insulator-defect dataset IDID, which is 3.4 percentage points higher than the accuracy achieved by the original algorithm. Moreover, the highest detection rate achieved by the proposed algorithm is 166 frame/s, which is 69.4% higher than that achieved by the original algorithm. These results prove that the improved algorithm meets the requirements of real-time transmission-line detection.Herein, an improved insulator defect-detection algorithm, YOLOv5, is proposed to overcome the shortcomings, including inconspicuous target features and poor detection of small targets when detecting trapped insulators using unmanned aerial vehicles, which cannot satisfy both detection speed and accuracy. First, ConvNeXt is applied to the YOLOv5 reference network to improve its ability to extract the features of obscure targets. Moreover, a coordinate attention mechanism is introduced into the reference network to improve its detection accuracy with respect to small targets in an image. Then, the improved model is pruned to eliminate its redundant channels, thus reducing the number of model parameters and making the model more lightweight. The experimental results show that the improved model achieves an average detection accuracy of 93.84% with respect to the insulator-defect dataset IDID, which is 3.4 percentage points higher than the accuracy achieved by the original algorithm. Moreover, the highest detection rate achieved by the proposed algorithm is 166 frame/s, which is 69.4% higher than that achieved by the original algorithm. These results prove that the improved algorithm meets the requirements of real-time transmission-line detection.
- Dec. 25, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 24, 2412007 (2023)
- DOI：10.3788/LOP231032
Infrared Vehicle Detection Algorithm Based on Improved Shuffle-RetinaNet
Xiaochang Fan, Yu Liang, and Wei Zhang
In view of the low detection accuracy and high complexity of current multi-scale vehicle detection algorithms in infrared scenes, an infrared vehicle detection algorithm based on Shuffle-RetinaNet is proposed. On the basis of RetinaNet, the algorithm uses ShuffleNetV2 as the feature extraction network. A dual-branch attention module channel attention module is proposed, which adopts the dual-branch structure and adaptive fusion and enhances the ability to extract the key features of the target in infrared images. To optimize the feature fusion, the algorithm integrates cross-scale connection and fast normalized fusion in some feature layers to enhance the multi-scale feature expression. The calibration factor is set to enhance the task interaction of classification and regression, and the accuracy of target classification and locating is increased. A series of experiments are conducted on a self-built infrared vehicle dataset to verify the effectiveness of the proposed algorithm. The detection accuracy of this algorithm for the self-built vehicle dataset is 92.9%, the number of parameters is 11.74×106, and the number of floating-point operations is 24.35×109. The algorithm exhibits better detection performance on the public dataset FLIR ADAS. Experimental results indicate that the algorithm has advantages in detection accuracy and model complexity, giving it good application value in multi-scale vehicle detection tasks in infrared scenes.In view of the low detection accuracy and high complexity of current multi-scale vehicle detection algorithms in infrared scenes, an infrared vehicle detection algorithm based on Shuffle-RetinaNet is proposed. On the basis of RetinaNet, the algorithm uses ShuffleNetV2 as the feature extraction network. A dual-branch attention module channel attention module is proposed, which adopts the dual-branch structure and adaptive fusion and enhances the ability to extract the key features of the target in infrared images. To optimize the feature fusion, the algorithm integrates cross-scale connection and fast normalized fusion in some feature layers to enhance the multi-scale feature expression. The calibration factor is set to enhance the task interaction of classification and regression, and the accuracy of target classification and locating is increased. A series of experiments are conducted on a self-built infrared vehicle dataset to verify the effectiveness of the proposed algorithm. The detection accuracy of this algorithm for the self-built vehicle dataset is 92.9%, the number of parameters is 11.74×106, and the number of floating-point operations is 24.35×109. The algorithm exhibits better detection performance on the public dataset FLIR ADAS. Experimental results indicate that the algorithm has advantages in detection accuracy and model complexity, giving it good application value in multi-scale vehicle detection tasks in infrared scenes.
- Dec. 25, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 24, 2410006 (2023)
- DOI：10.3788/LOP230713
Research Progress on Laser-Induced Breakdown Spectroscopy in Biomedicine
Zehai Hou, Lianbo Guo, Weiliang Wang, Yanwu Chu, and Furi Lin
Minor changes to elements in organisms can have a direct impact on their metabolism and physiological processes. Rapid and accurate qualitative and quantitative analyses of these elements are critical for both metabolism detection and clinical disease diagnosis. As medical detection technology continues to develop and clinical demand continues to increase, researchers are seeking newer, faster, and more adaptable clinical analysis and diagnostic technologies. Because of its fast, real-time, and multi-element simultaneous detection capabilities, laser-induced breakdown spectroscopy (LIBS) technology has shown great promise in recent years for use in blood and pathological tissue detection and elemental distribution imaging. This paper provides a comprehensive review of the current research status and latest progress of LIBS technology in biomedical applications and evaluates the challenges and opportunities of LIBS technology in the application fields of blood detection, biological tissue analysis, and elemental imaging. The paper also provides suggestions for further promoting the application of LIBS technology in the biomedical field.Minor changes to elements in organisms can have a direct impact on their metabolism and physiological processes. Rapid and accurate qualitative and quantitative analyses of these elements are critical for both metabolism detection and clinical disease diagnosis. As medical detection technology continues to develop and clinical demand continues to increase, researchers are seeking newer, faster, and more adaptable clinical analysis and diagnostic technologies. Because of its fast, real-time, and multi-element simultaneous detection capabilities, laser-induced breakdown spectroscopy (LIBS) technology has shown great promise in recent years for use in blood and pathological tissue detection and elemental distribution imaging. This paper provides a comprehensive review of the current research status and latest progress of LIBS technology in biomedical applications and evaluates the challenges and opportunities of LIBS technology in the application fields of blood detection, biological tissue analysis, and elemental imaging. The paper also provides suggestions for further promoting the application of LIBS technology in the biomedical field.
- Dec. 25, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 24, 2400004 (2023)
- DOI：10.3788/LOP231055
Fabrication of Infrared Micro-Optical Components Using Femtosecond Laser
Yahua Niu, Shengzhi Sun, Xiaofeng Liu, and Jianrong Qiu
With the advancements in the field of infrared micro-optical technologies, the preparation of infrared micro-optical devices with high precision has attracted increased attention. There are many shortcomings in the traditional preparation technology. However, the femtosecond laser is suitable for preparing infrared micro-optical components because of its ultra-fast characteristics. Taking lens array, compound eye, grating, optical waveguide and photonic crystal as examples, the development of infrared micro-optical components with the femtosecond laser using different materials and manufacturing methods is introduced. The materials used are the infrared semiconductor, chalcogenide glass, and infrared polymer. The methods used are the femtosecond laser-induced chemical etching, femtosecond laser-assisted wet etching, and femtosecond laser-assisted dry etching. The applications and specific cases are discussed, and the future development trend of this technology is presented.With the advancements in the field of infrared micro-optical technologies, the preparation of infrared micro-optical devices with high precision has attracted increased attention. There are many shortcomings in the traditional preparation technology. However, the femtosecond laser is suitable for preparing infrared micro-optical components because of its ultra-fast characteristics. Taking lens array, compound eye, grating, optical waveguide and photonic crystal as examples, the development of infrared micro-optical components with the femtosecond laser using different materials and manufacturing methods is introduced. The materials used are the infrared semiconductor, chalcogenide glass, and infrared polymer. The methods used are the femtosecond laser-induced chemical etching, femtosecond laser-assisted wet etching, and femtosecond laser-assisted dry etching. The applications and specific cases are discussed, and the future development trend of this technology is presented.
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2314006 (2023)
- DOI：10.3788/LOP223127
All-Optical Control of Ultrahigh-Q Whispering Gallery Microspheres with Laser-Induced Graphene
Xueyang Zhang, Huomu Yang, Guoliang Deng, and Shouhuan Zhou
We present a method for all-optical control of silica whispering-gallery-mode microspheres with laser-induced graphene (LIG). Polyimide (PI) films are carbonized to LIG under CO2 laser irradiation. A 980 nm laser is used as a pump light to irradiate the LIG surface, and the transmission and reflection spectra of microcavities are studied. Experimental results show that the tuning scheme maintains a Q factor of approximately 108 throughout the tuning process, with a tuning range and sensitivity of approximately 1.09 nm and approximately 8.8 pm/mW, respectively. This scheme has the advantages of having no mechanical interference, an ultrahigh-Q factor, and a wide tuning range, thus extending all-optical tuning to applications in cavity quantum dynamics, nonlinear optics, and low-threshold lasers.We present a method for all-optical control of silica whispering-gallery-mode microspheres with laser-induced graphene (LIG). Polyimide (PI) films are carbonized to LIG under CO2 laser irradiation. A 980 nm laser is used as a pump light to irradiate the LIG surface, and the transmission and reflection spectra of microcavities are studied. Experimental results show that the tuning scheme maintains a Q factor of approximately
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2314004 (2023)
- DOI：10.3788/LOP222917
Effect of TiB2 Content on Microstructure and Properties of in-situ TiB/Ti6Al4V Composites
Weidong Huang, Yitao Zhu, Xu Huang, Lu Wang, and Xuheng Cheng
Ti6Al4V powder added TiB2 particles with a mass fraction of 0%~5% was melted using selective laser melting to produce the in-situ TiB/Ti6Al4V composites to study the forming properties, microstructure evolution, and mechanical properties of samples with different TiB2 additions. When the mass fraction of TiB2 added was 1%, the α′-Ti in the microstructure of the sample gradually disappeared, and a dispersedly distributed unit cell like TiB was observed. Further addition of TiB2 caused the strengthening method of the material to evolve from dispersion strengthening to whisker strengthening. Simultaneously, TiB in the grain boundary was enriched to form a dendritic structure and a network structure. The in-situ TiB as a nucleation point can remarkably enhance the hardness and friction performances of the TiB/Ti6Al4V sample under the action of fine grain strengthening, solid solution strengthening, and whisker strengthening. As a result, the microhardness of the sample increased from (336.8±6.64) HV to (498.07±12.56) HV, and the wear mechanism of the sample changed from adhesive wear to abrasive wear. When the mass fraction of TiB2 added was 1%, better wear performance was obtained.Ti6Al4V powder added TiB2 particles with a mass fraction of 0%~5% was melted using selective laser melting to produce the in-situ TiB/Ti6Al4V composites to study the forming properties, microstructure evolution, and mechanical properties of samples with different TiB2 additions. When the mass fraction of TiB2 added was 1%, the α′-Ti in the microstructure of the sample gradually disappeared, and a dispersedly distributed unit cell like TiB was observed. Further addition of TiB2 caused the strengthening method of the material to evolve from dispersion strengthening to whisker strengthening. Simultaneously, TiB in the grain boundary was enriched to form a dendritic structure and a network structure. The in-situ TiB as a nucleation point can remarkably enhance the hardness and friction performances of the TiB/Ti6Al4V sample under the action of fine grain strengthening, solid solution strengthening, and whisker strengthening. As a result, the microhardness of the sample increased from (336.8±6.64) HV to (498.07±12.56) HV, and the wear mechanism of the sample changed from adhesive wear to abrasive wear. When the mass fraction of TiB2 added was 1%, better wear performance was obtained.
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2314003 (2023)
- DOI：10.3788/LOP222681
Effect of Continuous Laser Cleaning Process Parameters on Surface Quality of Q235B Carbon Steel Rust Layer
Qiupei Wu, Xiangyang Sun, Jiapo Sun, Lianghua Han, and Lie Liu
In this study, the effects of laser power and cleaning speed on the surface macro- and micromorphology of Q235B carbon steel after the rust layer was cleaned with a continuous laser at 1080 nm, are investigated. Moreover, the effect of different process parameters on roughness is analyzed, and cross-sectional metallographic observations, element content, and compound analysis of the cleaned sample surface are carried out. From hardness test and electrochemical analysis, it is found that the minimum roughness of sample surface is 3.94 μm, and the laser cleaning effect is the best when the cleaning speed is constant at 100 mm·s-1, laser power is 4 kW, Fe content is the highest, and O content is the lowest. When the laser power is constant at 7 kW, and the cleaning speed increases from 100 mm·s-1 to 500 mm·s-1, the surface roughness of the specimen first decreases and then increases. The minimum roughness is 3.68 μm when the cleaning speed is 400 mm·s-1, at which time the laser can completely clean the rust layer in a single scan, and its cleaning efficiency is 20.6 m2·h-1. A remelted layer is generated on the surface of the cleaned substrate, which improves the corrosion resistance of the substrate surface after cleaning the rust layer, and its hardness is approximately 4 times higher than that of the steel when the laser power is 7 kW and the cleaning speed is 100 mm·s-1. A continuous rectangular spot laser cleaning model is created using Ansys software and its results are compared with the experimental results to determine the parameters and estimate their effects for the high power continuous laser cleaning process.In this study, the effects of laser power and cleaning speed on the surface macro- and micromorphology of Q235B carbon steel after the rust layer was cleaned with a continuous laser at 1080 nm, are investigated. Moreover, the effect of different process parameters on roughness is analyzed, and cross-sectional metallographic observations, element content, and compound analysis of the cleaned sample surface are carried out. From hardness test and electrochemical analysis, it is found that the minimum roughness of sample surface is 3.94 μm, and the laser cleaning effect is the best when the cleaning speed is constant at 100 mm·s-1, laser power is 4 kW, Fe content is the highest, and O content is the lowest. When the laser power is constant at 7 kW, and the cleaning speed increases from 100 mm·s-1 to 500 mm·s-1, the surface roughness of the specimen first decreases and then increases. The minimum roughness is 3.68 μm when the cleaning speed is 400 mm·s-1, at which time the laser can completely clean the rust layer in a single scan, and its cleaning efficiency is 20.6 m2·h-1. A remelted layer is generated on the surface of the cleaned substrate, which improves the corrosion resistance of the substrate surface after cleaning the rust layer, and its hardness is approximately 4 times higher than that of the steel when the laser power is 7 kW and the cleaning speed is 100 mm·s-1. A continuous rectangular spot laser cleaning model is created using Ansys software and its results are compared with the experimental results to determine the parameters and estimate their effects for the high power continuous laser cleaning process.
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2314002 (2023)
- DOI：10.3788/LOP222926
Atomic Emission Spectrometry Oil Detection Technology Based on Double Turntable Electrode Structure
Bin Chen, Xiao Fu, Fajie Duan, Yu Yan, Jinfan Huang, and Guoshun Zhong
Atomic emission spectroscopy of turntable electrode (RDE-AES) is widely used in oil detection because of its simple operation, no need for sample preparation, and high reliability. However, the arc is the primary light source used in this technology. The instability of the arc caused by the change in the discharge gap induced by electrode wear and other factors leads to errors between the analysis results of the last collected spectral data and the actual situation. Herein, a method using atomic emission spectrometry with an oil detection device based on the “double turntable” electrode structure is proposed. In other words, the rod electrode in the traditional “rod turntable” electrode structure is replaced by a turntable electrode that can rotate. Its significant advantage is that it reduces the detection error caused by electrode wear. For its structural physical modeling, the process of arc excitation is simulated by COMSOL multi physical field simulation software. Further, the changing rules of electrode gap, oil film thickness, and external excitation on the arc excitation effect are explored using the control variable method. The effects of the influencing factors on the arc excitation time and instantaneous excitation temperature are obtained, and the parameters are optimized according to the simulation results. The simulation results show that the excitation effect of the “double turntable" electrode structure is significantly improved compared with the traditional structure, and the excitation time and temperature are improved to a certain extent. In particular, the arc excitation effect is stable in mass testing, which verifies the progressiveness and practicability of this method. It also provides analytical support for in-depth research on oil detection using the method based on turntable electrode atomic emission spectrometry.Atomic emission spectroscopy of turntable electrode (RDE-AES) is widely used in oil detection because of its simple operation, no need for sample preparation, and high reliability. However, the arc is the primary light source used in this technology. The instability of the arc caused by the change in the discharge gap induced by electrode wear and other factors leads to errors between the analysis results of the last collected spectral data and the actual situation. Herein, a method using atomic emission spectrometry with an oil detection device based on the “double turntable” electrode structure is proposed. In other words, the rod electrode in the traditional “rod turntable” electrode structure is replaced by a turntable electrode that can rotate. Its significant advantage is that it reduces the detection error caused by electrode wear. For its structural physical modeling, the process of arc excitation is simulated by COMSOL multi physical field simulation software. Further, the changing rules of electrode gap, oil film thickness, and external excitation on the arc excitation effect are explored using the control variable method. The effects of the influencing factors on the arc excitation time and instantaneous excitation temperature are obtained, and the parameters are optimized according to the simulation results. The simulation results show that the excitation effect of the “double turntable" electrode structure is significantly improved compared with the traditional structure, and the excitation time and temperature are improved to a certain extent. In particular, the arc excitation effect is stable in mass testing, which verifies the progressiveness and practicability of this method. It also provides analytical support for in-depth research on oil detection using the method based on turntable electrode atomic emission spectrometry.
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2312003 (2023)
- DOI：10.3788/LOP223175
Self-Calibration Method for Roll Angle Measurements of Dual Beams Based on Polarization
Fengrui Ma, Fajie Duan, Wenzheng Liu, Xiao Fu, and Cong Zhang
When dual beams are used to measure the roll angle, the angle between the double-collimated beams is easily disturbed by environmental changes, mechanical deformation, and other factors, which seriously affect measurement accuracy. The roll angle measured by the dual beam is sensitive to the spot position. By contrast, the roll angle measured by polarization depends on the polarization state of the incident light and is relatively less affected by laser angle drift. Therefore, to improve the accuracy of the roll error in five-degree-of-freedom measurements at long distances, a polarization-based roll angle optical path is proposed. The path is used to measure the roll angle of the sensor at different positions, calculate the angle between the dual beams, and perform calibrations, thus improving measurement accuracy. The test results show that in the measurement range of 0.75?2.00 m, the roll angle measurement error after compensation is reduced by 88.97%. This meets the long-distance roll angle measurement requirements for high precision and easy installation.When dual beams are used to measure the roll angle, the angle between the double-collimated beams is easily disturbed by environmental changes, mechanical deformation, and other factors, which seriously affect measurement accuracy. The roll angle measured by the dual beam is sensitive to the spot position. By contrast, the roll angle measured by polarization depends on the polarization state of the incident light and is relatively less affected by laser angle drift. Therefore, to improve the accuracy of the roll error in five-degree-of-freedom measurements at long distances, a polarization-based roll angle optical path is proposed. The path is used to measure the roll angle of the sensor at different positions, calculate the angle between the dual beams, and perform calibrations, thus improving measurement accuracy. The test results show that in the measurement range of 0.75?2.00 m, the roll angle measurement error after compensation is reduced by 88.97%. This meets the long-distance roll angle measurement requirements for high precision and easy installation.
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2312002 (2023)
- DOI：10.3788/LOP223032
Absolute Measurement of Rectangular Flat Mirrors Using a Unidirectional Multi-Shifted Method
Guang Zhou, Weizheng Lei, Xiaohao Dong, and Jie Wang
With the development of advanced light sources, surface smoothness is becoming crucial for large-sized and high-precision mirrors. In this study, a two-dimensional (2D) absolute metrology method is developed that is based on the laser interferometer translational shear method for high aspect ratio rectangular flat mirrors used in X-rays. In this method, translational measurements in orthogonal directions are replaced with multiple shifted measurements in a single direction. In addition, a multi-matrix augmented zonal method reconstruction algorithm is derived to obtain the absolute surface shape of high-precision rectangular flat mirrors. Through simulations, the root mean square (RMS) of the residuals between the reconstructed absolute surface shape and initial surface shape is 0.03 nm (~λ/20000) without considering noise. Surface shape recovery under different numbers of translations and translation distances when considering Gaussian noise is then simulated and analyzed, and an experimental verification is performed for a rectangular plane mirror of 120 mm×40 mm. The measured absolute surface shape recovery is determined as 1.07 nm (λ/591) using the RMS of the absolute surface shape residual obtained by the three-plane method. Both simulations and experiments show that the proposed method can effectively obtain the 2D absolute surface shape of a high-precision rectangular planar mirror. Unidirectional translation thus lays the foundation for establishing multi-aperture stitching measurements based on absolute measurements.With the development of advanced light sources, surface smoothness is becoming crucial for large-sized and high-precision mirrors. In this study, a two-dimensional (2D) absolute metrology method is developed that is based on the laser interferometer translational shear method for high aspect ratio rectangular flat mirrors used in X-rays. In this method, translational measurements in orthogonal directions are replaced with multiple shifted measurements in a single direction. In addition, a multi-matrix augmented zonal method reconstruction algorithm is derived to obtain the absolute surface shape of high-precision rectangular flat mirrors. Through simulations, the root mean square (RMS) of the residuals between the reconstructed absolute surface shape and initial surface shape is 0.03 nm (~λ/20000) without considering noise. Surface shape recovery under different numbers of translations and translation distances when considering Gaussian noise is then simulated and analyzed, and an experimental verification is performed for a rectangular plane mirror of 120 mm×40 mm. The measured absolute surface shape recovery is determined as 1.07 nm (λ/591) using the RMS of the absolute surface shape residual obtained by the three-plane method. Both simulations and experiments show that the proposed method can effectively obtain the 2D absolute surface shape of a high-precision rectangular planar mirror. Unidirectional translation thus lays the foundation for establishing multi-aperture stitching measurements based on absolute measurements.
- Dec. 10, 2023
- Laser & Optoelectronics Progress
- Vol. 60, Issue 23, 2312001 (2023)
- DOI：10.3788/LOP222992
© Copyright 2018-2021 | Chinese Laser Press. All Rights Reserved 沪ICP备15018463号-20