Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101701
Chen Gang, Yang Jun
Global Navigation Satellite Systems (GNSSs) have become an almost indispensable part of our infrastructure and modern life, but in a specific environment have its limitations, or cannot be applied. Pseudollites are ground-based transmitters that can provide similar navigation measurements as GNSSs. GH-LPS developed by NUDTis a new pseudolite system allows centimeter-accuracy single-point positioning for a mobile or static device. Carrier phase observations are commonly used in high-precision positioning applications, due to their millimeter precisions. Any cycle slip, if remained undetected, would deteriorate the high accuracy positioning. Hence, it is crucial to handle the cycle slips in the carrier phase measurements for GH-LPS. Carrier phase-based RAIM (CRAIM) algorithm is proposed to detect and estimate small cycle slips. In order to evaluate the performance, a kinematic test was conducted. The experiment demonstrated that the CRAIM method can be used to solve the problem of small (e.g. 1 cycle) cycle slips and GH-LPS can provide single point positioning with centimeter-level accuracy and sub-centimeter level precision.
{"title":"Cycle slip detection and estimation for ground-based high-precision local positioning system","authors":"Chen Gang, Yang Jun","doi":"10.1109/ICEMI46757.2019.9101701","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101701","url":null,"abstract":"Global Navigation Satellite Systems (GNSSs) have become an almost indispensable part of our infrastructure and modern life, but in a specific environment have its limitations, or cannot be applied. Pseudollites are ground-based transmitters that can provide similar navigation measurements as GNSSs. GH-LPS developed by NUDTis a new pseudolite system allows centimeter-accuracy single-point positioning for a mobile or static device. Carrier phase observations are commonly used in high-precision positioning applications, due to their millimeter precisions. Any cycle slip, if remained undetected, would deteriorate the high accuracy positioning. Hence, it is crucial to handle the cycle slips in the carrier phase measurements for GH-LPS. Carrier phase-based RAIM (CRAIM) algorithm is proposed to detect and estimate small cycle slips. In order to evaluate the performance, a kinematic test was conducted. The experiment demonstrated that the CRAIM method can be used to solve the problem of small (e.g. 1 cycle) cycle slips and GH-LPS can provide single point positioning with centimeter-level accuracy and sub-centimeter level precision.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132232956","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101835
Kong Deshan, Jiang Dong, Zhao Yanchao, Liu Xukun, Wang Deyu
The magnetic levitation system has the advantages of small friction and non-contact. For better application, it is necessary to carry out in-depth analysis and research on the magnetic levitation system and to process the data of the suspension. Since the rejection magnetic levitation system has a better balance range than the suction type magnetic levitation system, a rejection magnetic levitation system is designed. The dynamic equation of the rejection magnetic levitation system is established by analysis and the simulink system simulation is carried out according to the established dynamic equation. The stability balance of the repulsion magnetic levitation system is proved. The measurement data of the magnetic levitation system under external vibration conditions were measured by experiments and field experiments, and the external interference signals were removed by wavelet analysis. The experimental results verify that the designed rejection magnetic suspension system can work stably and has strong impact resistance. This design analysis method provides a theoretical basis for the wide application of magnetic levitation systems
{"title":"Research on rejection magnetic levitation model simulation and wavelet analysis of its signal","authors":"Kong Deshan, Jiang Dong, Zhao Yanchao, Liu Xukun, Wang Deyu","doi":"10.1109/ICEMI46757.2019.9101835","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101835","url":null,"abstract":"The magnetic levitation system has the advantages of small friction and non-contact. For better application, it is necessary to carry out in-depth analysis and research on the magnetic levitation system and to process the data of the suspension. Since the rejection magnetic levitation system has a better balance range than the suction type magnetic levitation system, a rejection magnetic levitation system is designed. The dynamic equation of the rejection magnetic levitation system is established by analysis and the simulink system simulation is carried out according to the established dynamic equation. The stability balance of the repulsion magnetic levitation system is proved. The measurement data of the magnetic levitation system under external vibration conditions were measured by experiments and field experiments, and the external interference signals were removed by wavelet analysis. The experimental results verify that the designed rejection magnetic suspension system can work stably and has strong impact resistance. This design analysis method provides a theoretical basis for the wide application of magnetic levitation systems","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130122905","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101479
Liu Qi, Z. Hongyu
With the rapid development of virtual reality technology, virtual reality technology has been widely used in many fields, especially in the field of education and training. Traditional measurement training is subject to economics, repeatability, safety, etc. The independent practice and teaching interaction in the training process are not ideal. This paper proposes the application concept of integrating virtual reality technology into the measurement training system. The trainee can interact with the virtual reality environment in real time, effectively avoiding the waste of measurement resources and improving the safety and repeatability of measurement training. At the same time, the system can provide the super immersion feeling that the traditional training mode can not provide, so that the trainees can feel the real measurement environment, improve the field measurement support ability, and effectively enhance the training effect. For the multi-party joint guarantee measurement task, the distributed characteristics of the system can be utilized, so that the metering personnel can still perform joint training in the same training scenario in different regions, and realize systematic joint training and joint training.
{"title":"Application of virtual reality technology in distributed measurement training system","authors":"Liu Qi, Z. Hongyu","doi":"10.1109/ICEMI46757.2019.9101479","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101479","url":null,"abstract":"With the rapid development of virtual reality technology, virtual reality technology has been widely used in many fields, especially in the field of education and training. Traditional measurement training is subject to economics, repeatability, safety, etc. The independent practice and teaching interaction in the training process are not ideal. This paper proposes the application concept of integrating virtual reality technology into the measurement training system. The trainee can interact with the virtual reality environment in real time, effectively avoiding the waste of measurement resources and improving the safety and repeatability of measurement training. At the same time, the system can provide the super immersion feeling that the traditional training mode can not provide, so that the trainees can feel the real measurement environment, improve the field measurement support ability, and effectively enhance the training effect. For the multi-party joint guarantee measurement task, the distributed characteristics of the system can be utilized, so that the metering personnel can still perform joint training in the same training scenario in different regions, and realize systematic joint training and joint training.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"231 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131478281","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101645
Zhang Pengsong, Wang Danyi, Gu Zhifei, Xu Jinghao, Ning Juan, Jiang Shanping, Yang Linhua
Aiming at the testing requirement of the light source of KM6 large solar simulator, a set of integrated measuring system for ignition and optical performance of high power short-arc xenon lamp is developed, which is used for function verification of xenon lamp trigger and power supply control system, measurement of optical performance parameters such as luminous efficiency, light distribution curve and optical efficiency of xenon lamp. Thus it provides reliable data support for the selection of light source, optical system design and operation of solar simulator.
{"title":"Development of integrative measurement system for ignition and optical performance of high-power xenon lamp","authors":"Zhang Pengsong, Wang Danyi, Gu Zhifei, Xu Jinghao, Ning Juan, Jiang Shanping, Yang Linhua","doi":"10.1109/ICEMI46757.2019.9101645","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101645","url":null,"abstract":"Aiming at the testing requirement of the light source of KM6 large solar simulator, a set of integrated measuring system for ignition and optical performance of high power short-arc xenon lamp is developed, which is used for function verification of xenon lamp trigger and power supply control system, measurement of optical performance parameters such as luminous efficiency, light distribution curve and optical efficiency of xenon lamp. Thus it provides reliable data support for the selection of light source, optical system design and operation of solar simulator.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"51 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126580968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101766
Li Yunfei, Ren Shiwei, Wan PeiWei, Wu Jinjie, Wang Ji, Li Mengshi
At the 4W1B test station of the Synchrotron Radiation Source in Beijing, the composite loss correction factor of the free-radiation ionization chamber of the synchrotron radiation was measured by the 12KeV single-energy X-ray obtained from the twin crystal. The corresponding free air ionization chamber and the traditional method for measuring the composite loss correction factor are introduced. Under the condition of voltage saturation, the method of changing the thickness of the absorption sheet is used to change the dose rate and calculate the composite loss correction factor. The result is shown in The corrected share of the selected 2400V operating voltage is 1.000000048.
{"title":"Study on correction factor of synchrotron radiation X-ray free air ionization chamber composite loss","authors":"Li Yunfei, Ren Shiwei, Wan PeiWei, Wu Jinjie, Wang Ji, Li Mengshi","doi":"10.1109/ICEMI46757.2019.9101766","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101766","url":null,"abstract":"At the 4W1B test station of the Synchrotron Radiation Source in Beijing, the composite loss correction factor of the free-radiation ionization chamber of the synchrotron radiation was measured by the 12KeV single-energy X-ray obtained from the twin crystal. The corresponding free air ionization chamber and the traditional method for measuring the composite loss correction factor are introduced. Under the condition of voltage saturation, the method of changing the thickness of the absorption sheet is used to change the dose rate and calculate the composite loss correction factor. The result is shown in The corrected share of the selected 2400V operating voltage is 1.000000048.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130986984","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101633
Zhou Pengyue, An Zhenghua, Wu Jinjie, Bai Zhanguo, Guo Siming
In order to meet the ground calibration requirements of Gravitational wave high-energy Electromagnetic Counterpart All-sky Monitor (GECAM) satellite gamma ray detector (GRD), two sets of single-energy x-ray calibration devices were established in China Metrology Institute (Changping). One set is a single crystal calibration device, and LiF crystal can realize continuous emission of (8.1-37.9) keV single-energy X-ray; The other set is a dual crystal calibration device. Si220 crystal and Si551 crystal can realize continuous light emission of single energy x-ray in the energy range of (38-160) keV. The X-ray energy spectrum, energy resolution and energy response of the calibration device were studied by standard high-purity germanium detectors. The test results show that the energy range of the calibration device is: (8-160) keV; monochromaticity: energy resolution 3.1%@8.1keV, 10.2%@79.9keV; monochromatic luminous flux 649cts/s@38. 0keV,678cts/s@79.9keV,298cts/s@121.8keV; monochromatic light ratio 96%@38.0keV, 97%@79.9keV, 96%@121.8keV. Combined with the actual test results of the energy point of the GRD, it indicates that the calibration device can meet the calibration requirements of the GRD.
{"title":"Research on calibration device of GECAM satellite gamma ray detector","authors":"Zhou Pengyue, An Zhenghua, Wu Jinjie, Bai Zhanguo, Guo Siming","doi":"10.1109/ICEMI46757.2019.9101633","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101633","url":null,"abstract":"In order to meet the ground calibration requirements of Gravitational wave high-energy Electromagnetic Counterpart All-sky Monitor (GECAM) satellite gamma ray detector (GRD), two sets of single-energy x-ray calibration devices were established in China Metrology Institute (Changping). One set is a single crystal calibration device, and LiF crystal can realize continuous emission of (8.1-37.9) keV single-energy X-ray; The other set is a dual crystal calibration device. Si220 crystal and Si551 crystal can realize continuous light emission of single energy x-ray in the energy range of (38-160) keV. The X-ray energy spectrum, energy resolution and energy response of the calibration device were studied by standard high-purity germanium detectors. The test results show that the energy range of the calibration device is: (8-160) keV; monochromaticity: energy resolution 3.1%@8.1keV, 10.2%@79.9keV; monochromatic luminous flux 649cts/s@38. 0keV,678cts/s@79.9keV,298cts/s@121.8keV; monochromatic light ratio 96%@38.0keV, 97%@79.9keV, 96%@121.8keV. Combined with the actual test results of the energy point of the GRD, it indicates that the calibration device can meet the calibration requirements of the GRD.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133018961","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101506
Wang Xinpeng, Men Yabin, Wang Guangjie, Zhang Dongliang, Dong Tao
In order to study the occurrence regularity and detection method of ocean evaporation duct phenomena, and to evaluate the height and intensity parameters of the duct, a fast detection system for ocean evaporation duct is designed. The system consists of measuring buoy, receiving terminal and data processing software. Data acquisition of temperature, humidity and air pressure at different height is completed by the buoy based on ZigBee wireless networking technology. And they are transmitted to the receiving terminal by the communication radio inside the buoy body. After data quality control and calculation processing, the atmosphere modified refractive index profile is calculated by software, then the height and intensity parameters of evaporation duct can be obtained. The sea trial results show that the system can obtain the height and intensity parameters quickly in the case of evaporation duct appearance, and has the characteristics of high measurement efficiency, good practicability and stable operation.
{"title":"Design and implementation of fast detection system for ocean evaporation duct","authors":"Wang Xinpeng, Men Yabin, Wang Guangjie, Zhang Dongliang, Dong Tao","doi":"10.1109/ICEMI46757.2019.9101506","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101506","url":null,"abstract":"In order to study the occurrence regularity and detection method of ocean evaporation duct phenomena, and to evaluate the height and intensity parameters of the duct, a fast detection system for ocean evaporation duct is designed. The system consists of measuring buoy, receiving terminal and data processing software. Data acquisition of temperature, humidity and air pressure at different height is completed by the buoy based on ZigBee wireless networking technology. And they are transmitted to the receiving terminal by the communication radio inside the buoy body. After data quality control and calculation processing, the atmosphere modified refractive index profile is calculated by software, then the height and intensity parameters of evaporation duct can be obtained. The sea trial results show that the system can obtain the height and intensity parameters quickly in the case of evaporation duct appearance, and has the characteristics of high measurement efficiency, good practicability and stable operation.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133461282","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101454
Li Xingxin, Zhu Liqiang, Yu Zujun, W. Yanqin
Object detection plays an important role in intrusion detection of railway safety monitoring system. Generally, high-resolution image has to be down-sampled in order to reduce the amount of computation, resulting in missed detections of distant objects. This paper proposes an auxiliary input framework based on vanishing point (VP) detection to preserve the image resolution of distant objects in railway alarm region. We improved the VP detection network based on CNN classification, which consists of two branches, estimating the x-y coordinates respectively. Auxiliary input based on VP can improve the accuracy of target detection, especially for distant targets. Experiments on public data sets show that the proposed model overperforms single-branch model and regression model.
{"title":"Adaptive auxiliary input extraction based on vanishing point detection for distant object detection in high-resolution railway scene","authors":"Li Xingxin, Zhu Liqiang, Yu Zujun, W. Yanqin","doi":"10.1109/ICEMI46757.2019.9101454","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101454","url":null,"abstract":"Object detection plays an important role in intrusion detection of railway safety monitoring system. Generally, high-resolution image has to be down-sampled in order to reduce the amount of computation, resulting in missed detections of distant objects. This paper proposes an auxiliary input framework based on vanishing point (VP) detection to preserve the image resolution of distant objects in railway alarm region. We improved the VP detection network based on CNN classification, which consists of two branches, estimating the x-y coordinates respectively. Auxiliary input based on VP can improve the accuracy of target detection, especially for distant targets. Experiments on public data sets show that the proposed model overperforms single-branch model and regression model.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133132062","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101785
Cai Yaxing, W. Tao, Ma Pingxin
For on-orbit replacement of ORU module of traditional spacecraft, it is necessary to set the module to be replaced to the state of power off and disabled before replacement operation. The related transactions of module have been interrupted for a long time, and the normal function of spacecraft has been greatly affected. This paper presents an on-orbit hot swap detection method for ORU modules. The hot swap detection process is described from both communication links and electrical aspects. The method is used to realize live hot swap and plug-and-play of ORU modules in spacecraft, shorten the interruption time of module services, and provide a new plug-and-play detection method for the development of plug-and-play spacecraft.
{"title":"A hot swap detection method for ORU modules of spacecraft","authors":"Cai Yaxing, W. Tao, Ma Pingxin","doi":"10.1109/ICEMI46757.2019.9101785","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101785","url":null,"abstract":"For on-orbit replacement of ORU module of traditional spacecraft, it is necessary to set the module to be replaced to the state of power off and disabled before replacement operation. The related transactions of module have been interrupted for a long time, and the normal function of spacecraft has been greatly affected. This paper presents an on-orbit hot swap detection method for ORU modules. The hot swap detection process is described from both communication links and electrical aspects. The method is used to realize live hot swap and plug-and-play of ORU modules in spacecraft, shorten the interruption time of module services, and provide a new plug-and-play detection method for the development of plug-and-play spacecraft.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"64 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115334325","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-11-01DOI: 10.1109/ICEMI46757.2019.9101821
Jing Yongquan, Wu Tianshu, Li Jin, Zhang Zhijia, Gao Chao
With the development and maturity of deep learning algorithms, CNN have emerged in the field of computer vision. Image recognition is one of the important research directions in the field of computer vision. The traditional image recognition method is to extract features by constructing feature descriptors and then classify them by classifiers, such as gradient direction histogram and support vector machine. These methods generally have the problems of poor robustness and insufficient ability to extract features in complex application scenarios. At the same time, convolutional neural network has not been well applied in image recognition due to its large amount of computation and slow speed. With the development of GPU, the parallel computing capability has been greatly improved. This paper designs a GPU acceleration method for the driver’s seatbelt detection system based on CNN. The system is based on the Deconv-SSD target detection algorithm for vehicle detection, the Squeeze-YOLO algorithm for vehicle front windshield location, and the semantic segmentation for seat belt detection. Based on the characteristics of GPU, through the off-line merging bath normlization and convolution layer, Tensorrt model conversion technology to realize the GPU optimization speed. The results show that the proposed acceleration method can effectively improve the detection efficiency.
{"title":"GPU acceleration design method for driver’s seatbelt detection","authors":"Jing Yongquan, Wu Tianshu, Li Jin, Zhang Zhijia, Gao Chao","doi":"10.1109/ICEMI46757.2019.9101821","DOIUrl":"https://doi.org/10.1109/ICEMI46757.2019.9101821","url":null,"abstract":"With the development and maturity of deep learning algorithms, CNN have emerged in the field of computer vision. Image recognition is one of the important research directions in the field of computer vision. The traditional image recognition method is to extract features by constructing feature descriptors and then classify them by classifiers, such as gradient direction histogram and support vector machine. These methods generally have the problems of poor robustness and insufficient ability to extract features in complex application scenarios. At the same time, convolutional neural network has not been well applied in image recognition due to its large amount of computation and slow speed. With the development of GPU, the parallel computing capability has been greatly improved. This paper designs a GPU acceleration method for the driver’s seatbelt detection system based on CNN. The system is based on the Deconv-SSD target detection algorithm for vehicle detection, the Squeeze-YOLO algorithm for vehicle front windshield location, and the semantic segmentation for seat belt detection. Based on the characteristics of GPU, through the off-line merging bath normlization and convolution layer, Tensorrt model conversion technology to realize the GPU optimization speed. The results show that the proposed acceleration method can effectively improve the detection efficiency.","PeriodicalId":419168,"journal":{"name":"2019 14th IEEE International Conference on Electronic Measurement & Instruments (ICEMI)","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2019-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115498914","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：