Pub Date : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208050
V. Dyadichev, S. Stoyanchenko, A. Dyadichev
The paper presents a suggested complex of mathematical models for searching for the traffic structure in flexible manufacturing systems. It shows that the usage of the elaborated models allows increasing significantly the efficiency of the manufacturing systems. This paper offers the methodology of searching for a rational AMS traffic structure. The basis of the developed methodology is the combined method, which combines imitating and analytical modelling. The effectiveness of the offered approach is shown by the example of traffic optimization in the automatized flexible assembly system. The methodology offered in the paper allows considering the influence of incidental factors on the AMS functioning process.
{"title":"Mathematical Model for Traffic Optimization in Flexible Manufacturing Systems","authors":"V. Dyadichev, S. Stoyanchenko, A. Dyadichev","doi":"10.1109/RusAutoCon49822.2020.9208050","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208050","url":null,"abstract":"The paper presents a suggested complex of mathematical models for searching for the traffic structure in flexible manufacturing systems. It shows that the usage of the elaborated models allows increasing significantly the efficiency of the manufacturing systems. This paper offers the methodology of searching for a rational AMS traffic structure. The basis of the developed methodology is the combined method, which combines imitating and analytical modelling. The effectiveness of the offered approach is shown by the example of traffic optimization in the automatized flexible assembly system. The methodology offered in the paper allows considering the influence of incidental factors on the AMS functioning process.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133494927","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208052
A. Filimonov, N. Filimonov
the polyhedral methodology for solving problems of discrete terminal control by dynamic objects under resource constraints conditions is discussed. It is based on formalism of multidimensional optimization and three theoretical constructions like predictive extrapolation of controlled object’s movements, the immersion principle of the input optimization problem to a number of more simple computational problems algorithmically, and also the mechanism of extreme guidance. Due to polyhedral formalization, all the considered optimization problems have the polyhedral structure and are reduced to linear programming problems. As an example, the problem of advanced impulse control for stabilization of the lateral hovercraft’s movement is considered.
{"title":"The Polyhedral Methodology of Formalization of Discrete Terminal Control Processes with Resource Constraints","authors":"A. Filimonov, N. Filimonov","doi":"10.1109/RusAutoCon49822.2020.9208052","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208052","url":null,"abstract":"the polyhedral methodology for solving problems of discrete terminal control by dynamic objects under resource constraints conditions is discussed. It is based on formalism of multidimensional optimization and three theoretical constructions like predictive extrapolation of controlled object’s movements, the immersion principle of the input optimization problem to a number of more simple computational problems algorithmically, and also the mechanism of extreme guidance. Due to polyhedral formalization, all the considered optimization problems have the polyhedral structure and are reduced to linear programming problems. As an example, the problem of advanced impulse control for stabilization of the lateral hovercraft’s movement is considered.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130302564","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208067
S. V. Shalobanov, S. S. Shalobanov
The paper considers an algorithm for searching for defects of topological links between dynamic blocks: breaking one link between dynamic blocks. The use of well-known algorithms of this kind is complicated by the need to use models with trial deviations of the parameters, or the need to analyze the signs of signal transmissions. The disadvantage of the algorithm based on trial changes in the topological relationships of the model is that it uses the task of these changes in the model, which is a time-consuming task. The disadvantage of the method of analysis of transmission signs using a normalized diagnostic feature, as well as using a binary diagnostic feature, is that additional calculations of the transmission signs of the signals from the outputs of the blocks to the control points are necessary. Below, we consider an algorithm for searching for defects in the form of breaking links between dynamic blocks, based on the use of a topological sensitivity model of integral estimates of the object's output signals. The procedure for obtaining normalized diagnostic features and calculating a quantitative measure of the distinguishability of defects is determined.
{"title":"Diagnosing Continuous Dynamic Systems Using Topological Sensitivity Functions","authors":"S. V. Shalobanov, S. S. Shalobanov","doi":"10.1109/RusAutoCon49822.2020.9208067","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208067","url":null,"abstract":"The paper considers an algorithm for searching for defects of topological links between dynamic blocks: breaking one link between dynamic blocks. The use of well-known algorithms of this kind is complicated by the need to use models with trial deviations of the parameters, or the need to analyze the signs of signal transmissions. The disadvantage of the algorithm based on trial changes in the topological relationships of the model is that it uses the task of these changes in the model, which is a time-consuming task. The disadvantage of the method of analysis of transmission signs using a normalized diagnostic feature, as well as using a binary diagnostic feature, is that additional calculations of the transmission signs of the signals from the outputs of the blocks to the control points are necessary. Below, we consider an algorithm for searching for defects in the form of breaking links between dynamic blocks, based on the use of a topological sensitivity model of integral estimates of the object's output signals. The procedure for obtaining normalized diagnostic features and calculating a quantitative measure of the distinguishability of defects is determined.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116970347","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208159
A. Lapin, N. E. Zubov
This study deals with the problem of stabilizing spatial motion of a maneuverable aircraft with three controls and nine state parameters. In the task it is required to provide specified dynamic properties of the controlled plant at minimal control actions. For numeric matrices describing one of flight modes of a typical maneuverable aircraft the parameterized set of controllers by state vector is synthesized basing on the generalized Bass – Gura formula. This formula allows analytically calculating modal control for multi-input objects and simplifies parameterization of this control. The task of row-by-row minimization of gain matrix quadratic norm at fixed poles values is solved analytically by choosing the values of supplementary parameters which do not change the closed-loop control system spectrum. A technique for parameterization of two rows of three is proposed. In parameterized rows the minima are obtained at the condition of forced zeroing the coefficients corresponding to linear velocity. Based on simulation results the poles are chosen at which along with optimal supplementary parameters one may synthesize a controller allowing for specified range of initial conditions to obtain the most rapid transient with required dynamics and minimal control actions.
{"title":"Minimization of Control Signals at Stabilizing Spatial Motion of a Maneuverable Aircraft","authors":"A. Lapin, N. E. Zubov","doi":"10.1109/RusAutoCon49822.2020.9208159","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208159","url":null,"abstract":"This study deals with the problem of stabilizing spatial motion of a maneuverable aircraft with three controls and nine state parameters. In the task it is required to provide specified dynamic properties of the controlled plant at minimal control actions. For numeric matrices describing one of flight modes of a typical maneuverable aircraft the parameterized set of controllers by state vector is synthesized basing on the generalized Bass – Gura formula. This formula allows analytically calculating modal control for multi-input objects and simplifies parameterization of this control. The task of row-by-row minimization of gain matrix quadratic norm at fixed poles values is solved analytically by choosing the values of supplementary parameters which do not change the closed-loop control system spectrum. A technique for parameterization of two rows of three is proposed. In parameterized rows the minima are obtained at the condition of forced zeroing the coefficients corresponding to linear velocity. Based on simulation results the poles are chosen at which along with optimal supplementary parameters one may synthesize a controller allowing for specified range of initial conditions to obtain the most rapid transient with required dynamics and minimal control actions.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116252021","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208078
S. Zagolilo, A. Semenov, M. Semenova
The article presents a multimotor electric drive system of a roadheader. We made a computer model in the MatLab suite using a Simulink block library and the SimPowerSystems application. An AM-75 roadheader was chosen for the study. We provided a description of the circuit diagram of the primary electric equipment of the roadheader that includes motors of the boom-type implement, rotating flight conveyor, and the picker with two pallet handles. Additional parameters of asynchronous motors, such as active resistances and inductances of the stator and rotor winding, inductive coupling, reduced power, rated current, and structural and winding factors of motors were calculated for the modeling purposes using the Mathcad software. We developed a computer model of direct start of all the asynchronous motors with a production process-based interconnection. Modeling results are presented as time graphs of the primary motor parameters: angular frequency of revolution and electromagnetic torque. We obtained mains voltage and current graphs, as well as an active power consumption graph. The obtained results were qualitatively assessed by determining the relative error of the modeled parameters and the calculated data. In the process of evaluation and analysis of the modeling results, we identified insignificant errors in motor parameters. This indicates a faithful realization of the computer model and a possibility of using it for engineering calculations.
{"title":"Computer Modeling of a Multimotor Electric Drive System in the MatLab Suite","authors":"S. Zagolilo, A. Semenov, M. Semenova","doi":"10.1109/RusAutoCon49822.2020.9208078","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208078","url":null,"abstract":"The article presents a multimotor electric drive system of a roadheader. We made a computer model in the MatLab suite using a Simulink block library and the SimPowerSystems application. An AM-75 roadheader was chosen for the study. We provided a description of the circuit diagram of the primary electric equipment of the roadheader that includes motors of the boom-type implement, rotating flight conveyor, and the picker with two pallet handles. Additional parameters of asynchronous motors, such as active resistances and inductances of the stator and rotor winding, inductive coupling, reduced power, rated current, and structural and winding factors of motors were calculated for the modeling purposes using the Mathcad software. We developed a computer model of direct start of all the asynchronous motors with a production process-based interconnection. Modeling results are presented as time graphs of the primary motor parameters: angular frequency of revolution and electromagnetic torque. We obtained mains voltage and current graphs, as well as an active power consumption graph. The obtained results were qualitatively assessed by determining the relative error of the modeled parameters and the calculated data. In the process of evaluation and analysis of the modeling results, we identified insignificant errors in motor parameters. This indicates a faithful realization of the computer model and a possibility of using it for engineering calculations.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"30 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123116729","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208129
A. Martyshkin, D. Trokoz, I. I. Salnikov
This study is devoted to the challenges of motion planning for mobile robots with smart machine vision systems. Motion planning for mobile robots in the environment with obstacles is a problem to deal with when creating robots suitable for operation in real-world conditions. The solutions found today are predominantly private, and are highly specialized, which prevents judging of how successful they are in solving the problem of effective motion planning. The purpose of this study: develop and investigate a motion planning algorithm for a mobile robot with a smart machine vision system. The research subject for this article is a motion planning algorithm for a mobile robot with a smart machine vision system. This study provides a review of domestic and foreign mobile robots that solve the motion planning problem in a known environment with unknown obstacles. The following navigation methods are considered for mobile robots: local, global, individual. In the course of work and research, a mobile robot prototype has been built, capable of recognizing obstacles of regular geometric shapes, as well as plan and correct the movement path. Environment objects are identified and classified as obstacles by means of digital image processing methods and algorithms. Distance to the obstacle and relative angle are calculated by photogrammetry methods, image quality is improved by linear contrast enhancement and optimal linear filtering using the Wiener-Hopf equation. Virtual tools, related to mobile robot motion algorithm testing, have been reviewed, which led us to selecting Webots software package for prototype testing. Conclusions have been drawn regarding the concluded research.
{"title":"Development and Investigation of a Motion Planning Algorithm for a Mobile Robot with a Smart Machine Vision System","authors":"A. Martyshkin, D. Trokoz, I. I. Salnikov","doi":"10.1109/RusAutoCon49822.2020.9208129","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208129","url":null,"abstract":"This study is devoted to the challenges of motion planning for mobile robots with smart machine vision systems. Motion planning for mobile robots in the environment with obstacles is a problem to deal with when creating robots suitable for operation in real-world conditions. The solutions found today are predominantly private, and are highly specialized, which prevents judging of how successful they are in solving the problem of effective motion planning. The purpose of this study: develop and investigate a motion planning algorithm for a mobile robot with a smart machine vision system. The research subject for this article is a motion planning algorithm for a mobile robot with a smart machine vision system. This study provides a review of domestic and foreign mobile robots that solve the motion planning problem in a known environment with unknown obstacles. The following navigation methods are considered for mobile robots: local, global, individual. In the course of work and research, a mobile robot prototype has been built, capable of recognizing obstacles of regular geometric shapes, as well as plan and correct the movement path. Environment objects are identified and classified as obstacles by means of digital image processing methods and algorithms. Distance to the obstacle and relative angle are calculated by photogrammetry methods, image quality is improved by linear contrast enhancement and optimal linear filtering using the Wiener-Hopf equation. Virtual tools, related to mobile robot motion algorithm testing, have been reviewed, which led us to selecting Webots software package for prototype testing. Conclusions have been drawn regarding the concluded research.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"34 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124522217","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208192
I. Bolodurina, L. Zabrodina
In this paper, we consider the neural network approach to solving optimal control problems with mixed constraints at the stage of optimization of the functional approximations through evolutionary algorithms. Based on the necessary optimality conditions and the Lagrange multiplier method, the initial optimal control problem is reduced to a nonlinear optimization problem and the corresponding approximation model of the neural network for the control function and the trajectory is presented. The convergence of the neural network approach using a genetic algorithm, a population-based gravitational search algorithm, and a basic particle swarm algorithm was studied. Also, the results obtained are compared with the operation of the gradient descent algorithm. Computational experiments have shown that evolutionary algorithms for optimizing functions use the least number of iterations to achieve a given accuracy, but multi-agent methods of gravitational search and particle swarming show the longest execution time per iteration. The genetic optimization algorithm showed the fastest convergence rate relative to the total execution time of the algorithm.
{"title":"Investigation of the Evolutionary Optimization Algorithms for the Neural Network Solution of the Optimal Control Problems","authors":"I. Bolodurina, L. Zabrodina","doi":"10.1109/RusAutoCon49822.2020.9208192","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208192","url":null,"abstract":"In this paper, we consider the neural network approach to solving optimal control problems with mixed constraints at the stage of optimization of the functional approximations through evolutionary algorithms. Based on the necessary optimality conditions and the Lagrange multiplier method, the initial optimal control problem is reduced to a nonlinear optimization problem and the corresponding approximation model of the neural network for the control function and the trajectory is presented. The convergence of the neural network approach using a genetic algorithm, a population-based gravitational search algorithm, and a basic particle swarm algorithm was studied. Also, the results obtained are compared with the operation of the gradient descent algorithm. Computational experiments have shown that evolutionary algorithms for optimizing functions use the least number of iterations to achieve a given accuracy, but multi-agent methods of gravitational search and particle swarming show the longest execution time per iteration. The genetic optimization algorithm showed the fastest convergence rate relative to the total execution time of the algorithm.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124750187","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208042
A. Gusev, K. Andreev, V. Nikiforov
This article covers a specific end-to-end design technology. This technology is based on constructing a mathematical model of a uniaxial gyroscopic stabiliser (UGS) by transforming a physical model of an object created in SolidWorks using standard Matlab tools. Mathematical modelling was carried out to verify the stability of the obtained model. The main stages of end-to-end design (CAD-design) of a dynamic system are formed using the example of a uniaxial gyroscopic stabiliser. Conclusions were drawn on the application of the proposed technology in the design of dynamic systems and the synthesis of control laws. Application of the proposed end-to-end design technology will reduce the number of errors and the design time of dynamic systems and the control law building. An application for the operating system was developed to automate the end-to-end design process.
{"title":"End-to-end Terminal Control Design Technology","authors":"A. Gusev, K. Andreev, V. Nikiforov","doi":"10.1109/RusAutoCon49822.2020.9208042","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208042","url":null,"abstract":"This article covers a specific end-to-end design technology. This technology is based on constructing a mathematical model of a uniaxial gyroscopic stabiliser (UGS) by transforming a physical model of an object created in SolidWorks using standard Matlab tools. Mathematical modelling was carried out to verify the stability of the obtained model. The main stages of end-to-end design (CAD-design) of a dynamic system are formed using the example of a uniaxial gyroscopic stabiliser. Conclusions were drawn on the application of the proposed technology in the design of dynamic systems and the synthesis of control laws. Application of the proposed end-to-end design technology will reduce the number of errors and the design time of dynamic systems and the control law building. An application for the operating system was developed to automate the end-to-end design process.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123571427","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208160
D. D. Vasilev, K. Moiseev, S. V. Sidorova
The article considers the results of designing structure and development of the control algorithm of the shutter system for applying two-component ultrathin WSi films at the VUP-11M vacuum coating system with a rotating substrate holder. The algorithm of the operator’s operation in manual mode for obtaining two-component films is analyzed and its disadvantages are described. A procedure has been developed for combining the angle of the substrate in the program and the angle of the substrate in the vacuum chamber. The interface and algorithm of the program for controlling a stepper motor are described. The program allows to rotate the substrate and see its position on the equipment scheme from above. In the tab widget, the operator can calibrate the position of the substrate, rotate the substrate by a given angle, rotate the substrate around its axis or between certain angles for a specified time or number of times, and carry out the process of deposition of two-component films in automatic mode. The use of an automatic film deposition mode increases the reproducibility of film properties.
{"title":"Automation of the Process of Ultrathin Two-Component Films Deposition from Two Sputtering Sources","authors":"D. D. Vasilev, K. Moiseev, S. V. Sidorova","doi":"10.1109/RusAutoCon49822.2020.9208160","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208160","url":null,"abstract":"The article considers the results of designing structure and development of the control algorithm of the shutter system for applying two-component ultrathin WSi films at the VUP-11M vacuum coating system with a rotating substrate holder. The algorithm of the operator’s operation in manual mode for obtaining two-component films is analyzed and its disadvantages are described. A procedure has been developed for combining the angle of the substrate in the program and the angle of the substrate in the vacuum chamber. The interface and algorithm of the program for controlling a stepper motor are described. The program allows to rotate the substrate and see its position on the equipment scheme from above. In the tab widget, the operator can calibrate the position of the substrate, rotate the substrate by a given angle, rotate the substrate around its axis or between certain angles for a specified time or number of times, and carry out the process of deposition of two-component films in automatic mode. The use of an automatic film deposition mode increases the reproducibility of film properties.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122552588","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 : 2020-09-01DOI: 10.1109/RusAutoCon49822.2020.9208097
Polina Demochkina, A. Savchenko
In this paper, we address the problem of detecting small objects on high-quality X-ray imagesusing deep neural networks. We propose to implement the two-stage approach, in which, firstly, input image issplit into partially overlapping blocks to make small objects more discriminative for detection. Secondly, the small blocks are fed into conventional single-shot detectors. These detectors are trained using the blocks of the training images extracted by the same procedure.Two datasets of X-ray images from the customs inspection complex are examined in the experimental study. It was shown thatthe proposed algorithm with data augmentationleads tomore precise results when compared to the conventional technique:ourmethod outperforms the traditional approach by 5.4 - 25.7% depending on the type of used backbone convolutional neural network.
{"title":"Improving the Accuracy of One-Shot Detectors for Small Objects in X-ray Images","authors":"Polina Demochkina, A. Savchenko","doi":"10.1109/RusAutoCon49822.2020.9208097","DOIUrl":"https://doi.org/10.1109/RusAutoCon49822.2020.9208097","url":null,"abstract":"In this paper, we address the problem of detecting small objects on high-quality X-ray imagesusing deep neural networks. We propose to implement the two-stage approach, in which, firstly, input image issplit into partially overlapping blocks to make small objects more discriminative for detection. Secondly, the small blocks are fed into conventional single-shot detectors. These detectors are trained using the blocks of the training images extracted by the same procedure.Two datasets of X-ray images from the customs inspection complex are examined in the experimental study. It was shown thatthe proposed algorithm with data augmentationleads tomore precise results when compared to the conventional technique:ourmethod outperforms the traditional approach by 5.4 - 25.7% depending on the type of used backbone convolutional neural network.","PeriodicalId":101834,"journal":{"name":"2020 International Russian Automation Conference (RusAutoCon)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125267632","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：