Pub Date : 2020-07-01DOI: 10.1109/ACIRS49895.2020.9162625
Zhu Qin, Siyuan Chen, Xiliang Xu, Minghui Zhao
Smart mines are not only the ultimate transformation model of the traditional coal industry but the best approach to drive the intelligence of classical mines as well. To realize the transformation of mines from tradition to wisdom, this paper proposes the key technologies for building smart mines and discusses the construction of its systems. Furthermore, through the design of technology and system platform, the application technologies such as the Internet of Things, big data, artificial intelligence, etc. are organically integrated into the system construction of the smart mine, to achieve the overall intelligence of mine development, excavation, transportation, selection, safety production management, and also realize the safe, efficient and green mine production and operation objectives.
{"title":"Research on Key Technologies and System Construction of Smart Mine","authors":"Zhu Qin, Siyuan Chen, Xiliang Xu, Minghui Zhao","doi":"10.1109/ACIRS49895.2020.9162625","DOIUrl":"https://doi.org/10.1109/ACIRS49895.2020.9162625","url":null,"abstract":"Smart mines are not only the ultimate transformation model of the traditional coal industry but the best approach to drive the intelligence of classical mines as well. To realize the transformation of mines from tradition to wisdom, this paper proposes the key technologies for building smart mines and discusses the construction of its systems. Furthermore, through the design of technology and system platform, the application technologies such as the Internet of Things, big data, artificial intelligence, etc. are organically integrated into the system construction of the smart mine, to achieve the overall intelligence of mine development, excavation, transportation, selection, safety production management, and also realize the safe, efficient and green mine production and operation objectives.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"79 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"117085787","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 : 2020-07-01DOI: 10.1109/acirs49895.2020.9162600
Xu Zhang, Bin Xue, Feng Jing
Real-world object recognition is an important and difficult robot vision problem. In this paper, a real-world multi-angle and multi-attitude deformable object recognition method for robot system, named RCOR, is proposed based on lightweight deep multimodal distance learning (DMDL). (1) Deep multimodal convolutional neural network (DMCNN) is proposed to improve the transformation abilities of CNNs and enhance feature maps’ resolutions. (2) Deep distance metric learning (DDML) is presented to relieve the problem of lacking adequate labeled data and efficiently reduce redundancy. (3) To apply RCOR into embedded vision applications in real-world environment, a light weight DCNN, Mobile-XB, is proposed. Extensive experiments demonstrate that the proposed approach significantly outperforms state-of-the-arts. And it performs well on computationally limited platforms.
{"title":"Real-World Object Recognition for Robot Based on Lightweight Deep Multimodal Distance Learning","authors":"Xu Zhang, Bin Xue, Feng Jing","doi":"10.1109/acirs49895.2020.9162600","DOIUrl":"https://doi.org/10.1109/acirs49895.2020.9162600","url":null,"abstract":"Real-world object recognition is an important and difficult robot vision problem. In this paper, a real-world multi-angle and multi-attitude deformable object recognition method for robot system, named RCOR, is proposed based on lightweight deep multimodal distance learning (DMDL). (1) Deep multimodal convolutional neural network (DMCNN) is proposed to improve the transformation abilities of CNNs and enhance feature maps’ resolutions. (2) Deep distance metric learning (DDML) is presented to relieve the problem of lacking adequate labeled data and efficiently reduce redundancy. (3) To apply RCOR into embedded vision applications in real-world environment, a light weight DCNN, Mobile-XB, is proposed. Extensive experiments demonstrate that the proposed approach significantly outperforms state-of-the-arts. And it performs well on computationally limited platforms.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"171 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130147377","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 : 2020-07-01DOI: 10.1109/acirs49895.2020.9162609
Xu Zhang, Tao Xie, Liqun Ren, Linna Yang
Speckle noise interference and inshore ship detection have remained difficult problems associated with the detection of ships in SAR images. The segmentation method based on the Markov random field (MRF) model can classify the pixels of an image for target detection. However, the traditional MRF method is not robust against speckle noise and cannot use the high-level information of the image. To overcome these problems, this paper proposes a hybrid non-local MRF model based on superpixel segmentation for the detection of ship targets at sea and in dock areas in SAR images. Superpixel segmentation can segment an SAR image into blocks of pixels with similar attributes to suppress the effect of coherent speckle noise in the image. In addition, a hybrid non-local energy function is proposed, and it can comprehensively consider local and non-local information and greatly improve the accuracy of image segmentation. The experimental results show that the proposed method can improve the detection accuracy and significantly reduce the false alarm rate compared with the those of the traditional MRF method.
{"title":"Ship Detection Based on Superpixel-Level Hybrid Non-local MRF for SAR Imagery","authors":"Xu Zhang, Tao Xie, Liqun Ren, Linna Yang","doi":"10.1109/acirs49895.2020.9162609","DOIUrl":"https://doi.org/10.1109/acirs49895.2020.9162609","url":null,"abstract":"Speckle noise interference and inshore ship detection have remained difficult problems associated with the detection of ships in SAR images. The segmentation method based on the Markov random field (MRF) model can classify the pixels of an image for target detection. However, the traditional MRF method is not robust against speckle noise and cannot use the high-level information of the image. To overcome these problems, this paper proposes a hybrid non-local MRF model based on superpixel segmentation for the detection of ship targets at sea and in dock areas in SAR images. Superpixel segmentation can segment an SAR image into blocks of pixels with similar attributes to suppress the effect of coherent speckle noise in the image. In addition, a hybrid non-local energy function is proposed, and it can comprehensively consider local and non-local information and greatly improve the accuracy of image segmentation. The experimental results show that the proposed method can improve the detection accuracy and significantly reduce the false alarm rate compared with the those of the traditional MRF method.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130440195","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}
Given the inevitable existence of some static or dynamic obstacles in some local areas of the underground roadway and the characteristics of the roadway environment where the obstacles are located, this paper proposes a method of obstacle detection for underground UAV based on monocular vision. This method can extract obstacles from the complex tunnel background, and get the two-dimensional information of obstacles in the image after preprocessing and provide an important basis for the subsequent obstacle avoidance.
{"title":"Image Preprocessing of Obstacle Avoidance for Underground Unmanned Aerial Vehicle Based on Monocular Vision","authors":"Minghui Zhao, Dongbo Wang, Xiliang Xu, Siyuan Chen, Zhu Qin","doi":"10.1109/acirs49895.2020.9162606","DOIUrl":"https://doi.org/10.1109/acirs49895.2020.9162606","url":null,"abstract":"Given the inevitable existence of some static or dynamic obstacles in some local areas of the underground roadway and the characteristics of the roadway environment where the obstacles are located, this paper proposes a method of obstacle detection for underground UAV based on monocular vision. This method can extract obstacles from the complex tunnel background, and get the two-dimensional information of obstacles in the image after preprocessing and provide an important basis for the subsequent obstacle avoidance.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"1986 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130527951","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 : 2020-07-01DOI: 10.1109/ACIRS49895.2020.9162616
R. Zou, Shaoping Wang, Zhiling Wang, Pan Zhao, Pengfei Zhou
The traditional parking path planning method based on geometric curve has the characteristics of simple calculation and high real-time performance, but it needs high requirements for initial parking position. In practical application, when the vehicle deviates from the path, the automatic parking task will be terminated in order to ensure safety and prevent collision. In this paper, an improved method is proposed. Firstly, according to the location of the target point of the parking space, the reverse driving method is used; according to the collision constraints and the minimum parking space constraints, the parking starting area is generated, which solves the problem of high requirements for the parking initial position based on the planning method. Then, the parking area is sampled to generate a smooth parking area based on the method of state space sampling, which can realize the dynamic adjustment of the parking track and solve the problem that the vehicle cannot complete the parking when it deviates from the planned path. Finally, the simulation software is applied to verify the results. The experimental results show that the method is effective.
{"title":"A Reverse Planning Method of Autonomous Parking Path","authors":"R. Zou, Shaoping Wang, Zhiling Wang, Pan Zhao, Pengfei Zhou","doi":"10.1109/ACIRS49895.2020.9162616","DOIUrl":"https://doi.org/10.1109/ACIRS49895.2020.9162616","url":null,"abstract":"The traditional parking path planning method based on geometric curve has the characteristics of simple calculation and high real-time performance, but it needs high requirements for initial parking position. In practical application, when the vehicle deviates from the path, the automatic parking task will be terminated in order to ensure safety and prevent collision. In this paper, an improved method is proposed. Firstly, according to the location of the target point of the parking space, the reverse driving method is used; according to the collision constraints and the minimum parking space constraints, the parking starting area is generated, which solves the problem of high requirements for the parking initial position based on the planning method. Then, the parking area is sampled to generate a smooth parking area based on the method of state space sampling, which can realize the dynamic adjustment of the parking track and solve the problem that the vehicle cannot complete the parking when it deviates from the planned path. Finally, the simulation software is applied to verify the results. The experimental results show that the method is effective.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"331 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134330703","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 : 2020-07-01DOI: 10.1109/acirs49895.2020.9162612
Sriram Marisetty, Pavan Kumar Pothula, Pon Deepika, C. K. Vinay, Vikas Vazhayil, Madhav Rao
Mechanical drilling and perforation of bone is one of the most standard practices followed in neurosurgery, but is a highly specialised task performed by trained neurosurgeons. Craniotomy refers to removing part of the skull bone to access the underlying tissues, which is currently performed manually involving a series of precautionary procedures. The non-planar surface of the brain makes the manual drilling process over a series of marked points less accurate. Guided robotic drilling is one such alternative towards the contemporary craniotomy method that leads to increased accuracy, thereby minimizing permanent secondary damages to the patient under treatment. The paper proposes a design towards achieving the first step in building an autonomous neurosurgical tool for craniotomy practices over a non-planar surface. The proposed device caters towards reaching to a specified target location in the surgical space, and orient the drilling head system perpendicular to the specified point on the non-planar surface to perform needle insertion, or deep drilling. In addition, the proposed tool follows the guided path in a perpendicular orientation along the series of drilling points on a non-planar skull surface, where the skull surface is derived from IR imaging of the patient in the operative field. The real-time 3D model of the patient’s head is reconstructed post multiple scanning from the handheld portable IR camera. The developed autonomous drilling system is driven by the path planned and annotated by the neurosurgeon on the 3D model generated from the handheld IR camera in the intraoperative space. The autonomous drill positioning system is validated experimentally on a 3D printed human head, and the system placement, along with the orientation errors and accuracy in guided path showed promising results.
{"title":"System Design of an Automated Drilling Device for Neurosurgical Applications","authors":"Sriram Marisetty, Pavan Kumar Pothula, Pon Deepika, C. K. Vinay, Vikas Vazhayil, Madhav Rao","doi":"10.1109/acirs49895.2020.9162612","DOIUrl":"https://doi.org/10.1109/acirs49895.2020.9162612","url":null,"abstract":"Mechanical drilling and perforation of bone is one of the most standard practices followed in neurosurgery, but is a highly specialised task performed by trained neurosurgeons. Craniotomy refers to removing part of the skull bone to access the underlying tissues, which is currently performed manually involving a series of precautionary procedures. The non-planar surface of the brain makes the manual drilling process over a series of marked points less accurate. Guided robotic drilling is one such alternative towards the contemporary craniotomy method that leads to increased accuracy, thereby minimizing permanent secondary damages to the patient under treatment. The paper proposes a design towards achieving the first step in building an autonomous neurosurgical tool for craniotomy practices over a non-planar surface. The proposed device caters towards reaching to a specified target location in the surgical space, and orient the drilling head system perpendicular to the specified point on the non-planar surface to perform needle insertion, or deep drilling. In addition, the proposed tool follows the guided path in a perpendicular orientation along the series of drilling points on a non-planar skull surface, where the skull surface is derived from IR imaging of the patient in the operative field. The real-time 3D model of the patient’s head is reconstructed post multiple scanning from the handheld portable IR camera. The developed autonomous drilling system is driven by the path planned and annotated by the neurosurgeon on the 3D model generated from the handheld IR camera in the intraoperative space. The autonomous drill positioning system is validated experimentally on a 3D printed human head, and the system placement, along with the orientation errors and accuracy in guided path showed promising results.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133812737","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 : 2020-07-01DOI: 10.1109/acirs49895.2020.9162599
A. Cheraghi, Asma Ben Janete, Kalman Graffi
Flocking behavior in swarms is phenomenal. Many small entities are interacting to form a large shape and to protect themselves from bigger entities. An examples are fish, they become a flock to be safe from predators. The way of how flocks move without getting separated, having a constant speed and not interfering each other is known as the flocking problem. In this paper, we build up a the three-zones flocking model on a two dimensional triangular graph to solve the flocking problem and evaluate the impact of changes in the separation, alignment and cohesion zone sizes on the flocking behavior.
{"title":"Robot Swarm Flocking on a 2D Triangular Graph","authors":"A. Cheraghi, Asma Ben Janete, Kalman Graffi","doi":"10.1109/acirs49895.2020.9162599","DOIUrl":"https://doi.org/10.1109/acirs49895.2020.9162599","url":null,"abstract":"Flocking behavior in swarms is phenomenal. Many small entities are interacting to form a large shape and to protect themselves from bigger entities. An examples are fish, they become a flock to be safe from predators. The way of how flocks move without getting separated, having a constant speed and not interfering each other is known as the flocking problem. In this paper, we build up a the three-zones flocking model on a two dimensional triangular graph to solve the flocking problem and evaluate the impact of changes in the separation, alignment and cohesion zone sizes on the flocking behavior.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125604880","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 : 2020-07-01DOI: 10.1109/acirs49895.2020.9162620
Max Paul Bauer, Anthony Ngo, Michael M. Resch
Compared to existing intelligent vehicles software such as Advanced Driver Assistance Systems (ADAS) the complexity of Autonomous Driving (AD) software exploded. As a result of that also the development and testing complexity increased proportionally. To master this challenge modern simulation based testing approaches are an essential element. Some of the main challenges for the underlying simulation platform are the large set of simulation use cases, the huge amount of simulation runs and the great testing depth even to the level of unit tests. To find an adequate architectural paradigm for simulation platforms we first analyse the simulation requirements. From this perspective we compare simulation platforms with the established Operating System (OS) Kernel platform. We propose a concept that maps terminology of such simulation kernel and the general architecture itself to an AD simulation platform. Finally, we discuss how design principles known from operating system kernels can be reused for the development of sustainable and modular simulation platforms.
{"title":"Usage of Operating System Kernel Paradigms to Build Modern Simulation Platforms for Autonomous Driving","authors":"Max Paul Bauer, Anthony Ngo, Michael M. Resch","doi":"10.1109/acirs49895.2020.9162620","DOIUrl":"https://doi.org/10.1109/acirs49895.2020.9162620","url":null,"abstract":"Compared to existing intelligent vehicles software such as Advanced Driver Assistance Systems (ADAS) the complexity of Autonomous Driving (AD) software exploded. As a result of that also the development and testing complexity increased proportionally. To master this challenge modern simulation based testing approaches are an essential element. Some of the main challenges for the underlying simulation platform are the large set of simulation use cases, the huge amount of simulation runs and the great testing depth even to the level of unit tests. To find an adequate architectural paradigm for simulation platforms we first analyse the simulation requirements. From this perspective we compare simulation platforms with the established Operating System (OS) Kernel platform. We propose a concept that maps terminology of such simulation kernel and the general architecture itself to an AD simulation platform. Finally, we discuss how design principles known from operating system kernels can be reused for the development of sustainable and modular simulation platforms.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"212 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114997494","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 : 2020-07-01DOI: 10.1109/ACIRS49895.2020.9162604
Yuxin Zheng, Runfeng Zhang, Xiaohan Yuan
The elevator car wall parameters are the pivot of parameters during the elevator design. To meet the individualized requirements of elevator design, and reduce the labor cost during design, the parametric design of elevator car wall model has important research significance in the design of elevator car. In this paper, genetic algorithm is utilized to optimize the internal parameters of the support vector machine method in order to establish the elevator car wall parameter prediction model based on GA-SVM. Based on the actual design data of Shanghai General Elevator Company over the years, 100 sets of data were simulated and predicted. The experimental results indicate that the average absolute percentage error of GA-SVM method is only 0.92%, and the relative error is 2.62%. The prediction accuracy is much better than BP neural network method. Most importantly, the GA-SVM method can effectively reduce the traditional labor cost of the elevator car design. Therefore, it is of great significance to the simulation design and manufacturing of the elevator car prototype.
{"title":"Parametric Design of Elevator Car Wall Based on GA-SVM Method","authors":"Yuxin Zheng, Runfeng Zhang, Xiaohan Yuan","doi":"10.1109/ACIRS49895.2020.9162604","DOIUrl":"https://doi.org/10.1109/ACIRS49895.2020.9162604","url":null,"abstract":"The elevator car wall parameters are the pivot of parameters during the elevator design. To meet the individualized requirements of elevator design, and reduce the labor cost during design, the parametric design of elevator car wall model has important research significance in the design of elevator car. In this paper, genetic algorithm is utilized to optimize the internal parameters of the support vector machine method in order to establish the elevator car wall parameter prediction model based on GA-SVM. Based on the actual design data of Shanghai General Elevator Company over the years, 100 sets of data were simulated and predicted. The experimental results indicate that the average absolute percentage error of GA-SVM method is only 0.92%, and the relative error is 2.62%. The prediction accuracy is much better than BP neural network method. Most importantly, the GA-SVM method can effectively reduce the traditional labor cost of the elevator car design. Therefore, it is of great significance to the simulation design and manufacturing of the elevator car prototype.","PeriodicalId":293428,"journal":{"name":"2020 5th Asia-Pacific Conference on Intelligent Robot Systems (ACIRS)","volume":"287 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2020-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115213626","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：