Pub Date : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268337
Seongyun Park, Pyeongyeon Lee, Jeong-Joon Ahn, S. Park, Youngmi Kim, W. Na, Jonghoon Kim
Nowadays, the usage of lithium-ion batteries is an increase highly for electric vehicles (EVs), energy storage systems (ESSs), and portable electrical devices. The electrical characteristics of lithium-ion batteries are changed by discharge/charge current magnitudes, depth of discharge (DoD), environment temperature, degradation, and so on. In addition, the mechanical stress such as vibration and shock are degraded due to deformation of electrode or stress of electrolyte. The previous literatures of vibration are limited to the conditions of electric vehicle and satellites. These vibrations are tested on a single axis. However, earthquake vibration is applied to three axes simultaneously. A lot of literatures have analyzed the change of single cell's electrical characteristics with mathematical analysis method. In this paper, lithium-ion battery module which is consisted of 14 series and 20 parallel by 18650 cylindrical cells is tested to analyze the change of electrical characteristics such as cell-to-cell voltage difference, internal resistance, discharge capacity and temperature difference in module by the earthquake vibration.
{"title":"Electrical characteristics analysis of 18650 lithium-ion battery pack with the earthquake vibration condition","authors":"Seongyun Park, Pyeongyeon Lee, Jeong-Joon Ahn, S. Park, Youngmi Kim, W. Na, Jonghoon Kim","doi":"10.23919/ICCAS50221.2020.9268337","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268337","url":null,"abstract":"Nowadays, the usage of lithium-ion batteries is an increase highly for electric vehicles (EVs), energy storage systems (ESSs), and portable electrical devices. The electrical characteristics of lithium-ion batteries are changed by discharge/charge current magnitudes, depth of discharge (DoD), environment temperature, degradation, and so on. In addition, the mechanical stress such as vibration and shock are degraded due to deformation of electrode or stress of electrolyte. The previous literatures of vibration are limited to the conditions of electric vehicle and satellites. These vibrations are tested on a single axis. However, earthquake vibration is applied to three axes simultaneously. A lot of literatures have analyzed the change of single cell's electrical characteristics with mathematical analysis method. In this paper, lithium-ion battery module which is consisted of 14 series and 20 parallel by 18650 cylindrical cells is tested to analyze the change of electrical characteristics such as cell-to-cell voltage difference, internal resistance, discharge capacity and temperature difference in module by the earthquake vibration.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"140 1","pages":"360-364"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75822492","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268271
Dong-Jin Choi, Ji-hoon Han, Sang-Uk Park, Sun-Ki Hong
Maintenance of an industrial electric motor is very important. The most commonly used algorithm for deep learning motor diagnosis using deep learning is CNN, which is one of the representative supervised learning algorithms. However, the failure diagnosis algorithm made with the CNN algorithm is vulnerable to this data. For this reason, an algorithm that complements this has been proposed, and that is to use the RNN and K-means algorithms. The method using RNN has a cyclic neural network structure, so it can grasp the similarity of data. K-means also uses the Euclidean distance method to grasp the similarity between data and classify the data using it. Due to the characteristics of these two algorithms, even if a disturbance is an input, if the similarity of data is high, it is determined as similar data. In this paper, two algorithms were used to perform fault diagnosis and two experiments were conducted to understand the differences and characteristics of the two algorithms. As a result of experiment 1 classifying only normal failures, experiment 2 experimented by increasing the number of failures to be classified. In the case of RNN, the results of experiments 1 and 2 showed similar accuracy. However, in the case of the algorithm using K-means, the accuracy decreased as the number of classifications increased.
{"title":"Comparison of motor fault diagnosis performance using RNN and K-means for data with disturbance","authors":"Dong-Jin Choi, Ji-hoon Han, Sang-Uk Park, Sun-Ki Hong","doi":"10.23919/ICCAS50221.2020.9268271","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268271","url":null,"abstract":"Maintenance of an industrial electric motor is very important. The most commonly used algorithm for deep learning motor diagnosis using deep learning is CNN, which is one of the representative supervised learning algorithms. However, the failure diagnosis algorithm made with the CNN algorithm is vulnerable to this data. For this reason, an algorithm that complements this has been proposed, and that is to use the RNN and K-means algorithms. The method using RNN has a cyclic neural network structure, so it can grasp the similarity of data. K-means also uses the Euclidean distance method to grasp the similarity between data and classify the data using it. Due to the characteristics of these two algorithms, even if a disturbance is an input, if the similarity of data is high, it is determined as similar data. In this paper, two algorithms were used to perform fault diagnosis and two experiments were conducted to understand the differences and characteristics of the two algorithms. As a result of experiment 1 classifying only normal failures, experiment 2 experimented by increasing the number of failures to be classified. In the case of RNN, the results of experiments 1 and 2 showed similar accuracy. However, in the case of the algorithm using K-means, the accuracy decreased as the number of classifications increased.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"518 1","pages":"443-446"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82997339","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268390
Kanghyun Park, Hyeongkeun Lee, Hunmin Yang, Se-Yoon Oh
Despite the advances in deep learning, training instance segmentation models like convolutional neural networks still tend to depend on enormous training data that are expensive and require labor to annotation. To avoid labor-intensive procedure, synthetic data can be an alternative because it is easy to generate and automatically segmented. However, it is challenging to train instance segmentation model that perform well at real world using only synthetic data because of domain gap. It is wrong direction to put a lot of effort into solving these problems by making synthetic data more photorealistic. In this paper, we suggest how to learn the instance segmentation model using synthetic data with artificial distractors. The performance has been improved about 7% by adding flying distractors compared to original synthetic data.
{"title":"Improving Instance Segmentation using Synthetic Data with Artificial Distractors","authors":"Kanghyun Park, Hyeongkeun Lee, Hunmin Yang, Se-Yoon Oh","doi":"10.23919/ICCAS50221.2020.9268390","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268390","url":null,"abstract":"Despite the advances in deep learning, training instance segmentation models like convolutional neural networks still tend to depend on enormous training data that are expensive and require labor to annotation. To avoid labor-intensive procedure, synthetic data can be an alternative because it is easy to generate and automatically segmented. However, it is challenging to train instance segmentation model that perform well at real world using only synthetic data because of domain gap. It is wrong direction to put a lot of effort into solving these problems by making synthetic data more photorealistic. In this paper, we suggest how to learn the instance segmentation model using synthetic data with artificial distractors. The performance has been improved about 7% by adding flying distractors compared to original synthetic data.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"1 1","pages":"22-26"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82840285","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}
As recent research on deep learning methods has been actively conducted, a number of deep learning methods have been proposed. In this paper, we propose a method of removing the desired object from an image using generative adversarial networks(GANs) structure. We composed the network in which two GANs are fused. The first GAN erases the target object from the input image, and the second GAN generates an image that fills the empty space with the background. Through this network, we can erase the desired object from the input image and get an image with the erased part filled with the background without any object detection method. We show that the removal of people and vehicles from images of roads using the CityScapes Dataset.
{"title":"Object Removal and Inpainting from Image using Combined GANs","authors":"Jeongwon Pyo, Yuri Goncalves Rocha, Arpan Ghosh, Kwanghee Lee, Gun-Gyo In, Tae-Yong Kuc","doi":"10.23919/ICCAS50221.2020.9268330","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268330","url":null,"abstract":"As recent research on deep learning methods has been actively conducted, a number of deep learning methods have been proposed. In this paper, we propose a method of removing the desired object from an image using generative adversarial networks(GANs) structure. We composed the network in which two GANs are fused. The first GAN erases the target object from the input image, and the second GAN generates an image that fills the empty space with the background. Through this network, we can erase the desired object from the input image and get an image with the erased part filled with the background without any object detection method. We show that the removal of people and vehicles from images of roads using the CityScapes Dataset.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"55 1","pages":"1116-1119"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90774401","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268383
Martin Kosinka, Z. Slanina, Michal Petruzela, Vojtech Blazek
The article deals with the design and implementation of a control unit for energy flow management in Vehicle To Home systems. Vehicle To Home systems use energy from an electric car battery to power a smart home and store excess energy back into the electric vehicle’s battery. A car with a Chademo interface was chosen for the tests. The designed control unit consists of a single-board computer Raspberry Pi 3B +, a designed printed circuit board and an electricity meter with communication via Modbus protocol. The control unit connects the superior system of the smart house, the battery management system and the developed two-way converter enabling the connection of the electric vehicle to the energy infrastructure of the smart house.
{"title":"Control system for V2H applications","authors":"Martin Kosinka, Z. Slanina, Michal Petruzela, Vojtech Blazek","doi":"10.23919/ICCAS50221.2020.9268383","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268383","url":null,"abstract":"The article deals with the design and implementation of a control unit for energy flow management in Vehicle To Home systems. Vehicle To Home systems use energy from an electric car battery to power a smart home and store excess energy back into the electric vehicle’s battery. A car with a Chademo interface was chosen for the tests. The designed control unit consists of a single-board computer Raspberry Pi 3B +, a designed printed circuit board and an electricity meter with communication via Modbus protocol. The control unit connects the superior system of the smart house, the battery management system and the developed two-way converter enabling the connection of the electric vehicle to the energy infrastructure of the smart house.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"71 1","pages":"916-921"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89382722","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268208
Jae-Seong Yoon, Sanghyeon Bae, Tae-Yong Kuc
This paper studies the human recognition, tracking, and clustering method in an indoor environment using a 3D lidar sensor and discusses two major issues in clustering. The first problem is when the Euclidean distance-based clustering is used, where a wall and a person are frequently clustered into one object. The other issue is that there is some noise due to reflective materials such as glass or marble. In order to cluster objects and recognize humans in this environment, we proposed a pre-processing sequence module for clustering. The pre-processing module composed in 5 steps that can remove walls around the robot and reduce the point cloud noise. We embedded this whole process in the robot system and it works while the robot is in motion.
{"title":"Human Recognition and Tracking in Narrow Indoor Environment using 3D Lidar Sensor","authors":"Jae-Seong Yoon, Sanghyeon Bae, Tae-Yong Kuc","doi":"10.23919/ICCAS50221.2020.9268208","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268208","url":null,"abstract":"This paper studies the human recognition, tracking, and clustering method in an indoor environment using a 3D lidar sensor and discusses two major issues in clustering. The first problem is when the Euclidean distance-based clustering is used, where a wall and a person are frequently clustered into one object. The other issue is that there is some noise due to reflective materials such as glass or marble. In order to cluster objects and recognize humans in this environment, we proposed a pre-processing sequence module for clustering. The pre-processing module composed in 5 steps that can remove walls around the robot and reduce the point cloud noise. We embedded this whole process in the robot system and it works while the robot is in motion.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"73 5 1","pages":"978-981"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87771464","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268288
Myung-oh Kim, Useok Jeong, D. Choi, Duk-Yeon Lee, Bo-Hyeong Seo, Dong-Wook Lee
Continuum robots are utilized in various fields, such as surgical catheters, and used to assist human muscular strength. To make continuum robots more practical, it is essential to miniaturize or reduce their weights. For tendon-driven robots, the weight of motors is the most important factor to be concerned. Accordingly, we suggest a method to control the tendon-driven continuum robot with a single motor to lighten the weight, which needed more than two motors. However, it is difficult to control the tendon-driven robots with one motor to track the desired trajectory as the lengths of two or more tendons change when the shape of the tendon-driven robot changes. To overcome this issue, we designed a radius-variable pulley, using which the required lengths for respective tendons can be achieved when only a single motor is operating.
{"title":"Tendon-Driven Continuum Robot Systems with only A Single Motor and A Radius-Changing Pulley","authors":"Myung-oh Kim, Useok Jeong, D. Choi, Duk-Yeon Lee, Bo-Hyeong Seo, Dong-Wook Lee","doi":"10.23919/ICCAS50221.2020.9268288","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268288","url":null,"abstract":"Continuum robots are utilized in various fields, such as surgical catheters, and used to assist human muscular strength. To make continuum robots more practical, it is essential to miniaturize or reduce their weights. For tendon-driven robots, the weight of motors is the most important factor to be concerned. Accordingly, we suggest a method to control the tendon-driven continuum robot with a single motor to lighten the weight, which needed more than two motors. However, it is difficult to control the tendon-driven robots with one motor to track the desired trajectory as the lengths of two or more tendons change when the shape of the tendon-driven robot changes. To overcome this issue, we designed a radius-variable pulley, using which the required lengths for respective tendons can be achieved when only a single motor is operating.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"29 1","pages":"945-949"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76012889","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268237
Gyujin Na, Y. Eun
This paper addresses an active probing signal-based attack detection method for autonomous vehicular systems. Employing active probing signals for attack detection may become a common method for detecting replay attacks performed using prerecorded sensor data. Conventional replay attack detection methods usually operate by injecting active probing signals into the control inputs and simultaneously checking whether the effect appears on the output signals. When active probing signals are used in vehicular systems, they may change the vehicle acceleration and steering angle. The tracking performance can degrade; inspired by this issue, we develop an attack detection method employing disturbance observers. The attack detection method compensates for the effect of active probing signals and detects malicious attacks, including replay attacks. To validate the effectiveness of the proposed method, several simulations are carried out.
{"title":"Active Probing Signal-Based Attack Detection Method for Autonomous Vehicular Systems","authors":"Gyujin Na, Y. Eun","doi":"10.23919/ICCAS50221.2020.9268237","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268237","url":null,"abstract":"This paper addresses an active probing signal-based attack detection method for autonomous vehicular systems. Employing active probing signals for attack detection may become a common method for detecting replay attacks performed using prerecorded sensor data. Conventional replay attack detection methods usually operate by injecting active probing signals into the control inputs and simultaneously checking whether the effect appears on the output signals. When active probing signals are used in vehicular systems, they may change the vehicle acceleration and steering angle. The tracking performance can degrade; inspired by this issue, we develop an attack detection method employing disturbance observers. The attack detection method compensates for the effect of active probing signals and detects malicious attacks, including replay attacks. To validate the effectiveness of the proposed method, several simulations are carried out.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"40 1","pages":"53-59"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76150047","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}
Localization technology is essential for robots. The map created to recognize the location mainly contains metric information. However, in a changing environment, a Semantic Map containing Semantic object information is required a multi-modal sensor composed of multiple types and multiple sensors[RGBD, thermal, night vision, global shutter camera, microphone, 16 channel laser sensor(=Lidar)] was created for semantic information recognition and semantic map creation in various environments, and calibration was performed to integrate the coordinate system. After that, we introduce the method of generating the metric map according to the configuration of the multi-modal sensor. Also, we propose a method to obtain a single accurate location by integrating the location recognition results obtained from various maps. This can be used to specify the position of the semantic object. Finally, it can be expected that the semantic object and the semantic map information obtained through the multi-modal sensor can be used for various different sensor configurations and various types of robots.
{"title":"Study on multi-modal sensor system based sematic navigation map building","authors":"Gi-Deok Bae, Taeyoung Uhm, Young-Ho Choi, Junghwan Hwang","doi":"10.23919/ICCAS50221.2020.9268414","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268414","url":null,"abstract":"Localization technology is essential for robots. The map created to recognize the location mainly contains metric information. However, in a changing environment, a Semantic Map containing Semantic object information is required a multi-modal sensor composed of multiple types and multiple sensors[RGBD, thermal, night vision, global shutter camera, microphone, 16 channel laser sensor(=Lidar)] was created for semantic information recognition and semantic map creation in various environments, and calibration was performed to integrate the coordinate system. After that, we introduce the method of generating the metric map according to the configuration of the multi-modal sensor. Also, we propose a method to obtain a single accurate location by integrating the location recognition results obtained from various maps. This can be used to specify the position of the semantic object. Finally, it can be expected that the semantic object and the semantic map information obtained through the multi-modal sensor can be used for various different sensor configurations and various types of robots.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"23 1","pages":"1195-1197"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76534076","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 : 2020-10-13DOI: 10.23919/ICCAS50221.2020.9268299
Nuri Kim, Yunho Choi, Minjae Kang, Songhwai Oh
The goal of the optimal viewpoint path estimation is to generate a path to the optimal viewpoint location where the robot can best see the Point of Interest (POI). There are several learning-based methods to find an optimal viewpoint, but these methods are limited to a specific object POI and it is necessary to newly learn in a situation where a new POI is added, and not robust to the environment changes. In this paper, we propose an algorithm that generates a path to the optimal viewpoint by using the geometrical features of the environment in the situation where the target POI is in the field of view. This method makes it easy to add new POIs and is robust to environmental changes because it uses semantic and geometric information. We assume that the robot can make a simple estimation of the geometric characteristics of the surrounding environment by using pretrained networks or by using sensor values. We collected the Kwanjeong street dataset for testing our algorithm. In this dataset, the distance accuracy of our method to reach the optimal viewpoint of the POI achieved 81.8% and 70.9% for template matching accuracy.
{"title":"GOPE: Geometry-Aware Optimal Viewpoint Path Estimation Using a Monocular Camera","authors":"Nuri Kim, Yunho Choi, Minjae Kang, Songhwai Oh","doi":"10.23919/ICCAS50221.2020.9268299","DOIUrl":"https://doi.org/10.23919/ICCAS50221.2020.9268299","url":null,"abstract":"The goal of the optimal viewpoint path estimation is to generate a path to the optimal viewpoint location where the robot can best see the Point of Interest (POI). There are several learning-based methods to find an optimal viewpoint, but these methods are limited to a specific object POI and it is necessary to newly learn in a situation where a new POI is added, and not robust to the environment changes. In this paper, we propose an algorithm that generates a path to the optimal viewpoint by using the geometrical features of the environment in the situation where the target POI is in the field of view. This method makes it easy to add new POIs and is robust to environmental changes because it uses semantic and geometric information. We assume that the robot can make a simple estimation of the geometric characteristics of the surrounding environment by using pretrained networks or by using sensor values. We collected the Kwanjeong street dataset for testing our algorithm. In this dataset, the distance accuracy of our method to reach the optimal viewpoint of the POI achieved 81.8% and 70.9% for template matching accuracy.","PeriodicalId":6732,"journal":{"name":"2020 20th International Conference on Control, Automation and Systems (ICCAS)","volume":"91 1","pages":"1062-1067"},"PeriodicalIF":0.0,"publicationDate":"2020-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79606110","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：