Pub Date : 2011-06-05DOI: 10.1109/IVS.2011.5940541
Y. Li, N. Haas, Sharath Pankanti
This paper describes our recent work on developing an intelligent headlight control system using machine learning-based approaches. Specifically, such a system aims to automatically control a vehicle's beam state (high beam or low beam) during a night-time drive based on the detection of oncoming/overtaking/leading traffics as well as urban areas from the videos captured by a camera. Two machine learning-based approaches, namely, support vector machine (SVM) and AdaBoost, have been applied to accomplish this task. The architect of each approach, as well as its detailed processing modules, will be elaborated in the paper. The system has been extensively tested both online and offline to validate the robustness and effectiveness of the two proposed approaches. A detailed performance study along with some comparisons between the two approaches will be reported at the end.
{"title":"Intelligent headlight control using learning-based approaches","authors":"Y. Li, N. Haas, Sharath Pankanti","doi":"10.1109/IVS.2011.5940541","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940541","url":null,"abstract":"This paper describes our recent work on developing an intelligent headlight control system using machine learning-based approaches. Specifically, such a system aims to automatically control a vehicle's beam state (high beam or low beam) during a night-time drive based on the detection of oncoming/overtaking/leading traffics as well as urban areas from the videos captured by a camera. Two machine learning-based approaches, namely, support vector machine (SVM) and AdaBoost, have been applied to accomplish this task. The architect of each approach, as well as its detailed processing modules, will be elaborated in the paper. The system has been extensively tested both online and offline to validate the robustness and effectiveness of the two proposed approaches. A detailed performance study along with some comparisons between the two approaches will be reported at the end.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"23 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114028439","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940579
J. Ferreira, V. Monteiro, J. Afonso, A. Silva
In this work is proposed the design of a system to create and handle Electric Vehicles (EV) charging procedures, based on intelligent process. Due to the electrical power distribution network limitation and absence of smart meter devices, Electric Vehicles charging should be performed in a balanced way, taking into account past experience, weather information based on data mining, and simulation approaches. In order to allow information exchange and to help user mobility, it was also created a mobile application to assist the EV driver on these processes. This proposed Smart Electric Vehicle Charging System uses Vehicle-to-Grid (V2G) technology, in order to connect Electric Vehicles and also renewable energy sources to Smart Grids (SG). This system also explores the new paradigm of Electrical Markets (EM), with deregulation of electricity production and use, in order to obtain the best conditions for commercializing electrical energy.
{"title":"Smart electric vehicle charging system","authors":"J. Ferreira, V. Monteiro, J. Afonso, A. Silva","doi":"10.1109/IVS.2011.5940579","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940579","url":null,"abstract":"In this work is proposed the design of a system to create and handle Electric Vehicles (EV) charging procedures, based on intelligent process. Due to the electrical power distribution network limitation and absence of smart meter devices, Electric Vehicles charging should be performed in a balanced way, taking into account past experience, weather information based on data mining, and simulation approaches. In order to allow information exchange and to help user mobility, it was also created a mobile application to assist the EV driver on these processes. This proposed Smart Electric Vehicle Charging System uses Vehicle-to-Grid (V2G) technology, in order to connect Electric Vehicles and also renewable energy sources to Smart Grids (SG). This system also explores the new paradigm of Electrical Markets (EM), with deregulation of electricity production and use, in order to obtain the best conditions for commercializing electrical energy.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114807035","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940581
Junnian Wang, Qingnian Wang, Chuanxue Song, Liang Chu, Yingying Wang
Drive torque coordinated control for four wheel independent-drive electric vehicle with differential drive assisted steering system (DDAS) is presented in this paper. Firstly, the technical detail of DDAS system is reviewed. Then a structure composed of two layers coordinating control loop is proposed. Moreover, the control algorithm of each layer is introduced. At last, the simulations of parking at stand still and double-lane-change maneuvering in several different adhesion conditions are performed. The simulation results show that the coordinating control system can ensure the DDAS system to work correctly and correct the bad influence of DDAS system on vehicle stability.
{"title":"Coordinated control of differential drive assisted steering system with vehicle stability enhancement system","authors":"Junnian Wang, Qingnian Wang, Chuanxue Song, Liang Chu, Yingying Wang","doi":"10.1109/IVS.2011.5940581","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940581","url":null,"abstract":"Drive torque coordinated control for four wheel independent-drive electric vehicle with differential drive assisted steering system (DDAS) is presented in this paper. Firstly, the technical detail of DDAS system is reviewed. Then a structure composed of two layers coordinating control loop is proposed. Moreover, the control algorithm of each layer is introduced. At last, the simulations of parking at stand still and double-lane-change maneuvering in several different adhesion conditions are performed. The simulation results show that the coordinating control system can ensure the DDAS system to work correctly and correct the bad influence of DDAS system on vehicle stability.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114873738","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940521
Benjamin Ranft, Timo Schönwald, B. Kitt
Many advanced driver assistance systems (ADAS) and autonomous vehicles require 3D information available from (stereo) camera systems. The corresponding task of estimating disparity or optical flow is computationally demanding, so meeting real-time update rates at high image resolutions has proven to be challenging. Modern parallel hardware seems suitable for this task only if its processing power can be efficiently accessed by parallel software implementations. In this paper we present a case study comparing different hardware platforms by two variants of block matching-based estimation. These platforms include two x86-compatible multicore systems, a graphics processing unit (GPU) and a 64-core embedded design. We introduce relevant features of each architecture and describe their effects on the applied algorithms, parallelization approaches and implementations. Target platforms are evaluated concerning computational performance, energy efficiency and programmer productivity.
{"title":"Parallel matching-based estimation - a case study on three different hardware architectures","authors":"Benjamin Ranft, Timo Schönwald, B. Kitt","doi":"10.1109/IVS.2011.5940521","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940521","url":null,"abstract":"Many advanced driver assistance systems (ADAS) and autonomous vehicles require 3D information available from (stereo) camera systems. The corresponding task of estimating disparity or optical flow is computationally demanding, so meeting real-time update rates at high image resolutions has proven to be challenging. Modern parallel hardware seems suitable for this task only if its processing power can be efficiently accessed by parallel software implementations. In this paper we present a case study comparing different hardware platforms by two variants of block matching-based estimation. These platforms include two x86-compatible multicore systems, a graphics processing unit (GPU) and a 64-core embedded design. We introduce relevant features of each architecture and describe their effects on the applied algorithms, parallelization approaches and implementations. Target platforms are evaluated concerning computational performance, energy efficiency and programmer productivity.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"371 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122152882","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940558
Philip Lenz, Julius Ziegler, Andreas Geiger, Martin Roser
Modern driver assistance systems such as collision avoidance or intersection assistance need reliable information on the current environment. Extracting such information from camera-based systems is a complex and challenging task for inner city traffic scenarios. This paper presents an approach for object detection utilizing sparse scene flow. For consecutive stereo images taken from a moving vehicle, corresponding interest points are extracted. Thus, for every interest point, disparity and optical flow values are known and consequently, scene flow can be calculated. Adjacent interest points describing a similar scene flow are considered to belong to one rigid object. The proposed method does not rely on object classes and allows for a robust detection of dynamic objects in traffic scenes. Leading vehicles are continuously detected for several frames. Oncoming objects are detected within five frames after their appearance.
{"title":"Sparse scene flow segmentation for moving object detection in urban environments","authors":"Philip Lenz, Julius Ziegler, Andreas Geiger, Martin Roser","doi":"10.1109/IVS.2011.5940558","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940558","url":null,"abstract":"Modern driver assistance systems such as collision avoidance or intersection assistance need reliable information on the current environment. Extracting such information from camera-based systems is a complex and challenging task for inner city traffic scenarios. This paper presents an approach for object detection utilizing sparse scene flow. For consecutive stereo images taken from a moving vehicle, corresponding interest points are extracted. Thus, for every interest point, disparity and optical flow values are known and consequently, scene flow can be calculated. Adjacent interest points describing a similar scene flow are considered to belong to one rigid object. The proposed method does not rely on object classes and allows for a robust detection of dynamic objects in traffic scenes. Leading vehicles are continuously detected for several frames. Oncoming objects are detected within five frames after their appearance.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124206330","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940573
M. Saffarian, R. Happee
A car following assisting system named as Rear Window Notification Display (RWND) is developed in order to assist drivers to interact effectively with vehicles equipped with Cooperative Adaptive Cruise Control (CACC) systems. The interface quantifies the acceleration of the instrumented lead car and the following distance in an intuitive way for the human driver on the rear window of the leader. Results of tests with human subjects in a driving simulator indicate that this interface reduces time headway and decreases the variance in time headway that drivers adopt, especially in manoeuvres that involve short term speed changes of the leading car. This system can accelerate the introduction of cooperative driving due to its effectiveness with low penetration rates.
{"title":"Supporting drivers in car following: A step towards cooperative driving","authors":"M. Saffarian, R. Happee","doi":"10.1109/IVS.2011.5940573","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940573","url":null,"abstract":"A car following assisting system named as Rear Window Notification Display (RWND) is developed in order to assist drivers to interact effectively with vehicles equipped with Cooperative Adaptive Cruise Control (CACC) systems. The interface quantifies the acceleration of the instrumented lead car and the following distance in an intuitive way for the human driver on the rear window of the leader. Results of tests with human subjects in a driving simulator indicate that this interface reduces time headway and decreases the variance in time headway that drivers adopt, especially in manoeuvres that involve short term speed changes of the leading car. This system can accelerate the introduction of cooperative driving due to its effectiveness with low penetration rates.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124213160","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940481
Nizar Chatti, A. Gehin, B. O. Bouamama, R. Merzouki
A supervision system oversees the operating state of an autonomous vehicle through the availability of the functions and services provided by the vehicle components. However, functional representations do not take into account the system dynamic behaviour and suffer from subjective definitions. For this reason, Bond Graph Models, through their behavioural, structural and causal properties are used to overcome the limitations of functional models. We obtain a new tool taking the form of a finite automaton, to provide fault identification and the reconfiguration conditions of a system. Each operating mode, corresponding to a vertex of the automaton, is associated with a set of services from a functional point of view and is defined accurately by a behavioral bond graph model. Furthermore, the service availability (associated to the Bond Graph elements) and the conditions for passage from one mode to another are analysed by fault detection and isolation algorithms generated on the basis of the structural and causal properties of the bond graph tool. The proposed approach is illustrated by a traction system of an intelligent and autonomous vehicle.
{"title":"Online supervision of intelligent vehicle using functional and behavioral models","authors":"Nizar Chatti, A. Gehin, B. O. Bouamama, R. Merzouki","doi":"10.1109/IVS.2011.5940481","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940481","url":null,"abstract":"A supervision system oversees the operating state of an autonomous vehicle through the availability of the functions and services provided by the vehicle components. However, functional representations do not take into account the system dynamic behaviour and suffer from subjective definitions. For this reason, Bond Graph Models, through their behavioural, structural and causal properties are used to overcome the limitations of functional models. We obtain a new tool taking the form of a finite automaton, to provide fault identification and the reconfiguration conditions of a system. Each operating mode, corresponding to a vertex of the automaton, is associated with a set of services from a functional point of view and is defined accurately by a behavioral bond graph model. Furthermore, the service availability (associated to the Bond Graph elements) and the conditions for passage from one mode to another are analysed by fault detection and isolation algorithms generated on the basis of the structural and causal properties of the bond graph tool. The proposed approach is illustrated by a traction system of an intelligent and autonomous vehicle.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125261553","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940528
M. Matilainen, A. Tuononen
Intelligent vehicles require more accurate knowledge of tire-road friction conditions in order to utilize the full potential of each tire. Continuous and reliable friction potential estimation would have a very positive traffic safety impact, because, e.g., adaptive cruise control could adjust to an adequate safety distance and a collision mitigation system could expedite intervention in slippery conditions. This paper presents how tire friction potential can be determined from individually measured tie rod forces. The method is independent of tire tread stiffness, tire slip angle, or cornering stiffness. In addition, left- and right-hand-side front tire friction potentials can be estimated separately, which would be very beneficial in μ-split conditions.
{"title":"Tire friction potential estimation from measured tie rod forces","authors":"M. Matilainen, A. Tuononen","doi":"10.1109/IVS.2011.5940528","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940528","url":null,"abstract":"Intelligent vehicles require more accurate knowledge of tire-road friction conditions in order to utilize the full potential of each tire. Continuous and reliable friction potential estimation would have a very positive traffic safety impact, because, e.g., adaptive cruise control could adjust to an adequate safety distance and a collision mitigation system could expedite intervention in slippery conditions. This paper presents how tire friction potential can be determined from individually measured tie rod forces. The method is independent of tire tread stiffness, tire slip angle, or cornering stiffness. In addition, left- and right-hand-side front tire friction potentials can be estimated separately, which would be very beneficial in μ-split conditions.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129348698","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940500
T. Menard, Jeffrey Miller
In this paper, we present a comparison between the Apple iPhone 3G™ [2] and the iPhone 4™ [2] using the real-time vehicle tracking application FreeSim_Mobile [24]. The built-in GPS receiver and web capabilities of the iPhone™, coupled with a V2I architecture, are used to send a continuous flow of data to a central server for processing by FreeSim [13–15], which is a real-time traffic simulator. The proportional model algorithm [18] is then used to determine the time to traverse a roadway in order to report in real-time the current flow of traffic. At the University of Alaska Anchorage, we currently have vehicle tracking devices installed in 80 probe vehicles that traverse the Anchorage area. Due to the high cost associated with vehicle tracking devices, it is difficult to penetrate a large vehicular network on a finite amount of money, so we must look towards other available technologies, such as the constantly-expanding cellular network. In this paper we look at the iPhone 4™ capability of reporting accurate and reliable locations and compare it to the recent study of the iPhone 3G™ [24]. Drivers equipped with an iPhone 4™ cellular phone and a vehicle tracking device manually timed how long it took to travel along a 0.99 mile/1.59 kilometer section of roadway. The vehicle tracking device and the iPhone 4™ report speed and location every 10 seconds whereas the iPhone 3G™ reported every 8 seconds [24]. From this data, we calculated the amount of time to traverse the test section of roadway using the proportional model algorithm [18] and compared it to the actual amount of time it took to traverse the test section of roadway as manually timed. We found that the vehicle tracking device had an average error factor of 4.94% from the actual time to traverse the section of roadway (as determined by the stopwatch), whereas the iPhone 4™ was found to have an error factor of 1.10%. The outcome of the case study is used to determine that the iPhone 4™ has higher accuracy than a vehicle tracking device, though it is important to note that the iPhone™ is more limited than a device attached to a vehicle since it can only report its location. If paired with another third party OBD device, however, it can also send the same information as a vehicle tracking device.
{"title":"Comparing the GPS capabilities of the iPhone 4 and iPhone 3G for vehicle tracking using FreeSim_Mobile","authors":"T. Menard, Jeffrey Miller","doi":"10.1109/IVS.2011.5940500","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940500","url":null,"abstract":"In this paper, we present a comparison between the Apple iPhone 3G™ [2] and the iPhone 4™ [2] using the real-time vehicle tracking application FreeSim_Mobile [24]. The built-in GPS receiver and web capabilities of the iPhone™, coupled with a V2I architecture, are used to send a continuous flow of data to a central server for processing by FreeSim [13–15], which is a real-time traffic simulator. The proportional model algorithm [18] is then used to determine the time to traverse a roadway in order to report in real-time the current flow of traffic. At the University of Alaska Anchorage, we currently have vehicle tracking devices installed in 80 probe vehicles that traverse the Anchorage area. Due to the high cost associated with vehicle tracking devices, it is difficult to penetrate a large vehicular network on a finite amount of money, so we must look towards other available technologies, such as the constantly-expanding cellular network. In this paper we look at the iPhone 4™ capability of reporting accurate and reliable locations and compare it to the recent study of the iPhone 3G™ [24]. Drivers equipped with an iPhone 4™ cellular phone and a vehicle tracking device manually timed how long it took to travel along a 0.99 mile/1.59 kilometer section of roadway. The vehicle tracking device and the iPhone 4™ report speed and location every 10 seconds whereas the iPhone 3G™ reported every 8 seconds [24]. From this data, we calculated the amount of time to traverse the test section of roadway using the proportional model algorithm [18] and compared it to the actual amount of time it took to traverse the test section of roadway as manually timed. We found that the vehicle tracking device had an average error factor of 4.94% from the actual time to traverse the section of roadway (as determined by the stopwatch), whereas the iPhone 4™ was found to have an error factor of 1.10%. The outcome of the case study is used to determine that the iPhone 4™ has higher accuracy than a vehicle tracking device, though it is important to note that the iPhone™ is more limited than a device attached to a vehicle since it can only report its location. If paired with another third party OBD device, however, it can also send the same information as a vehicle tracking device.","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"29 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128698244","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 : 2011-06-05DOI: 10.1109/IVS.2011.5940404
Ikuro Sato, C. Yamano, H. Yanagawa
We propose a computer vision algorithm that detects obstacles crossing a vehicle's path with a monocular camera mounted on the vehicle. False positives are strongly suppressed even for low-resolution images by imposing constraints on feature-based optical flows. The constraints are derived from a model of crossing obstacle motion under perspective projection. A key concept in this model is “Relative Incoming Angle”, which is an angle between the camera's translational direction and relative velocity of a crossing obstacle with respect to the camera. We show a ROC curve that has been obtained by varying the Relative Incoming Angle using our dataset consisting of 18 scenes, 1456 frames. A representative point on the curve yields the detection rate of 59.7% and false positive rate of 2.6% (per-image).
{"title":"Crossing obstacle detection with a vehicle-mounted camera","authors":"Ikuro Sato, C. Yamano, H. Yanagawa","doi":"10.1109/IVS.2011.5940404","DOIUrl":"https://doi.org/10.1109/IVS.2011.5940404","url":null,"abstract":"We propose a computer vision algorithm that detects obstacles crossing a vehicle's path with a monocular camera mounted on the vehicle. False positives are strongly suppressed even for low-resolution images by imposing constraints on feature-based optical flows. The constraints are derived from a model of crossing obstacle motion under perspective projection. A key concept in this model is “Relative Incoming Angle”, which is an angle between the camera's translational direction and relative velocity of a crossing obstacle with respect to the camera. We show a ROC curve that has been obtained by varying the Relative Incoming Angle using our dataset consisting of 18 scenes, 1456 frames. A representative point on the curve yields the detection rate of 59.7% and false positive rate of 2.6% (per-image).","PeriodicalId":117811,"journal":{"name":"2011 IEEE Intelligent Vehicles Symposium (IV)","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-06-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128829192","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：