Vitor Santos - Academia.edu (original) (raw)

Papers by Vitor Santos

The concept of tracking the component origin and manufacturing specification encoded in the form ... more The concept of tracking the component origin and manufacturing specification encoded in the form of Data-Matrices has been gaining ground. This has translated into a search for cost-effective alternatives to machine readable, printable codes such as barcodes. The advantage of Data-Matrices assumes special relevance in the case of miniature components and in the case of complex, multi-part assemblies in which a large amount of information must be encoded within a small foot-print. We propose the application of state of the art techniques in multi-scale, local image feature extraction (Corners and lines) to read the binary data in Data-Matrices. The columns of binary data are arranged in the form of a sequence of binary features each of which codes some information. Our aim is to increase the applicability of the codes by developing applications in object classification without supervision, achieving tolerance to incomplete information about the objects to be classified. The classification process is achieved by using techniques that are commonly utilized in bio-informatics to align protein sequence which recently have been applied to Place recognition problems in Mobile Robot Navigation.

This paper presents and demonstrates the validity of a solution to implement a 3D laser scanner b... more This paper presents and demonstrates the validity of a solution to implement a 3D laser scanner based on a traditional 2D laser range finder. Besides the mechanical and control issues addressed to achieve physically the 3D sensor, the adaptation of some previously developed techniques to merge range and intensity image illustrates the potential of such a unit. The result is a promising system both for navigation and 3D model reconstruction at a commercial price 10 or 20 times lower than current commercial 3D laser scanners in spite of some limitations related to maximal range and accuracy.

This paper describes a system intended to identify and track dynamic targets that may change appe... more This paper describes a system intended to identify and track dynamic targets that may change appearance while moving. The full system includes a pan and tilt unit to ease tracking and keep the interesting target in the center of the image. View-based Haar-like features are used for object recognition while template matching continues to track the object even when its view is not recognized by object recognition system. Some of the techniques used to improve the template matching performance are also presented. Preliminary results are given and system performs well up to 15 frames per second on a 320 x 240 image on an ordinary laptop computer.

Autonomous navigation in real roads has been a concern for several years now, especially for the ... more Autonomous navigation in real roads has been a concern for several years now, especially for the Intelligent Transportation Systems community. Several interesting results have been obtained since the early 90's, but the problem is so vast and manifold that even today it keeps many persons at academic and industrial levels engaged. These researchers are focused in the search for more, and more general, approaches for road and lane detection, traffic lights and many other perturbations on the normal flow of perception from the road and its entourage. Perception and navigation are the main two components of this huge problem, where the second depends largely on the performance of the first. In Portugal, a competition for Autonomous Driving was created in 2001. It intended to promote the development of solutions for these perception and navigation problems. Currently, the competition accounts for several challenges and a University project named ATLAS has developed a series of robots that have outperformed all the competitors for five years in a row. The paper presents the main techniques and algorithms that lead to this success and formulates the bases to migrate the solutions to real road conditions. Perception of road and its features are given the main focus.

The concept of tracking the component origin and manufacturing specification encoded in the form ... more The concept of tracking the component origin and manufacturing specification encoded in the form of Data-Matrices has been gaining ground. This has translated into a search for cost-effective alternatives to machine readable, printable codes such as barcodes. The advan-tage of Data-Matrices assumes special relevance in the case of miniature components and in the case of complex, multi-part assemblies in which a large amount of information must be encoded within a small foot-print. We propose the application of state of the art techniques in multi-scale, local image feature extraction (Corners and lines) to read the binary data in Data-Matrices. The columns of binary data are arranged in the form of a sequence of binary features each of which codes some information. Our aim is to increase the applicability of the codes by developing applications in object classification without supervision, achieving tolerance to incomplete information about the objects to be classified. The classifi...

Advances In Mobile Robotics - Proceedings of the Eleventh International Conference on Climbing and Walking Robots and the Support Technologies for Mobile Machines, 2008

This article presents a method for the creation of a topological map without having to previously... more This article presents a method for the creation of a topological map without having to previously create a geometric map. Independently obtained topological paths are compared pairwise to search for possible overlap in the view sequence. Each individual topological path is created from a sequence of raw data views that are sampled by leading the robot along a path in the environment. The multiple topological paths are 'stitched together' into a generic Topological map by identifying overlapping segments in the individual sequences. A general topological map can be created by considering all the multiple sequences or separate runs through the environment. Results on the merger of upto 8 seperate paths indicate that this method of mapping large environments, by selective touring through the environment, can generate robust topological maps.

Springer Tracts in Advanced Robotics, 2015

Recognition of traffic lights and signs is an important task for advanced driver assistance syste... more Recognition of traffic lights and signs is an important task for advanced driver assistance systems (ADAS), as well as for autonomous driving systems of driverless cars. This paper deals with the challenge of perceiving traffic signs for autonomous driving. It is concerned with the competition of fully autonomous robots that takes place in a track with the shape of a traffic road. A specific set of signs has been addressed and solved with very good results. A combination of two techniques based on blob analysis and pattern recognition has been used and selected results of the experiments will be presented along with the description of the main algorithms.

2007 7th IEEE-RAS International Conference on Humanoid Robots, 2007

The recent trend of humanoid robotics research has been deeply influenced by concepts such as dis... more The recent trend of humanoid robotics research has been deeply influenced by concepts such as distributed architectures, local control, force interaction and emergence of coordinated motions. A hypothesis is that feedback control from several sensors, such as force sensors and inertial devices, and more advanced control algorithms will be a key issue for the next developments. In this paper, we discuss how these concepts have been applied to a custom low-cost humanoid platform developed for research purposes. The humanoid robot is equipped with a rich set of sensors enabling the evaluation of simple feedback rules used online to control the robot. A great focus has been given to a force-driven controller based on the Jacobian transpose. A kind of intermediate local-level controller is implemented based on force sensing, providing robust and adaptive behaviour. The proposed ideas and concepts are introduced and validated in several hardware experiments.

Eighth International IFAC Symposium on Robot Control, 2006, 2006

... Conference Editor: Bruno Siciliano. Authors Jose RT Puga; Filipe MT Silva; Vitor MF Santos. I... more ... Conference Editor: Bruno Siciliano. Authors Jose RT Puga; Filipe MT Silva; Vitor MF Santos. Identifier 10.3182/20060906-3-IT-2910.00129. Index Terms 5-10 legged robots,biped walking,trajectory planning,COG control,force interaction control. ...

The concept of tracking the component origin and manufacturing specification encoded in the form ... more The concept of tracking the component origin and manufacturing specification encoded in the form of Data-Matrices has been gaining ground. This has translated into a search for cost-effective alternatives to machine readable, printable codes such as barcodes. The advantage of Data-Matrices assumes special relevance in the case of miniature components and in the case of complex, multi-part assemblies in which a large amount of information must be encoded within a small foot-print. We propose the application of state of the art techniques in multi-scale, local image feature extraction (Corners and lines) to read the binary data in Data-Matrices. The columns of binary data are arranged in the form of a sequence of binary features each of which codes some information. Our aim is to increase the applicability of the codes by developing applications in object classification without supervision, achieving tolerance to incomplete information about the objects to be classified. The classification process is achieved by using techniques that are commonly utilized in bio-informatics to align protein sequence which recently have been applied to Place recognition problems in Mobile Robot Navigation.

This paper presents and demonstrates the validity of a solution to implement a 3D laser scanner b... more This paper presents and demonstrates the validity of a solution to implement a 3D laser scanner based on a traditional 2D laser range finder. Besides the mechanical and control issues addressed to achieve physically the 3D sensor, the adaptation of some previously developed techniques to merge range and intensity image illustrates the potential of such a unit. The result is a promising system both for navigation and 3D model reconstruction at a commercial price 10 or 20 times lower than current commercial 3D laser scanners in spite of some limitations related to maximal range and accuracy.

This paper describes a system intended to identify and track dynamic targets that may change appe... more This paper describes a system intended to identify and track dynamic targets that may change appearance while moving. The full system includes a pan and tilt unit to ease tracking and keep the interesting target in the center of the image. View-based Haar-like features are used for object recognition while template matching continues to track the object even when its view is not recognized by object recognition system. Some of the techniques used to improve the template matching performance are also presented. Preliminary results are given and system performs well up to 15 frames per second on a 320 x 240 image on an ordinary laptop computer.

Autonomous navigation in real roads has been a concern for several years now, especially for the ... more Autonomous navigation in real roads has been a concern for several years now, especially for the Intelligent Transportation Systems community. Several interesting results have been obtained since the early 90's, but the problem is so vast and manifold that even today it keeps many persons at academic and industrial levels engaged. These researchers are focused in the search for more, and more general, approaches for road and lane detection, traffic lights and many other perturbations on the normal flow of perception from the road and its entourage. Perception and navigation are the main two components of this huge problem, where the second depends largely on the performance of the first. In Portugal, a competition for Autonomous Driving was created in 2001. It intended to promote the development of solutions for these perception and navigation problems. Currently, the competition accounts for several challenges and a University project named ATLAS has developed a series of robots that have outperformed all the competitors for five years in a row. The paper presents the main techniques and algorithms that lead to this success and formulates the bases to migrate the solutions to real road conditions. Perception of road and its features are given the main focus.

The concept of tracking the component origin and manufacturing specification encoded in the form ... more The concept of tracking the component origin and manufacturing specification encoded in the form of Data-Matrices has been gaining ground. This has translated into a search for cost-effective alternatives to machine readable, printable codes such as barcodes. The advan-tage of Data-Matrices assumes special relevance in the case of miniature components and in the case of complex, multi-part assemblies in which a large amount of information must be encoded within a small foot-print. We propose the application of state of the art techniques in multi-scale, local image feature extraction (Corners and lines) to read the binary data in Data-Matrices. The columns of binary data are arranged in the form of a sequence of binary features each of which codes some information. Our aim is to increase the applicability of the codes by developing applications in object classification without supervision, achieving tolerance to incomplete information about the objects to be classified. The classifi...

Advances In Mobile Robotics - Proceedings of the Eleventh International Conference on Climbing and Walking Robots and the Support Technologies for Mobile Machines, 2008

This article presents a method for the creation of a topological map without having to previously... more This article presents a method for the creation of a topological map without having to previously create a geometric map. Independently obtained topological paths are compared pairwise to search for possible overlap in the view sequence. Each individual topological path is created from a sequence of raw data views that are sampled by leading the robot along a path in the environment. The multiple topological paths are 'stitched together' into a generic Topological map by identifying overlapping segments in the individual sequences. A general topological map can be created by considering all the multiple sequences or separate runs through the environment. Results on the merger of upto 8 seperate paths indicate that this method of mapping large environments, by selective touring through the environment, can generate robust topological maps.

Springer Tracts in Advanced Robotics, 2015

Recognition of traffic lights and signs is an important task for advanced driver assistance syste... more Recognition of traffic lights and signs is an important task for advanced driver assistance systems (ADAS), as well as for autonomous driving systems of driverless cars. This paper deals with the challenge of perceiving traffic signs for autonomous driving. It is concerned with the competition of fully autonomous robots that takes place in a track with the shape of a traffic road. A specific set of signs has been addressed and solved with very good results. A combination of two techniques based on blob analysis and pattern recognition has been used and selected results of the experiments will be presented along with the description of the main algorithms.

2007 7th IEEE-RAS International Conference on Humanoid Robots, 2007

The recent trend of humanoid robotics research has been deeply influenced by concepts such as dis... more The recent trend of humanoid robotics research has been deeply influenced by concepts such as distributed architectures, local control, force interaction and emergence of coordinated motions. A hypothesis is that feedback control from several sensors, such as force sensors and inertial devices, and more advanced control algorithms will be a key issue for the next developments. In this paper, we discuss how these concepts have been applied to a custom low-cost humanoid platform developed for research purposes. The humanoid robot is equipped with a rich set of sensors enabling the evaluation of simple feedback rules used online to control the robot. A great focus has been given to a force-driven controller based on the Jacobian transpose. A kind of intermediate local-level controller is implemented based on force sensing, providing robust and adaptive behaviour. The proposed ideas and concepts are introduced and validated in several hardware experiments.

Eighth International IFAC Symposium on Robot Control, 2006, 2006

... Conference Editor: Bruno Siciliano. Authors Jose RT Puga; Filipe MT Silva; Vitor MF Santos. I... more ... Conference Editor: Bruno Siciliano. Authors Jose RT Puga; Filipe MT Silva; Vitor MF Santos. Identifier 10.3182/20060906-3-IT-2910.00129. Index Terms 5-10 legged robots,biped walking,trajectory planning,COG control,force interaction control. ...