A5063547251- Robotics and Sensor-Based Localization
- Soft Robotics and Applications
- 3D Surveying and Cultural Heritage
- Robot Manipulation and Learning
- Robotics and Automated Systems
- Control and Dynamics of Mobile Robots
- Advanced Vision and Imaging
- Remote Sensing and LiDAR Applications
- Robotic Locomotion and Control
- Advanced Optical Sensing Technologies
- Modular Robots and Swarm Intelligence
- Industrial Vision Systems and Defect Detection
- Robotic Path Planning Algorithms
- Image and Object Detection Techniques
- Advanced Manufacturing and Logistics Optimization
Universidade Tecnológica Federal do Paraná
2017-2020
DLG Automation (Brazil)
2017
Advanced Systems Laboratory
2017
Climbing robots are characterized by a secure surface coupling that is designed to prevent falling. The robot ability assured an adhesion method leading nonlinear dynamic models with time-varying parameters affect the robot’s mobility. Additionally, wheel friction and force of gravity also relevant issues can compromise climbing if they not well modeled. This work presents model-based torque controller for velocity tracking in four-wheeled specially inspect storage tanks. (MPC) compensates...
This work proposes a novel semantic perception system based on computer vision and machine learning techniques. The main goal is to identify objects in the environment extract their characteristics, allowing dynamic interaction with environment. composed of GPU processing source 3D sensor that provides RGB image PointCloud data. structured three steps: Lexical Analysis, Syntax Analysis finally an Anticipation. detects actual position (or tokens) environment, through combination PointCloud,...
This work presents a new concept of environment mapping aiming to identify dynamic objects in the scene, and then estimate their direction future position. The motion analysis is performed directly traditional 2D costmap through three different map frames. Costmap technique that uses an occupation grid used, where each cell has degree uncertainty about whether it occupied or not. pose mobile object estimated its motion' behavior, analyzing previous poses. A representation called RGB-costmap...
The use of GPU in point cloud processing usually represents a gain computation time more than ten times higher then on CPU, especially for large amounts data. This paper brings an evaluation the fusion three different systems (PC, Nvidia TX1 and TK1) using CPU GPU. objective is to find best way perform sensor autonomous inspection robot. Four scenarios were considered tests, with five neighbor radius. A table found each experiment was created allow quick comparison. As results, reduction...
This paper presents an omnidirectional RGB-D (RGB + Distance fusion) sensor prototype using actuated LIDAR (Light Detection and Ranging) RGB camera. Besides the sensor, a novel mapping strategy is developed considering scanning characteristics. The can gather 3D data from any direction by toppling in 90 degrees laser scan rotating it about its central axis. based on two environment maps, local map for instantaneous perception, global perception memory. 2D represents surface front of robot...