SENSOR FUSION - Laser and Camera Camera Laser Range Finder direct depth measurement illumination dependent wide accuracy span (till 200 m) accurate only for limited distances only 2 or 3 D contour info on colour and texture high computational time M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - LASER + CAMERA MEASUREMENT BY LASER and CAMERA • Laser rangefinders, principles and applications • Laser-Camera Calibration MEASUREMENT BY LASER and CAMERA: object recognition • Clustering and segmentation of the scene seen by the laser • Chamfer distance (or Hausdorff) MEASUREMENT BY LASER and CAMERA: object recognition • reprojection of the object model of CCD • Corner extraction • Matching and acceptance M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - LASER + CAMERA MEASUREMENT BY LASER and CAMERA: object recognition • Practice with real data. The scene will be a box of given size to be recognized MEASUREMENT BY LASER and CAMERA: object recognition • Practice with real data. SUPERQUADRICHE • General concepts SUPERQUADRICHE • Application to object recognition M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - esercitazione SENSOR FUSION of timeline signals - Complementary Filtering. Theory and applications. Example of simulation of an altimeter baro-inertial. SENSOR FUSION of timeline signals - Simulation PC in the classroom portion of the estimate by filtering between a barometer and an inertial platform SENSOR FUSION of timeline signals - Use of real data: - Measurement of the camera position by means of an object in motion on a plane by means of KLT, after having calibrated the worktop (using a grid placed on the floor) - Combined with the accelerometer data and complementary filtering Telecamera + oggetto sul piano con accelerometro solidale M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - sensor fusion + esercitazione M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - sensor fusion + tesina SENSOR FUSION - Statistical concepts accessories, Bayes' Theorem SENSOR FUSION - Application of Bayes' theorem to the fusion of information scalar and vector SENSOR FUSION - Kalman Filter SENSOR FUSION. Tutorial SLAM + Kalman. Mapping with laser scanner or camera SENSOR FUSION. Tutorial SLAM + Kalman. Mapping with laser scanner or camera M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - sensor fusion + tesina MOBILE ROBOT - Overview of applications. Localization issues, planning and control, holonomic and non-linear differential constraints. Conditions of integrability, Model Differential Drive. Recursive equations for odometry. MOBILE ROBOT - Models kinematic unicycle, bicycle and bicycle trailers with N MOBILE ROBOT - Problem of planning. Classification. Transformation of kinematic models in chained form. MOBILE ROBOT - Planning open-loop. Systems in chained form for the solution of the motion point-to-point with sinusoidal input, wise constant, polynomial. Calculation of Cartesian trajectories eligible MOBILE ROBOT - Planning open-loop. Clothoids and polar spline. Examples of calculation. MOBILE ROBOT - Controllability of systems that are not holonomic. Example of control system in chained form linearized around the desired trajectory M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Programma - robot mobili Exam: homework + 1 ORAL ARGUMENTS ON 2 CHOICES (between 4 topics, which does not coincide with that of the homework), [NOTE: 1 topic for mehanics area] Homework chose examples: trajectory control of manipulators by inverting the differential kinematics (CLASS) simulation and trajectory control for non-holonomic vehicles processing data for the calibration kinematics of an autonomous vehicle AGV SLAM using a laser scanner at 360 ° M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Modalità di esame L.Sciavicco, B. Siciliano, Robotica - Modellistica, pianificazione e controllo 3/ed, McGraw Mitchell Harvey, "Multi-Sensor Data Fusion: An Introduction" - Springer 2007 Ake Bjork, Numerical methods for least squares problems M. De Cecco, Lucidi del corso di Robotica e Sensor Fusion Luca Baglivo, M. De Cecco, Navigazione di Veicoli Autonomi Fondamenti di “sensor fusion” per la localizzazione L. Baglivo, Navigazione di Veicoli Autonomi (Localizzazione, Pianificazione e Controllo traiettoria) M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion Testi Consigliati VIDEO M. De Cecco - Lucidi del corso di Robotica e Sensor Fusion