Formulario robotica fronte

Dh transfrm matrix redundancy

n-m, n = Dom, n’=TCP DOF, m = Task DOF Ti = Te+Tc

Orientation planning

Convex combination total interval Bang-bang VEDI SOTTO ***ADDITIONAL CONSTRAINTS Gradient VISION

Preprocessing gradient

• Sobel operator
• Local enhancement histogram

Perception

1) Perception = cameras, camera calibration, stereoscopic vision GxGy

Preprocessing

2) Preprocessing = avg of the neighborhood (blur contours), median filter, avg of multiple images, stencils, enhancement, histogram, enhancement based on local properties, edge detection, gradient, laplacian, canny, thresholding

Segmentation

3) Segmentation = Local analysis, Global analysis with Hough diagram, segmentation through thresholding, pixel aggregation region (odom) growing

Process model

Mobile 4) Description = Chained code, normalization, characteristic shape (monodim. Funct. Descript / Fourier descr), texture (hist. moments)

Recognition

5) Recognition = pattern vector, decision function, Rule-driven, Data-driven, Hybrid, Training, Test & Validation

Meas. model ** Bicycle Kalman (Lin. Gauss.system) Unicycle (weighted samples for Particle Filters approx robot pose belief)

Inverse kinematics

DEF ODOMETRY particle set given motion model - prediction (new (assign at each particle a weight accord - update to sensor) particle wrt their weight) - resampling (resample Mapping Usa campioni random pesati per env assuming robot pose: Building a consistent repr of the approssimare la ddp (belief) del robot pose. Non-holonomic constr

Sensors types

• Stoch maps with probabilistic --Landmark-based repr.
• Dynamic range 1. Encoders, resolver - Occupancy grid - maps description of scenario - Resolution TR PLANNING
• 2. Inertial sensor: - Bandwidth - Heading sensors HD models of environment -- accuracy SLAM - inclinometers - precision
• MOBILE ** - compasses - gyrocompasses Incremental (more easy we calculate only v and a) - Accel-Gyrosc (mechanical e optical (Sagnac)) posit of landmarks in the state vector) (Include
• 3. Beacons
• 4. Distance/prox sens. Complex, can diverge due to non-lin
• 5. Accelerom/Inert. Measurm Unit
• 6. Vision Absolute

scriptumPost