Semantički model stabljike biljke za taktilnu manipulaciju

undergraduate thesis

• Title (english): A semantic model of a plant stem for tactile manipulation
• Author: Karlo Rajković
• Mentor: Matko Orsag (mentor)
• Committee member: Matko Orsag (predsjednik povjerenstva)
• Committee member: Marsela Polić (član povjerenstva)
• Committee member: Bruno Marić (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2020-07-08, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Electrical Engineering

Abstract

Rad započinje simuliranjem Franka Emika Panda robota u Gazebo virtualnom okruženju. Pomoću MoveIt paketa prilagođenog Franki, robotom se može jednostavno upravljati u simulaciji. Dodavanjem taktilnog senzora na alatni dio robota (eng. end effector), dobiven je način na koji se mogu pamtiti ostvareni kontakti robota sa ostalim predmetima. Sa Frankom opremljenom novodefiniranim senzorom sakupljaju se skupovi točaka (pointcloud-ovi) različitih jednostavnih oblika koji predstavljaju dijelove stabljike biljke. Dobiveni pointcloud-ovi se zatim iskoriste u obliku slika kao ulazni podatci za model strojnog učenja napravljenog korištenjem Keras API. Model strojnog učenja je nakon učenja na dovoljno velikom skupu podataka naučio prepoznavati između ravnog dijela stabljike, ploda i račvanja.

Abstract (english)

The work disseminated in this thesis starts with simulating Franka Emika Panda robot in Gazebo virtual environment. With the help of a newly created MoveIt ROS package, specially designed for Franka, the robot can be easily controlled in the simulation. Adding a contact sensor to the end effector part of the robot enables one to collect contact data when Franka interacts with other objects. Equipped with the newly defined contact sensor, Franka can collect pointclouds of simple geometrical objects, with each object representing a different part of a plant stem. Collected pointclouds are then used in a form of images as an input for a machine learning model, which is defined using Keras API in Python. After training (learning), the model was able to succesfully differentiate between different parts of a stem - a flat part, fruit and branching.

• Keywords: Franka; ROS; Gazebo; MoveIt; pointcloud; strojno učenje; Keras
• Keywords (english): Franka; ROS; Gazebo; MoveIt; Pointcloud; Machine learning; Keras
• Language: croatian
• URN:NBN: urn:nbn:hr:168:778165
• Study programme: Title: Electrical Engineering and Information Technology; Study programme type: university; Study level: undergraduate; Academic / professional title: sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije (sveučilišni/a prvostupnik/ prvostupnica (baccalaureus/baccalaurea) inženjer/inženjerka elektrotehnike i informacijske tehnologije)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-14 09:49:27