Vodenje mobilnega robota KUKA youBot na osnovi robotskega vida

ŽUPANEK, IZTOK (avtor), Mihelj, Matjaž (mentor)

V diplomskem delu se ukvarjamo s problemom vodenja robotskega manipulatorja KUKA youBot na osnovi robotskega vida. V nalogi sta predstavljena dva načina vodenja robota. Prvi je na osnovi barvnih markerjev, ki so nameščeni na manipulator in na objekte, s katerimi robot opravlja nalogo. V drugem delu vodenje poteka na osnovi podatkov iz programa za detekcijo markerjev ArUco. Markerji so nameščeni na mestih, kjer mora manipulator opraviti del naloge; ko z delom zaključi, se vrne v izhodiščno stanje. Na kratko so predstavljene metode osvetljevanja, na kaj je potrebno biti pozoren pri izbiri svetila in njegovi svetilni moči. Predstavljenih je tudi nekaj bolj ali manj ključnih napak, ki se pojavijo pri uporabi leč na kamerah, njihov vpliv na sliko in postopki, kako jih odpraviti. Nekoliko podrobneje so predstavljeni algoritem za detektiranje ArUco markerjev in metode kalibracije, s katerimi se odpravijo napake leče. Vodenje je bilo v celoti realizirano v programskem okolju MATLAB SIMULINK, kjer je bila v diagramu prehajanja stanj določena tudi naloga, ki jo je moral robot opraviti. Detekcija barvnih markerjev je bila v celoti izdelana v programskem okolju MATLAB, kjer sta na podlagi vgrajenih orodij Image Acquisition Toolbox in Image Processing Toolbox potekala zajem in obdelava slike.

• Jezik: Slovenski jezik
• Ključne besede: KUKA youBot, inverzna kinematika, strojni vid, kalibracija, napake leč, obdelava slike, ArUco markerji.
• Vrsta gradiva: Diplomsko delo
• Organizacija: FE - Fakulteta za elektrotehniko
• Leto izida: 2016
• PID: 20.500.12556/RUL-85553
• Datum objave v RUL: 16.09.2016

Sekundarni jezik

• Jezik: Angleški jezik
• Naslov: Vision-based control of mobile robot KUKA youBot

Izvleček

The objective of this study is the control of a robot manipulator KUKA youBot based on robot vision. The thesis deals with two modes of robot control. The first is based on colour markers that are installed on the manipulator and on objects, which are important for performing the task. In the second part, control is based on the information from the algorithm for detection of ArUco markers. The markers are placed at positions, where the manipulator is to carry out a part of its task. As soon as the task has been completed the robot returns to the starting position. Methods of illumination are presented briefly. Specified is what has to be taken into account when selecting the light and its luminosity. Furthermore, discussed are issues related to using lenses on cams, their influence on the picture and procedures for the elimination of errors as well as the camera calibration. Finally, the algorithm for detection of ArUco markers is described and task implementation is detailed. The robot control was entirely implemented in the MATLAB SIMULINK programme environment, where the robot task was determined in the StateFlow diagram. The detection of colour markers was created in MATLAB programme environment where capture and image processing were implemented based on Image Acquisition Toolbox and Image Processing Toolbox.

• Ključne besede: KUKA youBot, inverse kinematics, machine vision, calibration, lens errors, image processing, ArUco markers