2D irregular shape parts recognition for automatic assembly | |
| Author | Sakol Kitkasiwat |
| Subject(s) | Assembly -line method Shapes |
| Note | The degree of Master of Engineering in Mechatronics, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Abstract | This thesis is to develop matching algorithm for assembling 2D irregular shape parts. The algorithm was implemented based on a real process. The algorithm focuses on parts identification and matching. Given parts are always placed to a working area of assembling process randomly. This system has to assemble any parts following a design. In order to achieve this process, there are three main procedures as follow: image enhancement, part recognition and toolpath generation. Firstly, image enhancement, a given image and a designing image is transformed to one-pixel wide closed contour binary image to analyze the contour information after the both of them are acquired by capturing from a working area of an assembling process. This algorithm can extract contour information from raw images. Secondly, part recognition, the purposed algorithm of this process is to create index descriptions and match parts from index information. The parts images initially, are traced to create group of contours with x,y coordinates. Then the algorithm generates four index descriptions in which can specify the parts differences under a polygon-like concept. The index descriptions can show the estimated detail of each object shape. By using the index description, the algorithm then matches the given parts to a design and makes matched pairs. Lastly, toolpath generation, this technique has been developed to search a efficient toolpath of assembly. There are two sections for calculation in this process. the first section is to manipulate the part orientation in order to make a machine assemble part in a correct angle. The second section is to select a suitable toolpath of assembly. This study attempts to investigate the performance of the algorithm by testing with several types of part. The result of all experiments shows the efficiency of the algorithm for assembling polygon and non-polygon. |
| Year | 2014 |
| School | School of Engineering and Technology (SET) |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Microelectronics (ME) |
| Chairperson(s) | Pisut Koomsap; |
| Examination Committee(s) | Bohez, Erik L. J.;Manukid Parnichkun; |
| Scholarship Donor(s) | AIT Fellowship; |