Defect detection of transparent tube assembly | |
| Author | Nutthapong Wongpromkal |
| Call Number | AIT Thesis no.ISE-24-28 |
| Subject(s) | Quality control Image processing Computer vision |
| Note | A thesis submitted in partial fulfillment of the requirements for the degree of Master of Engineering in Mechatronics and Machine Intelligence |
| Publisher | Asian Institute of Technology |
| Abstract | In traditional production quality control, human evaluators are tasked with discerning and eliminating flawed components to guarantee that solely superior-quality merchandise reaches patrons. However, upholding uniformity in evaluative precision and periodicity proves arduous due to intrinsic human constraints, such as exhaustion and variability in discernment. To rectify these dilemmas, computer vision technology has been progressively utilized to supplant human evaluators in quality assurance procedures. Computer vision provides a substantial advantage in its capacity to execute meticulous, consistent evaluations at elevated velocity, thereby reducing human error and variability. This transition augments the dependability and efficacy of quality assurance, thereby ensuring enhanced product standards.This investigation elucidates a methodology for product quality assurance utilizing computer vision to identify inadequately assembled components. The specimen examined was the Z-305 from Mitani Siam Tech, a transparent tube with a translucent body. In light of the challenges presented by the transparency, two detection methodologies were employed.The initial methodology is Convolutional Neural Network Classification. Neural network image classification functions by processing an image through a succession of interconnected layers of neurons.The subsequent methodology is Rule-based Image Processing. Rule-based classification in image processing categorizes images by implementing predefined conditions base on specific extracted features, such as color, shape, and texture.This study aspires to enhance the precision and dependability of defect identification in transparent assemblies by contrasting two computer vision methodologies under the constraints of application and configuration of production apparatus. This study contributes to the field by addressing the unique challenges posed by transparent objects and demonstrating the effectiveness of computer vision techniques in improving product quality assurance. |
| Year | 2024 |
| Type | Thesis |
| School | School of Engineering and Technology |
| Department | Department of Industrial Systems Engineering (DISE) |
| Academic Program/FoS | Mechatronics and Machine Intelligence (MMI) |
| Chairperson(s) | Manukid Parnichkun |
| Examination Committee(s) | Huynh, Trung Luong;Mongkol Ekpanyapong |
| Scholarship Donor(s) | Mitani Scholarship |
| Degree | Thesis (M. Eng.) - Asian Institute of Technology, 2024 |