Development of an Automatic Optical Inspection System for Defect Detection of Dental Floss Stick
碩士 === 國立高雄應用科技大學 === 應用工程科學研究所 === 99 === Dental Floss Sticks are classified into personal hygiene items, therefore, the cleanliness and quality of the products will be stringent requirements. However, due to cost-down issue, the defect detection of dental floss sticks relies on manual inspection w...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2011
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/46417129747898327558 |
| Summary: | 碩士 === 國立高雄應用科技大學 === 應用工程科學研究所 === 99 === Dental Floss Sticks are classified into personal hygiene items, therefore, the cleanliness and quality of the products will be stringent requirements. However, due to cost-down issue, the defect detection of dental floss sticks relies on manual inspection which will lead to inconsistency of inspection results. Since recognition standard and concentration level of inspectors are not uniformly maintained in process. Therefore the standard of defect in detection process will be not consistency. In general, defect management by human recognition is incomplete which will limit product yield and quality.
The objectives of this study are to replace manual inspection by automatic optical inspection system. Study on the development of standardization and automation for three major flaws of dental floss sticks. This study contains layout about light sources and automatic control. Moreover, a computer code for defect detection is designed and an automatic optical inspection system is built for verification.
This study integrates five webcams to process image in one program by personal computer and controls external equipment by I/O card in same time. The effect of defect detection on performance of production line can be measured with larger than 0.09 mm2 of damage defects and dirty defects with gray-scale difference greater than 35. Shortage defects with length greater than 3mm can also be detected. The test results on ten pieces show that detection rates are 90.4% for damage defects and 54.5% for dirty defects and 84.8% for shortage defects.
|
|---|