Automatic Inspection System for Egg Packs using Image Processing Techniques

• Main Authors: Gwo-En Hwang, 黃國恩
• Other Authors: Yuan-Hsiang Chang
• Format: Others
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/51478688172842783299

碩士 === 中原大學 === 通訊工程碩士學位學程 === 103 === During the automatic packing of eggs, quality of eggs may vary during the conveying processes, resulting in good or crack eggs. In this study, a vision-based system for the automatic inspection of egg packs using image processing techniques is presented. The objective is to automatically classify if an egg in a pack is good or crack. The method includes: egg region localization, egg region correction, egg region segmentation, and egg quality classification. A set of images containing six-egg packs was formed for the performance evaluation. The preliminary results showed that our system could achieve a reasonable classification rate of 86% for good eggs and 100% for crack eggs, respectively. In conclusion, although this study is still preliminary, our system has the potentials to be incorporated in an egg packing conveyor to ease the labor-intensive and time-consuming tasks for quality controls.