Automated image processing for scratch detection on specular surfaces

---

Tezin Türü: Yüksek Lisans

Tezin Yürütüldüğü Kurum: Orta Doğu Teknik Üniversitesi, Mühendislik Fakültesi, Elektrik ve Elektronik Mühendisliği Bölümü, Türkiye

Tezin Onay Tarihi: 2018

Öğrenci: VOLKAN OKBAY

Danışman: GÖZDE AKAR

Özet:

In industry, problems due to human error, mechanical flaws and transportation may occur; besides, they need to be detected in fast and efficient ways. In order to eliminate failure of human inspection, automated systems come in action, usually image processing involved. This thesis work, targets one common mass production problem on specular surfaces, i.e. scratch detection. To achieve this, we have implemented two different prototypes. The low-cost system is based on basic line detection, and the mid-end system depends on learning based detection. Both systems are implemented on embedded platforms and performance comparisons are done. Detailed analysis is carried out on computational cost and detection performance. This real-world episode is done on a mechanical prototype in laboratory environment