Machine Vision for IC Inspection

Project Advisor :

Assoc. Prof. Dr. Pakorn Kaewtrakulpong

Candidate:

Mr. Pongsak Joompolpong

Mr. Wisuwat Bhoka

Abstract:

This project proposes a prototype of an automated visual inspectionsystem for IC quality inspection using Machine Vision techniques. This project composes of hardware and software parts. First, the ICs in a tube are fed into the inspection station in a one-byone manner. Then the software uses an image processing to inspect the mark on each IC and the completeness of its connector pins.The IC mark is evaluated by a shape matching method using a pre-recorded model obtained from a master specimen.The connector inspection is processed by Morphological techniques. When the inspection is completed, a separatoris used to direct the IC to its expected location, i.e., to an output tube for the accept product otherwise to a reject bin. From our experiment, the system can achieve 72.14 % accuracy and can inspect 12 products per minute.

คลิกที่รูปเพื่อดูรูปกิจกรรม

Back / ย้อนกลับ | update : 02/01/2556