Benchmarking, Research, Development, and Support for ORNL Automated Image and Signature Retrieval (AIR/ASR) Technologies

Tobin, K.W.

Oak Ridge National Laboratory

关键词: Management;    Ornl;    Architecture;    Manufacturing;    Defects;   

DOI  :  10.2172/940300
RP-ID  :  ORNL02-0655
RP-ID  :  DE-AC05-00OR22725
RP-ID  :  940300

美国|英语

来源: UNT Digital Library

【 摘 要 】

This report describes the results of a Cooperative Research and Development Agreement (CRADA) with Applied Materials, Inc. (AMAT) of Santa Clara, California. This project encompassed the continued development and integration of the ORNL Automated Image Retrieval (AIR) technology, and an extension of the technology denoted Automated Signature Retrieval (ASR), and other related technologies with the Defect Source Identification (DSI) software system that was under development by AMAT at the time this work was performed. In the semiconductor manufacturing environment, defect imagery is used to diagnose problems in the manufacturing line, train yield management engineers, and examine historical data for trends. Image management in semiconductor data systems is a growing cause of concern in the industry as fabricators are now collecting up to 20,000 images each week. In response to this concern, researchers at the Oak Ridge National Laboratory (ORNL) developed a semiconductor-specific content-based image retrieval method and system, also known as AIR. The system uses an image-based query-by-example method to locate and retrieve similar imagery from a database of digital imagery using visual image characteristics. The query method is based on a unique architecture that takes advantage of the statistical, morphological, and structural characteristics of image data, generated by inspection equipment in industrial applications. The system improves the manufacturing process by allowing rapid access to historical records of similar events so that errant process equipment can be isolated and corrective actions can be quickly taken to improve yield. The combined ORNL and AMAT technology is referred to hereafter as DSI-AIR and DSI-ASR.