Real-time image data acquisition and inspection system for integrated circuit wafer after sawing process

Abstract

The major defect in integrated circuit wafers after the sawing process is the unexpected cracking from the sawing lane that damages the die area. Traditional defect detection algorithms for integrated circuit wafers after the sawing process can be divided into two parts. The first part is locating the sawing lane area and the second is detecting defects in the sawing lane area. The traditional methods take too long to locate the sawing lane area. To overcome this, this paper proposes a novel method for locating the sawing lane area using low-resolution and gray-value input images. After the sawing lane area is extracted from the original image, defects can be detected by traditional defect detection methods in the sawing lane area or by the method proposed in this paper. Using the proposed defect detection method is faster and better for real-time operation than traditional methods since the proposed method is more concise. Finally, a pioneering image data acquisition and inspection system for integrated circuit wafers after the sawing process is proposed in this paper. Furthermore, the proposed system works in real time because its architecture operates quickly and its proposed algorithm is simple, using low-resolution, gray-value images. Therefore, the proposed system can operate in real time and maintain a high-level detection rate.