Image analysis – Applications – Manufacturing or product inspection
Reexamination Certificate
1998-09-30
2001-09-18
Mancuso, Joseph (Department: 2621)
Image analysis
Applications
Manufacturing or product inspection
C348S126000, C382S248000, C382S281000
Reexamination Certificate
active
06292583
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an apparatus for recognizing and detecting defect information such as an alien or foreign substance or substances occurring on a semiconductor wafer, a printed board or the like. Herein, the term “printed board” is intended to refer generically to a single-sided, a double-sided and a multi-layer printed boards inclusive of flexible printed circuits.
2. Description of the Related Art
A defect occurring on or a foreign substance (or material) mixing in a printed board or a semiconductor wafer during manufacturing process thereof causes a defective or faulty product. Therefore, it is necessary to rapidly detect any defect or foreign substance at the time when it has occurred on or mixed in the printed board or the semiconductor wafer.
There have been conventionally proposed image information processing apparatus each of which detects, recognizes or judges a defect or an alien substance as mentioned above on the basis of a photograph of a printed board, or an image or picture of a semiconductor wafer obtained by having taken it using an optical microscope or the like to input it to the apparatus. Almost all of the procedures of these image information processing apparatus are based on digital image processing in which an image is digitally treated or processed, and they perform the image processing using a computer.
Heretofore, as the defect detecting procedure using the image processing, the following have been mainly known:
(a) golden device comparison method;
(b) CAD (computer-aided design) data comparison method;
(c) die to die comparison method; and
(d) FFT method.
In the golden device comparison method (a), an image of an ideal semiconductor wafer having no defect (hereinafter referred to as golden device) is previously provided. An image of an object (wafer or printed board) to be inspected is compared with the image of the golden device to detect a defect of the object. The comparison between two images means a processing in which values of picture elements or pixels of the one image are taken from values of picture elements or pixels of the other image for their corresponding picture elements one by one to create a difference image with respect to picture element values between the two images. In case the image of the object to be inspected has no defect, the two images being compared are all the same, and the resulting difference image will be a flat image having its all picture element values of 0 throughout the entire image. However, in case the image of the object to be inspected has a defect or foreign substance, picture elements having values other than 0 appear concentratedly in that region where the image of the object has a defect or foreign substance. In this instance, therefore, the defect or the foreign substance can be detected by extracting a group of the picture elements each having a value other than 0 to obtain the size and/or the center of the gravity (centroid).
In the CAD data comparison method (b), image data of a manufactured wafer is compared with CAD image data which is data information of the device to be inspected designed by the CAD. On the basis of this comparison process, a defect or an alien substance on the device to be inspected can be detected similarly to the above-mentioned golden device comparison method (a).
In the die to die comparison method (c), the fact is utilized that a plurality of identical chips (hereinafter referred to as dies) are arrayed or disposed on a single wafer, and image data of adjacent dies are compared with each other. On the basis of this comparison process, a defect or a foreign substance on the device to be inspected can be detected similarly to the above-mentioned golden device comparison method (a) and the CAD data comparison method (b).
In the FFT method, the features are utilized that a defect on a wafer locally occurs, whereas a wiring pattern or printed circuit pattern thereon is periodic, and a defect on a wafer to be inspected is detected by removing a wiring pattern or printed circuit pattern thereon. That is, a defect on the wafer to be inspected can be detected by Fourier-transforming an input wafer pattern image by application of a two dimensional FFT (Fast Fourier-Transform) or the like thereto, removing a specific space frequency component corresponding to a wiring pattern or printed circuit pattern in a space frequency region by utilization of a band stop filler or the like, and thereafter, applying an inverse FFT thereto.
In the case that the above-mentioned golden device comparison method (a) is utilized, a golden device image must be provided in advance. In order to create a golden device image, a sample having no defect must be found out first. The operation for finding out the sample having no defect must carefully be performed by human eyes. In addition, in view of a recent tendency of “various kinds of products and a small amount of production of each”, there are provided various kinds of semiconductor wafer patterns and many changes in pattern design are effected. A golden device image must be created for each of all of those varieties and/or design changes. This requires a considerable labor or work load.
The above-mentioned CAD data comparison method (b) is efficient because it is not necessary to look for a golden device by human eyes like the above-mentioned item (a). Incidentally, in recent years, a high-density integration of a semiconductor chip has been in progress and hence, a conventional optical microscope cannot be used for acquiring a detailed image of a wafer pattern. For this end, a scanning electron microscope (SEM) or the like has been in use. A SEM image contains therein more noise as compared. to an optical microscope image. Therefore, when a differential image is created between CAD data having no noise and a SEM image, a disadvantage is caused that many noises other than defects are detected.
Furthermore, the methods of the above-mentioned items (a) and (b) both require a precise registration or positioning of the two images at the first time upon performing the comparison between both images.
In the above-mentioned die to die comparison method (c), no registration or positioning of two dies (images) is basically required because adjacent dies are compared with each other. However, there is a drawback in this method that if the two dies should have similar defects at the same positions thereof by accident, those defects cannot be detected. In addition, similarly to the above-mentioned item (b), noises other than defects are detected if the comparison is applied to a SEM image.
In the above-mentioned FFT method (d), it is possible to detect a defect on a device even though a SEM image thereof was used for comparison. However, this method requires a very long processing time. For example, when a two dimensional Fourier-transform is applied to a digital image comprised of 512×512 pixels, approximately 1.31×10
8
additions/subtractions and 8.39×10
8
multiplications are required. In addition, a complicated wiring or printed circuit pattern of an MPU or the like cannot be correctly removed by a filtering process in a space frequency region.
SUMMARY OF THE INVENTION
It is a first object of the present invention to provide an image information processing apparatus which can detect a defect on a device even from an unclear SEM image thereof from which a defect on the device cannot be correctly detected by use of the image comparison methods of the above-mentioned items (a), (b) and (c) because of a strong influence of noise.
It is a second object of the present invention to provide an image information processing apparatus which can detect a defect on a device from an independent image thereof to be inspected without using a golden device or a CAD data required in the image comparison methods of the above-mentioned items (a) and (b).
It is a third object of the present invention to provide an image information processing apparatus which can detect a defect on a device at
Advantest Corporation
Gallagher & Lathrop
Kassa Yosef
Lathrop David N.
Mancuso Joseph
LandOfFree
Image information processing apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image information processing apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image information processing apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2455946