Image analysis – Image enhancement or restoration – Object boundary expansion or contraction
Reexamination Certificate
1998-08-17
2001-12-04
Lee, Thomas D. (Department: 2724)
Image analysis
Image enhancement or restoration
Object boundary expansion or contraction
C382S282000, C382S215000, C358S453000
Reexamination Certificate
active
06327393
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to machine vision, and particularly to a user interface for visually partitioning and optionally transforming a region of interest within an image for further processing by a machine-vision vision system.
BACKGROUND
Optical character recognition (“OCR”) systems are used in many industrial applications. In semiconductor manufacturing processes, OCR is used to read laser-etched serial numbers on wafers at many stages of the process, such as during test and assembly steps. In a typical wafer reading application, a machine-vision vision system is directly integrated with a piece of process equipment, such as a wafer handler or sorter, or is positioned at a dedicated reading station. In either case, the vision system uses an imaging device to capture an image of each semiconductor wafer. The vision system then compares each character in the wafer serial number to a pretrained font stored in memory. The character is identified when a match in memory is located.
Size differences between the image of the characters and the pretrained fonts often hamper proper identification of the characters. More particularly, the pretrained fonts stored in memory form a template. Typically, the template is a matrix of pixels. Each of the characters of the image are also represented as a matrix of pixels. When the template is compared with the character in the image, each pixel of the image is compared with each pixel of the template. The character can only match the template if the character size in the image of the wafer and the character size of the template are substantially the same. When the size of the character in the image and the size of the pretrained font differ, no match is found, and, therefore, the character is not identified.
To correct the problem, typically the image of the wafer is scaled so that it will match the appropriate template. The machine-vision vision system determines the size of the characters in the image of the wafer as a percentage of the size of the characters stored in memory by using calibration data obtained during set-up, such as magnification. The vision system then scales the stored font so that it can accurately identify the characters at run-time.
The OCR calibration step typically requires an operator to describe how the source images are distorted. Commonly, the operator describes the source image by entering numeric values of angles and coordinates using a keyboard. When the area that needs to be characterized by the user is irregular, such as a polar region on a wafer, it becomes intuitively difficult for the operator to guess an arc that fits the area.
SUMMARY
A method is disclosed for partitioning a region of interest in a digital image using a deformable window, the region of interest is then parameterized and input into a machine-vision vision system for further processing. The method allows a user to intuitively segment the region of interest using a deformable window. The deformable window is displayed on a graphical user interface superimposed upon the digital image.
A user specifies a shape and position of the deformable window so as to enclose the region of interest that will be processed. The deformable window enables the user to identify the region of interest graphically, eliminating the need for the user to conceptually estimate the region and input numerical parameters for the region. The deformable window is defined by using a pointing device that, in a preferred embodiment, can be used to translate, rotate, curve, skew, and scale the deformable window.
Thereafter, the region of interest is parameterized and transformed to a destination window, and the parameters of the destination window are input into a machine-vision vision system for further processing. Optionally, the destination window can have any location or shape specified by the user, thereby, enabling further transformation of the input image.
The method and apparatus are particularly useful in OCR applications, wherein the destination window is a rectangular window appropriate for typical OCR processing. In this aspect, an embodiment of the method directed to parameterizing a polar region provides an advantage over the prior art. An advantage over the prior art is also provided by a further embodiment of the method that transforms the parameterized polar region. The transformed polar region is then input into an OCR application.
One advantage of the method of the invention is that a user no longer needs describe the region of interest by inputting parameters. The deformable window provides a more direct method for the user to communicate that information without requiring the user to characterize the region using numerical parameters. Another advantage of the invention is the ease with which an arc or other non-rectilinear regions of interest are identified and parameterized. In addition, these regions can then be transformed to a destination window having a size and shape of the user's choice that can then be used as input for further processing.
In further aspects, the invention provides an apparatus in accord with the method described above. The aforementioned and other aspects of the invention are evident in the drawings and in the description that follows.
REFERENCES:
patent: 5428721 (1995-06-01), Sato et al.
patent: 5485561 (1996-01-01), Iizuka et al.
Cognex 3000/4000/5000 Vision Tools Manual, Revision 7.4 590-0136, Cognex Corporation, Chapter 8—Optical Character Recognition, 1996, pp. 353-430.
Cognex 3000/4000/5000 Vision Tools Manual, Revision 7.4 590-0136, Cognex Corporation, Chapter 12—Scene Angle Finder, 1996, pp. 501-515.
Cognex Checkpoint Machine Vision System User's Guide, Cognex Corporation, Chapter 11—Using the Region Tool, 1995, pp. 311-318.
Calabresi Tracy
Cognex Corporation
Lee Thomas D.
LandOfFree
Method and apparatus to transform a region within a digital... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and apparatus to transform a region within a digital..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and apparatus to transform a region within a digital... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2570619