Optics: measuring and testing – Inspection of flaws or impurities – Containers or enclosures
Reexamination Certificate
2000-09-07
2002-12-10
Stafira, Michael P. (Department: 2877)
Optics: measuring and testing
Inspection of flaws or impurities
Containers or enclosures
C382S141000
Reexamination Certificate
active
06493079
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to the field of machine vision applications, and more particularly to a system and method for performing machine vision analysis of an object using a reduced number of cameras.
DESCRIPTION OF THE RELATED ART
In many machine vision applications, machine vision analysis is required of a three dimensional object, such as a rectangular prism or cube. For example, in manufacturing applications, machine vision analysis may be used to detect defects in a manufactured object by acquiring images of the object and using various types of image processing algorithms to analyze the images. As an example, a system to manufacture electrical components such as capacitors may use machine vision to examine each side of the capacitors in order to detect manufacturing defects, ensure that the capacitors are labeled, marked, or color coded properly, etc.
Machine vision applications may use image processing software operable to perform any of various types of image analysis or image processing functions or algorithms in examining an acquired image of an object. For example, pattern matching algorithms are often used, e.g., in order to compare the pattern information of the acquired image to the pattern information of a template image. Color matching algorithms may also be used, e.g., in order to compare the color information of the acquired image to the color information of a template image. Blob (Binary Large Object) analysis tools may also be used to examine an image. In various applications, pattern, color and/or Blob analysis information may be used to verify that: an object includes all necessary components in the correct locations, an object has the appropriate words, labels, or markings, an object surface is not scratched or otherwise defective, etc.
Any of type of camera or other device may be used to acquire the images to be analyzed in a machine vision application, including digital cameras, frame grabbers, line scan cameras, etc. As used herein, the term “camera” is intended to encompass all such devices.
Prior art machine vision systems have required a separate image to be acquired and analyzed for each individual side of an object for which machine vision analysis is necessary. For example, in order to analyze a capacitor having six sides, a prior art system uses
6
cameras to acquire six separate images and analyze each of these images.
FIG. 1
illustrates a prior art system in which a camera
10
is faced perpendicularly to a side of a cube-shaped object. Thus, the camera
10
may acquire an image including visual information of side
2
of the object but not including visual information from the other sides of the object. In order to analyze other sides of the object, additional images must be acquired. For example,
FIG. 2
illustrate a prior art system including four cameras
10
A,
10
B,
10
C, and
10
D, in which the cameras are able to acquire images of sides
2
,
1
,
3
, and
6
, respectively, of the object. In order to image the top side
4
and the bottom side
5
of the object, two additional images must be acquired, which requires two additional cameras in prior art systems.
In various situations, there are several disadvantages associated with the prior art approach described above. For example, the acquisition and processing of a separate image for each object side of interest may be inefficient. Thus, it may be desirable to reduce the number of images that need to be acquired in order to examine the object, and hence reduce the number of cameras required in a machine vision system. For example, the cost of a having a separate camera dedicated to each side of the object may be prohibitive. Also, it may be desirable to position the cameras in locations other than perpendicularly facing the object sides. For example, due to the physical environment of a particular manufacturing facility, it may be necessary to position the cameras elsewhere, e.g., facing the object along a diagonal axis extending through opposite comers of the object.
SUMMARY OF THE INVENTION
One embodiment of the present invention comprises a system and method for analyzing a physical object (or a plurality of physical objects) using a reduced number of cameras and/or a reduced number of acquired images. The object may be shaped in various ways and may have any of various sizes. In one exemplary embodiment, the object is shaped like a cube or rectangular prism having six sides.
The system and method may be utilized in a machine vision application, e.g., in order to inspect manufactured objects. The method may operate to acquire an image, or a plurality of images, of the object. hi order to reduce the number of images that need to be acquired, at least one of the acquired images includes visual information for two or more sides of the object. In one embodiment, all of the acquired images include visual information for two or more sides of the object. Such an image may be acquired by a camera positioned in any of various ways such that the camera can capture visual information for multiple sides of the object.
In one embodiment, at least one of the acquired images includes visual information for three or more sides of the object. For example, with respect to an object shaped as a six-sided rectangular prism, such an image may be acquired by a camera positioned along an imaginary diagonal axis passing through the middle of the object that connects two opposite comers of the rectangular prism, wherein one comer is defined by the intersection of the three planes comprising the three sides of the rectangular prism for which visual information is acquired and the other comer is defined by the intersection of the three planes comprising the other three sides of the rectangular prism (see FIGS.
5
and
6
). Such an image may also be acquired by a camera that is not positioned exactly along this axis but is positioned similarly such that the camera can acquire visual information for three sides of the object.
The acquired images of the object may then be received by a computer system and analyzed using image processing software. One embodiment of the invention comprises image processing software operable to analyze the acquired images, including the image(s) that each include visual information for multiple sides of the object. The image processing software may be operable to separate the visual information associated with each of the multiple sides of the object captured in an acquired image, e.g., to separate ROIs associated with each of the multiple sides captured in the acquired image. The image processing software may then separately analyze the visual information associated with each of the multiple sides of the object.
In one embodiment, the images may be received by a plurality of computer systems, or a plurality of image acquisition devices coupled to or comprised in a single computer system, in order to perform the analyses at a faster rate. For example, for each camera, a separate computer system connected to the camera may receive the image acquired by the camera and analyze the received images. Alternatively, for each camera, a separate image acquisition device or board coupled to the camera may capture the image, and on-board processor may analyze the captured image, or a host computer may analyze captured images from each of the image acquisition devices.
For each image that includes visual information for multiple sides of the object, analyzing the image may comprise determining regions of interest (ROIs) within the image, wherein each region of interest corresponds to one side of the object. The regions of interest within the image may be determined in various ways. For example, in one embodiment, the system is operable to acquire images such that the regions of interest have predetermined boundaries within the image. For example, the image acquisitions may be precisely coordinated with the movement and positioning of the object on an assembly line, using timing or motion control techniques. In another embodiment, the regions of interest co
Burgess Jason L.
Conley Rose & Tayon PC
Hood Jeffrey C.
National Instruments Corporation
Stafira Michael P.
LandOfFree
System and method for machine vision analysis of an object... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System and method for machine vision analysis of an object..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System and method for machine vision analysis of an object... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2960817