Optics: measuring and testing – By configuration comparison – With object being compared and scale superimposed
Reexamination Certificate
2001-11-29
2004-09-28
Fuller, Rodney (Department: 2851)
Optics: measuring and testing
By configuration comparison
With object being compared and scale superimposed
C356S388000
Reexamination Certificate
active
06798515
ABSTRACT:
FIELD OF THE INVENTION
This invention relates to imaging systems, and more particularly to methods and apparatuses for calibrating an imaging system.
BACKGROUND
In imaging applications, one often wants to measure the distance between two or more points in an image and then convert the result to some physical scale. The conversion between the image distance (i.e. pixels) and the physical distance (e.g. microns) is governed by a scale relationship, where the scale relationship is typically calculated by calibrating an imaging system.
As known in the art, the imaging system typically includes an imaging element, such as a charge-coupled-device (“CCD”) camera, for example, a digitizing element, such as a frame grabber or digital CCD, for example, and optionally, an external optical system, such as a microscope. As known in the art, the imaging system also includes the relationship(s) between the imaging element, the external optical system, if present, and the physical object(s) that are being imaged, i.e. the subject.
Calibration provides both the scale relationship, that relates image dimensions to physical dimensions, and positional information, that relates any point in the image to a physical-coordinate system, such as that which governs a robotic arm, for example.
FIG. 1
is an illustration of a simple calibration device
100
, which can be used to provide the scale relationship, and two images
110
,
114
of the calibration device. The calibration device
100
includes a rectangle
102
and two crosses
104
,
106
, which are positioned on opposite ends of the rectangle
102
. To calibrate, at least partially, an imaging system with the calibration device
100
, the distance
108
, designated d
pu
(physical units) herein, is measured in physical units, such as microns or millimeters, for example. Then, the image
110
is acquired of the calibration device
100
positioned horizontally, and the distance
112
, which is designated d
x
(pixels) herein, is measured in pixels. The horizontal resolution, designated r
x
, is computed by dividing the distance
112
in pixels by the distance
108
in physical units, i.e. r
x
=d
x
/d
pu
. Subsequently, the image
114
is acquired of the calibration device
100
positioned vertically, and the distance
116
in the image
114
, which is designated d
y
(pixels) herein, is measured. The vertical resolution, designated r
y
, is computed as r
y
=d
y
/d
pu
. With the x and the y resolution of the imaging system, a distance between any two points in an image acquired with that imaging system could be accurately measured and expressed Where the distance in physical units, designated l, is given as:
l
=sqrt((
h/r
x
)
2
+(
v/r
y
)
2
)
where l=the length in physical units,
h and v are the horizontal and vertical distance in pixels, respectively, and r
x
and r
y
are the horizontal and vertical resolution, respectively.
The above-described calibration uses a separate calibration object
100
and calibration image(s)
110
,
114
to provide the scale relationship between physical dimensions and image dimensions.
Other known calibration methods use one or more calibration images and/or object(s), where the other methods can be more complex or less complex than the above-described example.
Calibrating an imaging system can also compensate for one or more distortions. The above-described calibration removes and/or minimizes non-square pixel distortion, a term known in the art. Other more complex calibration methods, such as a grid of dots, remove and/or minimize optical distortion, for example. As known in the art, calibrating with a grid of dots is imaging a rectangular array of filled circles of known dimensions, located at predetermined points, and spaced apart at known distances. Also, the grid may, or may not, have markers that indicate the direction of the x- and y-axes and the location of the origin of the grid of dots. The imaging system is calibrated by comparing the image of the grid of dots to its known physical description.
Further, calibrating an imaging system can also provide positional information For example, positional information is determined by placing the calibration device
100
at a known physical position before acquiring the calibration images
110
,
114
.
Typically in industry, a system is calibrated at the beginning of a task, and the system is not re-calibrated for days and sometimes even months. If parts of the imaging system change, the scale relationship determined during calibration is not accurate for subsequent images. Consequently, any measurements of features derived from an image, such as potential defects, for example, are not reliable. Inspecting an image using an unreliable scale relationship can cause one to miss defects or erroneously classify an acceptable feature as a defect, i.e. false positives.
SUMMARY
A method is disclosed for, at least partially, calibrating an imaging system, where partially calibrating constitutes at least calculating a scale relationship between an image and the real world, such as 10 pixels equals 4 microns, for example. First, the method selects a characteristic associated with an object that has a known value. The known value of the characteristic and a measured value of the characteristic are used to calculate the scale relationship, where the measured value is derived from an image of the characteristic. The image is acquired by the imaging system that is being calibrated In the image, the characteristic is found and measured to provide the measured value. Lastly, once the scale relationship is calculated, the scale relationship is used to process the same image from which it was derived.
The invention recognizes that a known value of a characteristic of an object can be leveraged to calculate a scale relationship for an imaging system. Further, the invention recognizes that by using the known value, a single image can be processed, such as inspected for example, and used to calculate the scale relationship of the imaging system. Thus, the invention eliminates, or decreases the need for a separate calibration image(s) and/or object(s).
Several examples of characteristics are described including the value of: an inherent feature, a boundary thereof, and/or an aspect(s) thereof, such as a dimension(s), a provided feature, such as a fiducial, a boundary thereof, and/or aspect(s) thereof, or a relationship between more than one feature, such as a distance between centers of features, for instance.
In one embodiment, the imaging system is re-calibrated using more than one image of the same or multiple objects.
In a preferred embodiment, the image is inspected using the scale relationship derived therefrom. Specifically, the scale relationship, calculated from the known value and the measurement from the image, is used to calculate the physical dimensions and positions of defects located during the subsequent inspection.
In a preferred embodiment, the method is accomplished, in part, using a model of at least part of the object, where the model includes the characteristic that has the known value. For example, as hereinafter described, a model of an annular cladding of a fiber-optic end-face is created using the diameter of the annular cladding in microns, where the value of the diameter in the physical world is the known value. The model is used to find, and optionally measure, a pixel diameter of the annular cladding in the image.
The method is particularly suited for calculating a scale relationship of an image of a fiber-optic end-face acquired by an imaging system. As known in the industry, during imaging the fiber-optic end-face is not fixed in the z-plane. There is some leeway, because of the physical set-up, described hereinafter. Therefore, as recognized by the invention, without re-calibrating the imaging system, the movement of the fiber-optic end-face in the z-plane makes the scale relationship originally calculated for the imaging system unreliable. Consequently, the unreliable scale relationship will cause errors in apparent size
Bachelder Ivan
Wagman Adam
Calabresi Tracy
Cognex Technology and Investment Corporation
Fuller Rodney
LandOfFree
Method for calculating a scale relationship for an imaging... 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 for calculating a scale relationship for an imaging..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for calculating a scale relationship for an imaging... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3203186