Image analysis – Image transformation or preprocessing – Measuring image properties
Reexamination Certificate
1998-01-16
2001-02-06
Couso, Yon J. (Department: 2723)
Image analysis
Image transformation or preprocessing
Measuring image properties
C382S216000
Reexamination Certificate
active
06185343
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
This invention relates to a position detection system and method using image processing so as to find a rotation angle and a coordinate for a check object.
2. Description of the Related Art
Hitherto, the following method has been known as a method of finding a rotation angle from a reference position for a check object contained in an input image (check image) given by picking up an image by an image input unit such as a TV camera: The concentration difference between the check object and background is used to binarize the input image with respect to the concentration, thereby separating the check object from the background, then a moment of inertia of the check object is found and the center position of the check object and the rotation angle of a main axis are calculated. This kind of art is adopted for position detection of parts on a manufacturing line. For example, it is used for a robot eye to control a robot for detecting orientation of supplied parts, holding the parts, and attaching the parts to predetermined positions.
The art for finding the main axis of the check object as described above involves the following problem: If the outline of the check object is a circle, the main axis cannot be found. Thus, if a character, etc., is written in check object P like a coin and the check object P contains caorientation (see FIGS.
6
(
a
) and
6
(
b
)), the rotation angle of the check object P cannot be found. If the outline has rotation symmetry like a square, the main axis cannot be determined one, thus a similar problem arises. Further, even if main axis M (see FIGS.
7
(
a
) and
7
(
b
)) can be found, information representing orientation is not contained in the main axis M. Thus, for example, when the check object P is symmetric with respect to a line as shown in FIGS.
7
(
a
) and
7
(
b
), if it differs in position 180 degrees, the check object P shown in FIG.
7
(
b
) cannot be distinguished from that in FIG.
7
(
a
).
In contrast, a pattern matching method of collating a check object with a previously entered template, thereby finding the rotation angle of the check object is known as described in Japanese Patent Publication No. Hei. 6-89341. In the art described here, particularly, an image at a corner of a part is collated as the check object, thereby decreasing the processing amount. It is considered that the art can be used to detect the rotation angle of the check object as shown in FIG.
7
. However, since only the image at the corner is used, when the outline is a circle or a square, the problem remains unsolved.
Further, the following method has also been known as a method of finding a position (rotation angle and coordinate) for a check object contained in an input image (check image) given by picking up an image by an image input unit such as a TV camera:
As this kind of position detection system, hitherto a system disclosed in Japanese Patent Publication No. Hei. 6-160047 has been available. This system performs rough detection processing of detecting a rough position of an approximate position of a check object in a check range based on the comparison result of comparison processing of comparing images provided by compressing the check image and a correct solution image subjected to the detection displacement operation of displacing by rough detection displacement amount at a time in the check image, then performs precise detection processing of detecting a more precise position of the check object in the check range in the proximity of the rough position based on the comparison result of comparison processing of comparing the images provided by compressing the check image and the correct solution image subjected to the detection displacement operation of displacing by rough detection displacement amount at a time in the check image at a compression rate more moderate than the previous compression rate, and detects the position of the check object in the check range in multiple steps.
More particularly, the position detection system performs compression processing by extracting (or thinning out) some of the pixels making up the correct solution image.
The conventional position detection system, which detects the position of the check object in the check range in multiple steps, takes a shorter detection time than the time taken for accurately detecting the position of the check object in the check range from the beginning by one-step detection processing.
However, since the position detection system performs compression processing by extracting or thinning out some of the pixels making up the correct solution image, when the pixel extracted from the pixels making up the correct solution image does not have an average value and has a peculiar value, the position of the check object in the check range cannot precisely be detected (for example, when the noises are picked up).
SUMMARY OF THE INVENTION
To solve the aforementioned problems, an object of the present invention is to provide a position detection system and method using image processing and being capable of finding a rotation angle and a coordinate of a check object regardless of the shape of the check object.
According to a first aspect of the invention, there is provided a method for detecting a position of a check object using image processing, comprising the steps of: picking up an image of the check object so as to prepare a check image of the check object; preparing a reference template having a reference image of the check object; preparing a plurality of collation templates by rotating the reference template at various rotation angles; collating the collation templates with the check image so as to evaluate a correlation degree therebetween; and acknowledging a rotation angle of the reference template for preparing the collation template, which matches with the check image most exactly, as a rotation angle of the check object.
According to a second aspect of the invention, there is provided a method for detecting a position of a check object using image processing, comprising the steps of: (1) picking up an image of the check object so as to prepare a check image of the check object; (2) preparing a correct solution image having a reference image of the check object; (3) compressing the check image and the correct solution image at a given compression rate by an average compression processing which calculates an average value of a brightness in a predetermined area and adopts the average value as a brightness of one pixel of the compressed check image and the compressed correct solution image; (4) displacing the compressed correct solution image; (5) comparing the compressed check image and the compressed/displaced correct solution image so as to detect a rough position of the check object; (6) compressing the check image and the correct solution image after detecting the rough position at a compression rate more moderate than the previous compression rate by the average compression processing; (7) displacing the compressed correct solution image in a proximity of the rough position; (8) comparing the compressed check image and the compressed/displaced correct solution image so as to detect a more precise position than the rough position of the check object; and (9) repeating (3) to (8) steps while moderating the compression rate in multiple steps so as to detect the precise position of the check object.
According to a third aspect of the invention, there is provided a system for detecting a position of a check object using image processing, comprising: means for picking up an image of the check object so as to prepare a check image of the check object; means for preparing a reference template having a reference image of the check object; means for preparing a plurality of collation templates by rotating the reference template at various rotation angles; means for collating the collation templates with the check image so as to evaluate a correlation degree therebetween; and means for acknowledging a rotation angle of the reference tem
Hashimoto Yoshihito
Ikeda Kazutaka
Couso Yon J.
Matsushita Electric & Works Ltd.
Sughrue Mion Zinn Macpeak & Seas, PLLC
LandOfFree
Position detection system and method does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Position detection system and method, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Position detection system and method will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2614961