JP4210126B2 - Image processing method, image processing apparatus, image processing program, and recording medium on which image processing program is recorded - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an image processing method and an image processing device that perform graphic processing on image data input from an imaging device or the like, and in particular, an image processing method and an image processing device that detect the center of a circular image included in image data, and The present invention relates to an image processing program and a recording medium on which the image processing program is recorded.
[0002]
[Prior art]
In order to realize automation of product assembly and inspection work with high accuracy, an image processing apparatus that captures images of products and parts with a camera and performs various image processing on the obtained image data is required. Yes.
[0003]
In particular, when a product case or part has a circular part and this part is used as an alignment mark or inspection part during assembly, the center of the circle is detected by image processing, and the detected center is used as a reference. Move and inspect parts.
[0004]
The casing and parts have chips and burrs, and as shown in FIG. 6, if these are present in a circular portion, the detection accuracy at the center of the circle is greatly reduced. Therefore, many techniques for detecting the circle center of image data with high accuracy have been developed.
[0005]
For example, in the circle center position measurement method described in Patent Document 1, as shown in the circle center position detection flowchart of FIG. 7, in step A1, an edge point is detected by a differential filter for an image captured by the camera, and in step A2, From the obtained set of edge points, circular arc approximation is performed by the mean square error minimum method to estimate the position of the center of the circle, and in step A3, the set of edge points excluding the points away from the estimated circle circumference For, the circle approximation is again performed by the mean square error minimum method to estimate the position of the circle center.
[0006]
The feature of this method is that the center of the circle can be detected with high accuracy even if there is dust or a chip on the circumference by performing the circle approximation twice.
[0007]
The circular pattern measurement / position recognition device described in Patent Document 2 calculates an approximate expression of a circle from the detected edge point, and provides a reference point within a predetermined range from the center of the circle based on the approximate expression. The distance from each reference point to the edge point is calculated, a histogram of distance and frequency is created, and the reference point constituting the histogram having the highest frequency is set as the center of the circle.
[0008]
The feature of this apparatus is that it is possible to detect the center of a circle with high accuracy even when noise is present by removing noise detected as an edge by using a histogram.
[0009]
[Patent Document 1]
Japanese Patent Laid-Open No. 7-225843
[Patent Document 2]
JP-A-7-210654
[0010]
[Problems to be solved by the invention]
With the above-described conventional technology, it is possible to detect the center of a circle with high accuracy by eliminating the influence of chipping and noise on one circle. However, in practice, the center of the circle may be detected for two or more circular shapes. For example, in the alignment of parts, the alignment marks provided on the reference plane and the parts may have circular shapes with different radii, and the parts may be moved to match the centers of the circles. At this time, if the part circle overlaps the circumference of the reference plane while moving the part, the conventional technique calculates the center of the circle with the part circle missing, and the circle on the circumference There is a problem that the center of the circle cannot be calculated. Furthermore, in addition to noise and chipping, foreign matters are mixed, and if the circle is divided due to the foreign matters, it is difficult to accurately detect the center of the circle with the conventional technology.
[0011]
An object of the present invention is to provide an image processing method, an image processing apparatus, an image processing program, and a recording medium on which the image processing program is recorded, which detects each center position with high accuracy for a plurality of circles displayed as image data. It is.
[0012]
[Means for Solving the Problems]
The present invention is an image processing method for detecting the center position of each circle with respect to a plurality of circles including a circle having a size different from that of other circles displayed as image data,
A great circle region identification step for identifying a region including a great circle which is the largest circle among a plurality of circles;
A great circle candidate pixel detecting step of detecting edge pixels in the great circle region and detecting the detected edge pixels as great circle candidate pixels constituting the circumference of the great circle;
A great circle detection step for performing a multi-stage circle approximation on the detected great circle circumference candidate pixels and detecting the center position and radius of the great circle;
A small circle region specifying step of labeling within the great circle so as not to include the great circle edge based on the detected center position and radius, and specifying one or a plurality of regions including one small circle;
A small circle candidate pixel detecting step for detecting edge pixels in the small circle region, and detecting the detected edge pixels as small circle candidate point pixels constituting the circumference of the small circle;
A small circle detection step of performing multi-stage circle approximation on the detected small circle candidate pixels and detecting a center position and a radius of the small circle.
[0013]
According to the present invention, first, in the great circle region specifying step, a region including the great circle which is the largest circle among the plurality of circles is specified. In the great circle candidate pixel detection step, an edge pixel in the great circle region is detected, and the detected edge pixel is detected as a great circle candidate pixel constituting the circumference of the great circle.
[0014]
In the great circle detection step, multi-stage circle approximation is performed on the detected great circle candidate pixels, and the center position and radius of the great circle are detected. At this step, the center position of the great circle is determined.
[0015]
Next, in the small circle region specifying step, one or a plurality of regions including one small circle are specified by labeling within the large circle so as not to include the large circular edge based on the detected center position and radius. In the small circle candidate pixel detection step, an edge pixel in the specified small circle region is detected, and the detected edge pixel is detected as a small circle candidate point pixel constituting the circumference of the small circle.
[0016]
Multi-stage circle approximation is performed on the detected small circle candidate pixels in the small circle detection step to detect the center position and radius of the small circle. In this step, the center position of the small circle is determined.
[0017]
In this way, the center position and radius of the great circle are first detected, the great circle is determined from the detected center position and radius, and the center position of the small circle is detected within the determined great circle. Even when small circles overlap on the circumference, each can be detected with high accuracy.
[0018]
Further, in the great circle region specifying step, the present invention separates pixels constituting the background and pixels constituting other than the background by binarization processing,
A great circle region is specified based on a feature amount of a set of pixels constituting a portion other than the background.
[0019]
According to the present invention, the great circle is, for example, a marker provided in a part or a screw hole, so that the color and the surface state of the great circle are significantly different from those other than the great circle, so-called ground portions. Therefore, even in the image data when this is imaged, the pixel values are greatly different. Therefore, it is possible to accurately separate by binarization processing by appropriately setting the threshold value. In addition to the pixels constituting the circle, there are also pixels caused by foreign matters such as dust and noise. Since the pixels forming the circle form a set of a plurality of pixels, the region where the great circle exists is specified from the feature amount of the set.
[0020]
As described above, since the great circle region is specified from the pixel separation by the binarization processing and the feature amount of the pixel set, the region can be specified easily and with high accuracy.
[0021]
Further, in the small circle region specifying step, the present invention separates pixels constituting the great circle and pixels constituting other than the great circle by binarization processing,
A small circle region is specified based on a feature amount of a set of pixels constituting a portion other than the great circle.
[0022]
According to the present invention, the small circle is a marker or a screw different from the great circle, and therefore the density is greatly different even in image data when the small circle is imaged. Therefore, by appropriately setting the threshold value, the inside of the great circle can be separated with high accuracy by binarizing labeling. Since the pixels forming the small circle form a set of a plurality of pixels, the region where the small circle exists is specified from the feature amount of the set.
[0023]
As described above, since the great circle region is specified from the pixel separation by the binarization processing and the feature amount of the pixel set, the region can be specified easily and with high accuracy.
[0024]
In the invention, it is preferable that the feature amount is the number of pixels included in a set of pixels and a ferret diameter of the set of pixels.
[0025]
According to the present invention, the number of pixels included in the pixel set or the ferret diameter of the pixel set is used as the feature amount for specifying the great circle or the small circle. Pixels caused by foreign matters such as dust or noise are often smaller than circles, and therefore a set having a relatively large number of pixels, that is, a large area, is specified as a circle. In addition, when a large foreign object has been imaged, a circle may not be specified only by the number of pixels included in the pixel set. However, in the case of a foreign substance, since the shape is not fixed, a ferret diameter is used, and a set in which the ratio of the horizontal ferret diameter to the vertical ferret diameter is close to 1 may be specified as a circle.
[0026]
As described above, since the region where the circle exists is specified by the area of the set of pixels and the diameter of the ferret, it is possible to prevent erroneous detection of detecting foreign matter or noise as a circle.
[0027]
Further, according to the present invention, the multi-stage circle approximation includes a process for detecting a center position and a radius based on a circumference candidate pixel, and a circle from a circumference based on the detected center position and radius among the circumference candidate pixels. The process of making a pixel whose distance is closer than a predetermined distance as a new circumference candidate pixel is one step, and this is repeated a plurality of times.
[0028]
According to the present invention, the multi-stage circle approximation is performed by repeating the following one-stage process. The one-stage process consists of two processes. First, as the first process, the center position and the radius are detected based on the circumference candidate pixels. Next, as a second process, a pixel whose distance from the circumference of the approximate circle determined from the detected center position and radius among the circumference candidate pixels is closer than a predetermined distance is set as a new circumference candidate pixel. Returning to the first process again, a new approximate circle is detected from the new circumference candidate pixels obtained in the second process. By repeating these processes, the center position of the circle can be detected with high accuracy even when the circumference is missing or other circumferences overlap. In addition, it is desirable that the steps are repeated 3 to 5 times.
[0029]
Further, the present invention is based on the ratio of the pixels excluding the pixels constituting the great circle and the pixels constituting the background out of all the pixels in the small circle based on the center position and radius detected in the small circle detection step. The method further includes a verification step for verifying the validity of the small circle.
[0030]
According to the present invention, all the pixels in the small circle based on the center position and the radius detected in the small circle detection step are counted. For example, if a foreign object is partly arc-shaped, and there is a false detection that would detect this arc as part of the circumference, In many cases, the pixel is a constituent pixel. Therefore, when the ratio of the pixels excluding the pixels constituting the great circle and the pixels constituting the background is small, for example, 50% or less, it is determined as a false detection.
[0031]
Thus, by verifying the validity of the small circle once detected, the center position of the circle can be detected with higher accuracy.
[0032]
The present invention is also an image processing program for causing a computer to execute the above image processing method.
[0033]
According to the present invention, an image processing program for causing a computer to execute the above-described image processing method can be provided.
[0034]
The present invention also provides a computer-readable recording medium on which an image processing program for causing a computer to execute the above image processing method is recorded.
[0035]
According to the present invention, it can be provided as a computer-readable recording medium on which an image processing program for causing a computer to execute the above-described image processing method is recorded.
[0036]
Further, the present invention is an image processing apparatus that detects the center position of each circle with respect to a plurality of circles including circles that are different in size from other circles that are displayed as image data,
A great circle region specifying means for specifying a region including a great circle which is the largest circle among a plurality of circles;
Great circle candidate pixel detecting means for detecting edge pixels in the great circle region and detecting the detected edge pixels as great circle candidate pixels constituting the circumference of the great circle;
A great circle detection means for performing multi-stage circle approximation on the detected great circle circumference candidate pixels, and detecting the center position and radius of the great circle;
A small circle region specifying means for labeling within the great circle so as not to include the great circle edge based on the detected center position and radius, and specifying one or a plurality of regions including one small circle;
Small circle candidate pixel detecting means for detecting edge pixels in the small circle region and detecting the detected edge pixels as small circle candidate point pixels constituting the circumference of the small circle;
An image processing apparatus comprising: a small circle detecting unit that performs multi-stage circle approximation on the detected small circle candidate pixels and detects a center position and a radius of the small circle.
[0037]
According to the present invention, first, the great circle region specifying means specifies a region including the great circle which is the largest circle among the plurality of circles. The great circle candidate pixel detecting means detects edge pixels in the great circle region, and detects the detected edge pixels as great circle candidate pixels constituting the circumference of the great circle.
[0038]
The great circle detection means performs multi-stage circle approximation on the detected great circle candidate pixel, and detects the center position and radius of the great circle. The center position of the great circle is determined by the processing of the great circle detection means.
[0039]
Next, the small circle region specifying means performs labeling within the large circle so as not to include the great circle edge based on the detected center position and radius, and specifies one or a plurality of regions including one small circle. The small circle candidate pixel detecting means detects edge pixels in the specified small circle region, and detects the detected edge pixels as small circle candidate point pixels constituting the circumference of the small circle.
[0040]
The small circle detection means performs multi-stage circle approximation on the detected small circle candidate pixels, and detects the center position and radius of the small circle. The center position of the small circle is determined by the processing of the small circle detecting means.
[0041]
In this way, the center position and radius of the great circle are first detected, the great circle is determined from the detected center position and radius, and the center position of the small circle is detected within the determined great circle. Even when small circles overlap on the circumference, each can be detected with high accuracy.
[0042]
Further, the present invention is based on the ratio of the pixels excluding the pixels constituting the great circle and the pixels constituting the ground, out of all the pixels in the small circle based on the center position and radius detected by the small circle detecting means, It further has verification means for verifying the validity of the small circle.
[0043]
According to the present invention, all pixels in the small circle based on the center position and the radius detected by the small circle detecting means are counted. If the ratio of the pixels in the small circle excluding the pixels constituting the great circle and the pixels constituting the background is small, it is determined as erroneous detection.
[0044]
Thus, by verifying the validity of the small circle once detected, the center position of the circle can be detected with higher accuracy.
[0045]
DETAILED DESCRIPTION OF THE INVENTION
FIG. 1 is a flowchart showing an image processing method of the present invention.
[0046]
In the present embodiment, for example, a description will be given of a case where alignment is performed to move a component to a correct position with respect to a reference surface during product assembly and inspection.
[0047]
Specifically, a circular reference marker (fixed) is provided on the reference surface, and this is imaged with a fixed camera. A circular part marker is provided on a part of the part to be aligned, and the part is placed on the reference plane by the moving device. At this initial time point, as shown in FIG. 2A, an image in which the positions of the reference marker M1 and the component marker M2 are shifted is captured by the camera.
[0048]
The purpose of alignment is to make the center position of the reference marker M1 coincide with the center position of the component marker M2. Therefore, image processing is performed on the image at the initial time point, the center position of each marker is detected, and the direction and distance in which the component is moved to match are calculated. Based on the calculated parameters, the moving device moves the component. After the movement, the marker is imaged again. If it is deviated, the center position is detected, the movement parameter is calculated, and the part is moved. This is repeated until the center of the circle of the reference marker M1 and the center of the circle of the part marker M2 match. .
[0049]
As shown in FIG. 2B, the component marker M2 may overlap the circumference of the reference marker M1 at the initial point and when the component is moved. In the present invention, the center position of each marker is detected with high accuracy in such a case.
[0050]
First, in step S1, a region (great circle region) in the image data of the reference marker M1 that is a great circle is specified. Since the reference marker M1 is provided in a different color from the other parts of the reference surface, the imaged image data is subjected to binarization processing, and pixels having pixel values that are significantly different from the background are changed to a great circle. It is detected as a constituent pixel. A region where a great circle exists is identified based on the area of the aggregate region of the detected pixels or the ferret diameter. For example, when the area is based on an area, an area having an area (the number of pixels included in the set) of a predetermined value or more is specified as a great circle area. When based on the ferret diameter, an area in which the horizontal ferret diameter and the vertical ferret diameter are equal to or larger than a predetermined value and the ratio is close to 1 is specified as a great circle area. As a result, a small area and a long and narrow area due to dust or noise are excluded from the great circle area.
[0051]
When a plurality of great circles are included in the captured image data, a plurality of great circle regions are specified, and the following steps S2 to S8 may be performed for each great circle region.
[0052]
Next, in step S2, the candidate of the pixel which comprises the circumference of a great circle is detected. Since the circumference of the great circle is the boundary between the reference marker M1 and the background, the edge pixels in the great circle area identified in step S1 are detected as circumference candidate points.
[0053]
The edge pixel has a large density difference with the adjacent pixel, so the edge difference is calculated by calculating the density difference with the adjacent pixel, and the edge pixel is a pixel whose density difference exceeds a predetermined value, or a Sobel filter, etc. Can be detected by a “differential binarization method” or the like that binarizes after first differentiation.
[0054]
In step S3, multi-stage circle approximation is performed on the circumference candidate point detected in step S2, and the center position and radius of the great circle are detected.
[0055]
The multi-stage circle approximation will be described in detail with reference to FIG.
As shown in FIG. 2B, when the component marker M2 overlaps the circumference of the reference marker M1, the captured image data is a circumference candidate point as shown in FIG. 3A at the time of step S2. An image consisting of
[0056]
When circle approximation is performed on the circumference candidate point, as shown in FIG. 3B, the position deviated from the correct circle center C1 of the great circle due to the influence of the small circle portion is set as the circle center C1 ′ of the approximate circle. The radius is detected as a radius smaller than the radius of the great circle.
[0057]
Next, as shown in FIG. 3C, a ring R having a predetermined width around the circumference of the approximate circle is set, and among the circle candidate points, the circle candidate points located outside the ring. Is excluded.
[0058]
The circumference candidate point in the ring is set as a new circumference candidate point, and circle approximation is performed based on this. At this time, since candidate points that deviate significantly from the circumference of the great circle are excluded, a second approximate circle closer to the great circle than the first approximate circle is detected. A ring is set for the second approximate circle, and circumferential candidate points located outside the ring are excluded.
[0059]
Such circle approximation based on circumference candidate points and exclusion of candidate points outside the ring are set as one stage, and a great circle is determined by repeating a plurality of times. In order to detect an accurate great circle, it is necessary to repeat at least 2 times, and 3 to 5 times are desirable from the relationship between accuracy and throughput.
[0060]
In step S4, the region (small circle region) of the component marker M2 that is a small circle is specified in the great circle determined in step S3.
[0061]
A region where a small circle exists is specified based on the area of the aggregate region of the detected pixels or the ferret diameter. One small circle is included in the small circle region, and when there are a plurality of small circles, a plurality of small circle regions are specified.
[0062]
In step S5, edge pixels in the small circle area specified in step S4 are detected as circumference candidate points for the small circle.
[0063]
In step S6, multi-stage circle approximation is performed on the circumference candidate points of the small circle detected in step S5, and the center position and radius of the small circle are detected. The multistage circle approximation is performed in the same manner as the great circle approximation in step S3.
[0064]
In step S7, it is determined whether or not the center position has been detected for all the small circle regions. If all the small circle regions have been detected, the process proceeds to step S8. If any small circle region that has not been detected remains, the process returns to step S5. .
[0065]
In step S8, it is determined whether there are a plurality of detected small circles. If there is more than one, the process proceeds to step S9, and if there is one, the process proceeds to step S10.
[0066]
In step S9, it is verified whether the small circle detected in step S6 is valid. As in the present embodiment, when it is known in advance that there is one large circle and one small circle, when two or more small circle regions are specified and two or more small circles are detected, one is selected. The other small circles other than the circle are not actually small circles, but are considered to be false detections in which a large circle chip or a foreign object is erroneously detected as a small circle.
[0067]
Even if one small circle region is specified and one small circle is detected, it is conceivable that the small circle is erroneously detected. Therefore, the validity of even one small circle may be verified.
[0068]
The verification method calculates the ratio of the pixels constituting the small circle detected in step S4 in the small circle detected in step S6. If the ratio is less than a predetermined ratio (for example, 50% or less), false detection is detected. to decide. As shown in FIG. 4, when a small circle is detected between an actual small circle and a foreign object, only a small amount of pixels are included in the detected small circle in the circumferential portion. Most of them are pixels constituting a great circle. Therefore, since the ratio of the pixels excluding the pixels constituting the great circle to the small circle is small, it can be understood that the small circle is erroneously detected. If there are a plurality of small circles larger than the predetermined ratio, the one with the largest ratio is determined as a small circle. When there are a plurality of small circles, the small circles are determined in descending order of proportion.
[0069]
In step S10, it is determined whether or not the center position has been detected for all the great circle regions. If all the great circle regions have been detected, the process is terminated, and if there are any undetected great circle regions, the process returns to step S2. .
[0070]
As described above, after detecting the center position of the great circle in advance, the center position of the small circle is detected in the area of the great circle, so that the small circle overlaps the circumference of the great circle. However, the two circle centers can be detected with high accuracy.
[0071]
FIG. 5 is a block diagram illustrating a configuration of the image processing system 100. The image processing system 100 includes an imaging device 1, an image processing device 2, and a display device 3, and constitutes a component positioning system as described above.
[0072]
The imaging device 1 includes a CCD (charge coupled device) camera 11, an A / D (analog / digital) converter 12, a camera controller 13, a D / A converter 14, and a frame memory 15. The CCD camera 11 images a subject, the reference plane and parts in the above-described embodiment, and outputs the received light amount as an analog image signal. The A / D converter 12 converts the analog image signal output from the CCD camera 11 into digital data and outputs it as digital image data. The camera controller 14 stores the digital image data in the frame memory 15 frame by frame and outputs it to the D / A converter 14 for display on the display device 3. The D / A converter converts the digital image data output from the camera controller 14 into an analog image signal corresponding to the display device 3 and outputs the analog image signal to the display device 3. The display device 3 is realized by an LCD (Liquid Crystal Display), a CRT (Cathode Ray Tube) display, or the like, and displays an analog image signal output from the imaging device 1.
[0073]
The image processing apparatus 2 includes a CPU (Central Processing Unit) 21, a RAM (Random Access Memory) 22, a ROM (Read Only Memory) 23, and an I / O (Input / Output) controller 24. The CPU 21 controls the operation of the image processing device 2 based on a control program stored in the ROM 23. The image data being processed and the data being calculated are temporarily stored in the RAM 22. The I / O controller 24 is connected to an input device such as a keyboard and a mouse and a component moving device, and controls these input / output data.
[0074]
The CPU 21 and the ROM 23 constitute a great circle region specifying unit, a great circle candidate pixel detecting unit, a great circle detecting unit, a small circle region specifying unit, a small circle candidate pixel detecting unit, a small circle detecting unit, and a verification unit. Image data is acquired from the frame memory 15 via the camera controller 14 of the apparatus 2 and the image processing shown in the above-described flowchart is executed. When the center positions of the two circles are detected by image processing and, for example, the moving distance and moving direction of the component are calculated, these parameters are output to the moving device. The moving device performs positioning by moving the parts based on the acquired parameters.
[0075]
Another embodiment of the present invention is an image processing program for causing a computer to function as the image processing apparatus 2, and a computer-readable recording medium on which the image processing program is recorded. Accordingly, the image processing program and the recording medium on which the image processing program is recorded can be provided in a portable manner.
[0076]
The recording medium is read by a program reading device provided in a printer or a computer system, whereby an image processing program is executed.
[0077]
As an input means of the computer system, a flat bed scanner, a film scanner, a digital camera, or the like may be used. The computer system includes these input means, a computer that executes image processing and the like by loading a predetermined program, an image display device such as a CRT display and a liquid crystal display that displays the processing results of the computer, and computer processing It consists of a printer that outputs the results to paper. Furthermore, a modem or the like as a communication means for connecting to a server or the like via a network is provided.
[0078]
The recording medium is not limited to be read by the program reading device, and may be a microcomputer memory, for example, a ROM. The recorded program may be accessed and executed by the microprocessor, or the program read from the recording medium may be downloaded to the program storage area of the microcomputer and executed. This download function is assumed to be provided in the microcomputer in advance.
[0079]
Specific examples of the recording medium include a tape system such as a magnetic tape and a cassette tape, a magnetic disk such as a flexible disk and a hard disk, a CD-ROM (Compact Disc-Read Only Memory) / MO (Magneto Optical) disk / MD ( Optical discs such as Mini Disc) / DVD (Digital Versatile Disc), IC (Integrated Circuit) cards (including memory cards) / optical cards, etc., mask ROM, EPROM (Erasable Programmable Read Only Memory), This is a medium that carries a fixed program including a semiconductor memory such as an EEPROM (Electrically Erasable Programmable Read Only Memory) and a flash ROM.
[0080]
In the present embodiment, the computer may have a system configuration that can be connected to a communication network including the Internet, and the image processing program may be downloaded via the communication network. In the case of downloading a program from the communication network in this way, the download function may be provided in advance in the computer or installed from another recording medium. The download program may be executed via a user interface, or a predetermined URL (Uniform Resource
The program may be downloaded periodically from Locater).
[0081]
【The invention's effect】
As described above, according to the present invention, first, the center position and radius of the great circle are detected, the great circle is determined from the detected center position and radius, and the center position of the small circle is detected within the determined great circle. Therefore, even when small circles overlap on the circumference of the great circle, each can be detected with high accuracy.
[0082]
Further, according to the present invention, since the great circle region is specified from the pixel separation by the binarization process and the feature amount of the pixel set, the region can be specified easily and with high accuracy.
[0083]
Further, according to the present invention, the region where the circle exists is specified by the area of the set of pixels and the diameter of the ferret, so that it is possible to prevent erroneous detection of detecting a foreign object or noise as a circle.
[0084]
In addition, according to the present invention, by repeating the detection of the center position and radius and the determination of a new circumference candidate pixel, even when the circumference is missing or other circumferences overlap, with high accuracy. The center position of the circle can be detected.
[0085]
Further, according to the present invention, the center position of the circle can be detected with higher accuracy by verifying the validity of the small circle once detected.
[0086]
Further, according to the present invention, it is possible to provide an image processing program for causing a computer to execute the above image processing method, and a computer-readable recording medium on which the image processing program is recorded.
[Brief description of the drawings]
FIG. 1 is a flowchart showing an image processing method of the present invention.
FIG. 2 is a schematic view of an alignment marker.
FIG. 3 is a diagram showing a procedure of multi-stage circle approximation.
FIG. 4 is a diagram illustrating an example of a small circle that is erroneously detected.
5 is a block diagram showing a configuration of an image processing system 100. FIG.
FIG. 6 is a diagram illustrating an example of a component with a chip or a burr.
FIG. 7 is a flowchart showing a conventional circle center position detection process.
[Explanation of symbols]
1 Imaging device
2 Image processing device
3 display devices
11 CCD camera
12 A / D converter
13 Camera controller
14 D / A converter
15 frame memory
21 CPU
22 RAM
23 ROM
24 I / O controller
100 Image processing system

Claims (10)

An image processing method for detecting a center position of each circle with respect to a plurality of circles including a circle having a size different from that of other circles displayed as image data,
A great circle region identification step for identifying a region including a great circle which is the largest circle among a plurality of circles;
A great circle candidate pixel detecting step of detecting edge pixels in the great circle region and detecting the detected edge pixels as great circle candidate pixels constituting the circumference of the great circle;
A great circle detection step for performing a multi-stage circle approximation on the detected great circle circumference candidate pixels and detecting the center position and radius of the great circle;
A small circle region specifying step of labeling within the great circle so as not to include the great circle edge based on the detected center position and radius, and specifying one or a plurality of regions including one small circle;
A small circle candidate pixel detecting step for detecting edge pixels in the small circle region, and detecting the detected edge pixels as small circle candidate point pixels constituting the circumference of the small circle;
An image processing method comprising: a small circle detecting step of performing multi-stage circle approximation on the detected small circle candidate pixels and detecting a center position and a radius of the small circle. In the great circle region specifying step, the pixels constituting the background and the pixels other than the background are separated by binarization processing,
The image processing method according to claim 1, wherein a great circle region is specified based on a feature amount of a set of pixels constituting a portion other than the background. In the small circle region specifying step, pixels constituting the great circle and pixels constituting other than the great circle are separated by binarization processing,
3. The image processing method according to claim 1, wherein a small circle region is specified based on a feature amount of a set of pixels constituting a portion other than the great circle. 4. The image processing method according to claim 2, wherein the feature amount is the number of pixels included in a set of pixels and a Ferre diameter of the set of pixels. The multi-stage circle approximation includes a process of detecting a center position and a radius based on a circumference candidate pixel, and a distance from the circumference of the circle based on the detected center position and radius among the circumference candidate pixels is a predetermined distance. The image processing method according to any one of claims 1 to 4, wherein a process of setting a closer pixel as a new circumference candidate pixel is made one step and is repeated a plurality of times. The validity of the small circle based on the ratio of the pixels excluding the pixels constituting the great circle and the pixels constituting the background out of all the pixels in the small circle based on the center position and radius detected in the small circle detection step The image processing method according to claim 1, further comprising a verification step of verifying. An image processing program for causing a computer to execute the image processing method according to claim 1. A computer-readable recording medium recording an image processing program for causing a computer to execute the image processing method according to claim 1. An image processing device that detects the center position of each circle with respect to a plurality of circles including a circle having a size different from that of other circles displayed as image data,
A great circle region specifying means for specifying a region including a great circle which is the largest circle among a plurality of circles;
Great circle candidate pixel detecting means for detecting edge pixels in the great circle region and detecting the detected edge pixels as great circle candidate pixels constituting the circumference of the great circle;
A great circle detection means for performing multi-stage circle approximation on the detected great circle circumference candidate pixels, and detecting the center position and radius of the great circle;
A small circle region specifying means for labeling within the great circle so as not to include the great circle edge based on the detected center position and radius, and specifying one or a plurality of regions including one small circle;
Small circle candidate pixel detecting means for detecting edge pixels in the small circle region and detecting the detected edge pixels as small circle candidate point pixels constituting the circumference of the small circle;
An image processing apparatus, comprising: a small circle detecting unit that performs multi-stage circle approximation on the detected small circle candidate pixels and detects a center position and a radius of the small circle. The validity of the small circle based on the ratio of the pixels excluding the pixels constituting the great circle and the pixels constituting the ground out of all the pixels in the small circle based on the center position and radius detected by the small circle detection means The image processing apparatus according to claim 9, further comprising verification means for verifying.

Images