CN103456000B - Feature point matching method and device - Google Patents

Abstract

A feature point matching method, which includes: obtaining a feature point from an image of at least one object; obtaining the position Pn of a feature point; obtaining the position Pn' of the feature point corresponding to Pn; determining whether the threshold is greater than the offset; if If yes, the feature points are obtained.

Description

Feature point matching method and device thereof

technical field

The invention relates to a feature point matching method and its device, in particular to a method and its device applied to image detection.

Background technique

With the development of automation, in recent years, the industry has used robotic arms to achieve the purpose of factory automation, but most of the applications are still limited to repetitive processing of fixed tasks, and these tasks must be planned in advance and then operated by an experienced operator A teaching device to teach the robot arm to reciprocate along the fixed track.

However, in the electronic industry with miniaturization and various types of low-volume production, the robotic arms used in this industry will require more time for teaching, and require a lot of fixtures for various objects, which will virtually increase the cost. .

Therefore, in recent years, robotic handsets used in automated production will be combined with camera devices, such as fixed single CCD (Charge-coupled Device), Eye-in-Hand CCD and stereo vision module. Posture recognition ability and grasping ability to shorten the line change time and reduce the production cost of fixtures.

However, in the vision-guided robotic arm (VisionGuided Robot, VGR) system, all guiding tasks must be based on the feature points of the image of the object for judgment and position estimation, so the technology of obtaining feature points through the camera device is possible. room for discussion.

Contents of the invention

In one embodiment, the present disclosure provides a feature point matching method, which includes:

obtaining a feature point in an image of at least one object;

Obtain the position Pn of a feature point;

Obtain the position Pn' of the feature point corresponding to Pn;

Determine whether the threshold is greater than the offset; if yes, obtain the feature points.

In one embodiment, the present disclosure provides a feature point matching device, which includes:

a receiving unit having at least one light source;

an image capture device, which is arranged above the receiving unit; and

A central processing unit is electrically connected to the receiving unit and the image capture device respectively.

Description of drawings

FIG. 1 is a schematic diagram of an embodiment of a feature point matching device of the present disclosure.

FIG. 2 is a schematic diagram of another embodiment of the feature point matching device of the present disclosure.

FIG. 3 is a schematic flowchart of a feature point matching method of the present disclosure.

[Description of main component symbols]

10 undertaking unit

100 light sources

101 bearing plate

20 image capture device

200 image capture device

201 First Mobile Unit

202 Second imaging device

203 Second Mobile Unit

30 central processing unit

300 output ports

201A First Mobile Unit

203A Second Mobile Unit

S1～S9 steps

detailed description

The implementation manners of the present disclosure are described below through specific specific examples, and those skilled in the art can easily understand the present disclosure from the content disclosed in this specification.

Please refer to FIG. 1 , an embodiment of a feature point matching device of the present disclosure, which has a receiving unit 10 , an image capturing device 20 and a central processing unit 30 .

The receiving unit 10 has at least one light source 100 and a support plate 101, the light source 100 is arranged on the outer edge of the support plate 101, the support plate 101 can move in an axial direction relative to the light source 100, for example, the axial movement is the longitudinal axis Alternatively, the light source 100 can be a light emitting diode.

The image capturing device 20 has at least one image capturing device 200 , at least one first moving unit 201 , at least one second image capturing device 202 and at least one second moving unit 203 .

The first mobile unit 201 is disposed above the receiving unit 10 .

The first imaging device 200 is arranged on the first mobile unit 201, and the first imaging device 200 has a first viewing angle relative to the receiving unit 10. If the first imaging device 200 is movable, the first imaging device 200 is A plurality of feature point positions P _n are captured at different positions on an object 40 (as described later), assuming that the first imaging device 200 is immovable, and the support plate 101 moves in an axial direction, the first imaging device 200 is A plurality of feature point positions P _n are extracted from different positions on the object 40 .

The second moving unit 203 is arranged above the receiving unit 10. For example, the second moving unit 203 and the first moving unit 201 can be a combination of a linear slide rail and a slider, or a combination of a ball screw and a support base. combination.

The second imaging device 202 is arranged on the second mobile unit 203, and the second imaging device 202 has a first viewing angle relative to the receiving unit 10. For example, the second viewing angle and the first viewing angle can be 0 to 180 degrees, such as 15, 25, 30, 35, 40, 45, 50, 60, 70, 80, 90, 95, 100, 110, 120, 130, 140, 150, 160, 170 or 175 degrees, the second imaging device 202 and The first imaging device 201 can be a CCD (Charge-coupled Device) or a CMOS (Complementary Metal Oxide Semiconductor).

If the second imaging device 202 is movable, the second imaging device 202 captures a plurality of feature point positions _Pn (as described later) on an object 40 at different positions, assuming that the second imaging device 202 is It cannot be moved, the support plate 101 moves in an axial direction, and the second imaging device 202 captures a plurality of feature point positions P _n ′ on the object 40 at different positions.

The central processing unit 30 is electrically connected to the image capture device 20 and the receiving unit 10 respectively. The central processing unit 30 has at least one output port 300. For example, the output port 300 can be USB (Universal Serial Bus), IEEE1394a or IEEE1394b.

In addition, the central processing unit 30 controls the operations of the first moving unit 201 , the second moving unit 203 and the tray 101 , and receives images of the object 40 captured by the second imaging device 202 and the first imaging device 201 .

Please refer to FIG. 2, another embodiment of the feature point matching device of the present disclosure, in FIG. 1, the central processing unit 30 and the output port 300 are used in FIG. 1 , so the component symbols are used in the same way.

In this embodiment, the first moving unit 201A and the second moving unit 203A of the image capture device 20A are respectively changed into a robotic arm, and the first image capturing device 200 and the second image capturing device 202 are arranged at the free ends of the robotic arm .

The first moving unit 201A and the second moving unit 203A enable the first imaging device 200 and the second imaging device 202 to move longitudinally or laterally, respectively.

Please refer to Figure 3, the present disclosure is a feature point matching method, which includes:

S1: Obtain feature points.

A feature point is acquired in an image of at least one object 40, for example, the object 40 can be a thin object.

Please refer to FIG. 1 and FIG. 2 again, at least one object 40 is located on the support plate 101, and the light source 100 projects a bright light onto the surface of the object 40 in a horizontal direction, so that the surface of the object 40 forms a plurality of continuous bright spots. With a plurality of consecutive dark points, bright points or dark points can be regarded as feature points, and the bearing plate 101 is moved axially in time to highlight the feature points. For example, the feature points are obtained by using a corner detection (ConrnerDetection) or a Manually designate to obtain the matching position of an image point in the outline of the object 40. The image point is a feature point. If it is described by corner detection, the surface of the object 40 has bright points and multiple continuous dark points. These bright points or The dark point can be regarded as a feature point. If it is further explained, the center of the object 40 is darker than the bright point, so the center can be regarded as a corner point, and the bright point can be regarded as a feature point.

S2: Set a matching size.

The matching size is the number of pixels (pixle), which is the image of the object 40 captured by the first image capturing device 200 and the second image capturing device 202 described later.

S3: Obtain the position P _n of a feature point.

The first imaging device 200 obtains the position P _n of a feature point at the first viewing angle, where P _n is the position of the feature point of the image of the object 40, and n is a constant, such as 0, 1, 2, 3, 4, 5, 6 to n.

S4: Obtain the position P _n ′ of the feature point corresponding to P _n .

The second image capturing device 202 obtains the position P _n ′ of the feature point at the second viewing angle, where P _n ′ is the corresponding position of P _n in the image of the first viewing angle.

S5: Obtain the position P _n ″ of the feature point corresponding to P _n ′.

The first imaging device 200 searches the image in the reverse direction relative to P _n ′, and obtains the position P _n ″ of the feature point at the first viewing angle, where P _n ″ is the corresponding position of the image of P _n ′ at the first viewing angle.

S6: Judging whether the threshold is greater than the offset (ΔX _n > X _n ), the offset is X _n , X _n =P _n ″-P _n ', the threshold is ΔX _n , ΔX _n =X _n-1 -X _n , and the above-mentioned search and judgment of the feature point positions are processed by the central processing unit 30, or through the output of the output port 300, so that another computing unit performs calculations to obtain the feature point positions and the aforementioned judgments.

S7: If not, perform size reduction matching, that is, reduce pixels, and then perform S3 after the pixels are reduced.

S8: If yes, obtain the feature points, and then perform a stereoscopic distance measurement in S9, the first imaging device 201 and the second imaging device 202 capture the above-mentioned feature points, and capture the first imaging device 201 The obtained feature points and the coordinates of the feature points captured by the second imaging device 202 are substituted into a first beam intersection collinear function and a second beam intersection collinear function to calculate the three-dimensional space coordinates of the object 40, A detailed description thereof can be found in Taiwan Patent Application No. 099107942.

To sum up the above, the present disclosure is aimed at an image to obtain its feature point position, and then performs matching of two feature points in the image for the feature point position. After the image matching is successful, use the result to perform a reverse matching on the image to obtain two matching The offset of the result, the offset is a position offset, if the offset is less than the threshold, the searched size is a better size, so the feature matching accuracy can be obtained to overcome the image feature deformation caused by the difference in viewing angle.

However, the specific embodiments described above are only used to illustrate the present disclosure, rather than to limit the applicable scope of the present disclosure. Completed equivalent changes and modifications should still be covered by the protection scope of the appended claims.

Claims (17)

1. a characteristic point matching method, it includes:

A unique point is obtained at the image of at least one object;

Obtain the position P of a unique point _n;

Obtain corresponding P _nthe position P of unique point _n';

Judge whether that threshold value is greater than side-play amount; If so, then unique point is obtained,

Wherein one first image-taking device obtains the position P of a unique point in one first visual angle _n, P _nbe the position of the unique point of an object image, one second image-taking device obtains the position P of unique point in one second visual angle _n', P _n' be P _nin the correspondence position of the image at the first visual angle,

The wherein relative P of this first image-taking device _n' in oppositely searching image, and the position P of unique point is obtained in this first visual angle _n", P _n" be P _n' correspondence position of image in this first visual angle, wherein this side-play amount is X _n, X _n=P _n"-P _n', this threshold value is Δ X _n, Δ X _n=X _n-1-X _n.

2. characteristic point matching method as claimed in claim 1, wherein obtaining of this unique point utilizes one jiao to detect or manual type appointment.

3. characteristic point matching method as claimed in claim 1, it has setting one matching size further.

4. characteristic point matching method as claimed in claim 3, wherein this matching size is picture number.

5. in this, characteristic point matching method as claimed in claim 3, wherein judges whether that threshold value is greater than in side-play amount, if not, then carries out reducing this matching size, then carry out the position Pn obtaining a unique point.

6. characteristic point matching method as claimed in claim 1, if wherein obtain unique point, then carries out a stereoscopy passive ranging.

7. a Feature Points Matching device, it includes:

One accommodation unit, it has at least one light source;

One image capturing device, it is located at the top of this accommodation unit; And

One CPU (central processing unit), it is electrically connected this accommodation unit and this image capturing device respectively,

Wherein this image capturing device has at least one first image-taking device, and this first image-taking device has one first visual angle relative to this accommodation unit,

Wherein this image capturing device has at least one second image-taking device, and this second image-taking device has one second visual angle relative to this accommodation unit,

Wherein this first image-taking device obtains the position P of a unique point in this first visual angle at the image of at least one object _n, P _nfor the position of the unique point of this object image, this second image-taking device obtains the position P of unique point in this second visual angle at the image of this at least one object _n', P _n' be P _nin the correspondence position of the image at the first visual angle,

The wherein relative P of this first image-taking device _n' in oppositely searching image, and the position P of unique point is obtained in this first visual angle _n", P _n" be P _n' correspondence position of image in this first visual angle, wherein this side-play amount is X _n, X _n=P _n"-P _n', this threshold value is Δ X _n, Δ X _n=X _n-1-X _n.

8. Feature Points Matching device as claimed in claim 7, wherein this accommodation unit has a tray further, and this light source is located at the outer rim of this tray.

9. Feature Points Matching device as claimed in claim 8, wherein this tray can be one to move axially relative to this light source.

10. Feature Points Matching device as claimed in claim 9, wherein this light source is light emitting diode.

11. Feature Points Matching devices as claimed in claim 7, wherein this image capturing device has at least one first mobile unit, and this first image-taking device is located at this first mobile unit.

12. Feature Points Matching devices as claimed in claim 11, wherein this first mobile unit can be the combination of a linear slide rail and a slide block, the combination of a ball screw and a supporting seat or a mechanical arm.

13. Feature Points Matching devices as claimed in claim 7, wherein this first image-taking device and this second image-taking device can be an a CCD or CMOS.

14. Feature Points Matching devices as claimed in claim 13, wherein this image capturing device has at least one second mobile unit, and this second image-taking device is located at this second mobile unit.

15. Feature Points Matching devices as claimed in claim 14, wherein this second mobile unit can be the combination of a linear slide rail and a slide block, the combination of a ball screw and a supporting seat or a mechanical arm.

16. Feature Points Matching devices as claimed in claim 7, wherein this CPU (central processing unit) has at least one output port.

17. Feature Points Matching devices as claimed in claim 16, wherein this output port can be USB, IEEE1394a or IEEE1394b.