WO2018139810A1

WO2018139810A1 - Sensing apparatus for calculating position information of object in motion, and sensing method using same

Info

Publication number: WO2018139810A1
Application number: PCT/KR2018/000897
Authority: WO
Inventors: 박현진
Original assignee: 주식회사 골프존
Priority date: 2017-01-25
Filing date: 2018-01-19
Publication date: 2018-08-02
Also published as: KR20180087748A; TWI647424B; KR101932525B1; TW201827788A

Abstract

The purpose of the present invention is to provide a sensing apparatus and a sensing method which can simply, very accurately, and stably extract an object in motion from images captured while the object, such as a golf club, having a shape, a size, a material, reflectivity, and the like that cannot be specified, is moving. The sensing method, which is performed by the sensing apparatus for calculating position information of an object in motion according to the present invention, comprises the steps of: successively obtaining images at the angle of view at which the object in motion is viewed; generating difference calculation images between each of the successively-obtained images and a reference image, and generating difference calculation images between each two successive difference calculation images; and calculating position information of the object in motion from each of the difference calculation images between each two successive difference calculation images.

Description

Sensing device for calculating position information of a moving object and sensing method using the same

The present invention provides a sensing device that can extract the object from each image to calculate the position information of the object from the image by continuously acquiring and analyzing the image of the object to be exercised, and a sensing method using the same. In the sensing device and sensing method using the same to calculate the sensing data for the golf ball and the motion of the golf club by acquiring and analyzing the image of the golf club and the golf ball when hitting the golf ball by holding a golf swing It is about.

The technology for calculating the location information of the object by analyzing the image of the moving object is mainly used for interactive sports simulation through simulation of popular sports games such as baseball, soccer, basketball, and golf indoors or at specific places. It is used as a sensing device in various simulators and devices for enjoying it in the form of simulation.

If a moving object, for example, when a user swings a golf club with a golf club and hits the golf ball, it is possible to calculate where the moving golf club and the golf ball are located in every frame where the golf ball is taken, the golf club uses the position information calculated every frame. Various kinematic information such as track, speed, direction angle and height angle of the club and golf ball can be calculated and used to calculate the analysis information about the user's golf swing, or in a virtual golf simulation system such as screen golf An image may be used to implement a simulation image of a golf ball flying.

In an image obtained continuously for a certain moving object, the moving object is displayed at different positions every frame while the background part is almost fixed and does not change from frame to frame. If a difference image is generated by difference calculation on the pixel value (brightness value) of each pixel, the background part is removed from the difference image and only the moving object part remains. Difference technique is widely used to extract moving objects from.

FIG. 1 is a diagram for explaining the difference image technique as described above. The difference image Diff (ta, tb) of an image (frame_ta) photographed at time ta and an image (frame_tb) captured at time tb is illustrated in FIG. It is shown.

The difference operation is obtained by the difference in pixel values (brightness values) between pixels at the same position. If the difference between the pixel values of pixels at the same position of two images is greater than 0, the difference value is the pixel value of the corresponding pixel ( Brightness value), and if the difference between the pixel values of the pixels at the same position of the two images is less than zero, the difference image is made by replacing with zero.

FIG. 1 illustrates a case in which a difference operation is performed on 14 x 13 sized images in 14 x-axis directions and 13 x-axis directions.

The purpose of this difference is to extract the moving object (part A) from the frame_ta image. To this end, if a difference operation is performed on all pixels of two images using an image (frame_tb) at time tb, as shown in FIG. 1, a difference image of Diff (ta, tb) is generated.

As shown in FIG. 1, for example, when pixels (1,1) of the frame_ta image and pixels (1,1) of the frame_tb image each have a pixel value of 20, the difference image Diff (ta , tb)) on pixel (1,1) will be black when the pixel value becomes 0.

For example, when the pixel value of pixel (2,7) of the frame_ta image is 50 and the pixel value of pixel (2,7) of the frame_tb image is 20, the difference image Diff (ta, tb) The pixel (2,7) on the) is represented as a pixel having a pixel value of 30 when the pixel value is 50-20 = 30.

For example, when the pixel value of pixel (6,10) of the frame_ta image is 40 and the pixel value of pixel (6,10) of the frame_tb image is 100, according to the difference operation, 40-100 = -60 and the negative value is negative. Since -60 is replaced with 0, the pixel (6, 10) on the difference image Diff (ta, tb) has a pixel value of 0 and is displayed in black.

In this way, if the difference image of the frame_tb image of the frame_ta image is obtained, the object A can be extracted fairly accurately through the difference image if the reflectivity of the moving object due to lighting is always constant and no other variables exist. .

However, because the object is moving, the reflectance of the light may be different every frame, so the pixel values of the pixels that make up the object may be different every frame, and another object is captured in the image. In many cases, it is not always possible to accurately extract the object A through the difference image. Referring to the difference image Diff (ta, tb) of the frame_tb image of the frame_ta image in FIG. 1, it can be seen that the object A is substantially damaged by the difference operation and appears as A ′.

In particular, due to the influence of the object's color or light reflectance, the difference in ambient light, the shade of the light, the material of the object, etc., the object may have a pixel value that is almost similar to or darker than the background part of the image. It is very difficult to extract the golf club from the image by the difference image technique, and even if extracted, the location information calculated therefrom has a problem that the accuracy is inevitably deteriorated.

Unlike golf clubs, the shape and size of golf balls are standardized, and most of them are bright colors. Therefore, the golf balls can be extracted from the image fairly accurately by the difference image technique, even if some pixels of the golf ball portion are lost during the difference image process. The size is standardized so that the contours can be fitted fairly accurately.

However, golf clubs are not only very difficult to extract by the car image technique but also very difficult to extract by the image processing technique used to extract the golf balls as described above.)

In order to solve this problem, a technique of attaching a specific marker to the shaft and head of the golf club and finding the specific marker in the image to identify the position of the golf club has emerged. There is a fatal problem that you must use a specific golf club attached to a specific marker, and even if a specific marker is attached, the marker may not be shown or hidden completely on the image depending on the golf swing. There was a problem that the identification of the location became difficult.

Conventional prior art documents related to the present invention include Patent Application No. 10-2011-0025149, Patent Application No. 10-2011-0111875, Japanese Patent Application Laid-Open No. 2005-210666, and the like.

The present invention is to solve the conventional problems as described above, a simple method of moving the object in the image taken for the movement of the object having an unspecified shape, size, material, reflectivity, such as a golf club The present invention provides a sensing device for calculating position information of a moving object that can be extracted quite accurately and stably, and a sensing method using the same.

According to an embodiment of the present disclosure, a sensing method of a sensing device for calculating position information of a moving object includes: continuously obtaining an image at an angle of view of the moving object; Generating a difference operation image of two successive difference images for each difference operation image of the reference image for each of the successively acquired images; And calculating position information of the moving object from each of the successive computed images of the two successive computed images.

Also preferably, the step of generating the difference operation images of the two consecutive difference images may include determining an image of one frame among the continuously acquired images as the reference image.

Also preferably, the step of generating a difference operation image of each of the two successive difference images may include pixel values that are brighter than the background portion of each image by the difference calculation of the reference image for each of the successively acquired images. Generating a difference operation image each including both pixels having pixels and pixels having a darker pixel value than the background portion, and respectively performing a difference operation on two consecutive difference images of the generated difference images And generating a difference-difference image.

Also preferably, the step of generating the difference operation images of the two successive difference images may include the pixel value of each pixel and the corresponding position of the reference image in each of the successively acquired images. Generating an absolute difference image with an absolute value for the difference in pixel values as the pixel value; and if the difference between the pixel values of each corresponding pixel is greater than zero for the two successive absolute difference images, the value is converted to pixels; And calculating a difference-difference image by performing a difference calculation process of setting the value to a value less than zero to zero as the pixel value.

On the other hand, the sensing method of the sensing device for calculating the position information of the moving object according to another embodiment of the present invention, the method comprising: continuously obtaining an image at an angle of view looking at the moving object; Each successively acquired image includes a background portion, a position moving object appearing at a new position every frame, and a movement position overlapping portion in which a motion change position overlaps in a successive image. Performing image processing to extract the position moving object and the movement position overlapping portion and remove the background portion in each case; Extracting the moving object of the moving object by removing the moving position overlapping portion through the difference operation of two consecutive images with respect to the image processed images; And calculating position information of the extracted exercise object.

Also preferably, the performing of the image processing to remove the background portion comprises: a pixel having a brighter pixel value than the background portion of each image through the absolute value of the difference operation of the reference image for each of the successively acquired images; And image processing to include both pixels having pixel values that are darker than the background portion.

On the other hand, the sensing method of the sensing device for calculating the position information of the golf club exercising in accordance with the golf swing of the user according to an embodiment of the present invention, to continuously acquire an image at an angle of view looking at the golf swing of the user step; Generating an absolute difference image by taking an absolute value of a difference operation of a reference image for each of the successively acquired images; Generating a difference-difference image by performing a difference operation on two consecutive absolute difference images among the generated absolute difference images; And calculating position information of the exercise golf club from each of the generated difference-calculated images.

On the other hand, the sensing device for calculating the position information of the moving object according to an embodiment of the present invention, the camera unit for continuously obtaining an image at an angle of view looking at the moving object; An image processing unit which performs image processing for extracting a difference operation image of two successive difference images for each difference operation image of a reference image for each successively acquired image; And an information calculator configured to calculate positional information of the moving object from each of the successive computed images of the two successive computed images extracted by the image processor.

Also preferably, the image processing unit may determine an image of one frame from among the continuously acquired images as the reference image, and determine an absolute value of a value by the difference operation of the reference image for each of the continuously acquired images. And generate an absolute difference image as a pixel value, and generate a difference-difference image through difference operations of two successive absolute difference images.

On the other hand, the sensing device for calculating the position information of the golf club exercising according to the golf swing of the user according to an embodiment of the present invention, the camera unit for continuously obtaining an image at an angle of view looking at the golf swing of the user; Absolute difference image is generated by taking the absolute value of the difference operation of the reference image with respect to each of the successively acquired images, and performs a difference operation on two consecutive absolute value difference images among the generated absolute value difference images, respectively. An image processing unit which performs image processing to generate a difference-difference image; And an information calculator configured to calculate location information of the golf club to be exercised from each of the generated difference-calculated images.

The sensing device for calculating the position information of the moving object and the sensing method using the same according to the present invention has a non-specific shape, size, material, reflectivity, etc., such as a golf club, without attaching a special marker or additional equipment For each difference image of the reference image for each image acquired in succession with respect to the movement of the object, the moving object is extracted fairly accurately and reliably by generating a difference operation image of two consecutive difference images. There is an effect of calculating accurate position information therefrom.

FIG. 1 is a diagram for explaining a difference image technique, which is generally used in an image processing technology.

2 is a block diagram illustrating a configuration of a sensing device according to an embodiment of the present invention.

3 is a flowchart illustrating a process according to a sensing method of a sensing device according to an embodiment of the present invention.

4 is a diagram illustrating an example of images continuously acquired by a camera unit of a sensing apparatus according to an embodiment of the present invention.

FIG. 5 illustrates a process of generating an absolute difference image using the images illustrated in FIG. 4.

FIG. 6 is a diagram illustrating a process of generating a difference-difference image using the absolute value difference images shown in FIG. 5.

A detailed description of a sensing device for calculating position information of a moving object and a sensing method using the same according to the present invention will be described with reference to the accompanying drawings.

First, a configuration of a sensing device according to an embodiment of the present invention will be described with reference to FIG. 2. 2 is a block diagram illustrating a configuration of a sensing device according to an embodiment of the present invention.

The present invention is completed in the development process of the sensing device for extracting the golf ball and the golf club from the image to calculate the location information by analyzing the images taken when the user hits the golf ball by golf swing to the golf club In particular, the present invention provides a method for accurately and stably extracting an object that is difficult to extract from an image by a conventional image processing technique such as a golf club, and effectively extracts an object that is difficult to extract from an image such as a golf club. Therefore, using the sensing device and the sensing method according to the present invention, as well as a standardized object such as a golf ball, the object according to the motion can be effectively extracted through image processing to accurately calculate its position information. have.

Hereinafter, the golf club will be described mainly as an example of the above-mentioned 'working object', but as described above, the present invention can be equally applied to any object that is difficult to extract through image processing as well as the golf club. Of course.

As shown in FIG. 2, the sensing device according to the exemplary embodiment includes a camera unit 100 and a sensing processor 200, and the sensing processor 200 includes an image processor 210 and an information calculator. And 220.

The camera unit 100 is configured to continuously acquire an image at an angle of view looking at a moving object. In order to calculate position information in a three-dimensional space with respect to the moving object, the camera unit 100 has different viewing angles. For example, as illustrated in FIG. 2, a plurality of cameras acquiring images of the same object, for example, the first camera 110 and the second camera 120 are preferably synchronized with each other and configured in a stereo manner.

As described above, the plurality of

cameras

110 and 120 of the camera unit 100 are synchronized with each other and configured in a stereo manner, thereby obtaining the image and the second camera 120 acquired through the first camera 110 for the same object. The 2D information of the corresponding object extracted from each of the acquired images may be converted into 3D information.

2 illustrates a case where a plurality of

cameras

110 and 120 of the camera unit 100 acquires an image of a user swinging a golf club GC with a user U, but the present invention is limited thereto. For example, the present invention can be used for various sports simulation systems, such as when the user swings a baseball.

The sensing processor 200 collects an image from each of the

cameras

110 and 120 of the camera unit 100 and performs a predetermined image processing to extract a corresponding object, and an object extracted from the image. It may be configured to include an information calculation unit 220 for calculating the three-dimensional position information and the like from the two-dimensional position information of.

The sensing processor 200 extracts a moving object from each of the images collected by the

cameras

110 and 120 of the camera unit 100, calculates position information of the object, and calculates the position information of the corresponding object to the client 300. By transmitting the information, the client 300 may perform a unique function of the client 300 such as calculating new information or calculating analysis information using the position information of the received object.

For example, when the client 300 is implemented as a simulator used for a screen golf system, the location information of a golf ball and a golf club is received from the sensing processing unit 200 and simulation of the trajectory of the ball flying on a virtual golf course using the location information is received. You can implement the image.

In addition, in the case of implementing the client 300 as a golf swing analysis device, receiving the position information of the golf ball and the golf club from the sensing processing unit 200 using the analysis information on the user's golf swing, the diagnosis of the swing and It can be implemented to provide lesson information for solving this.

The image processing unit 210 processes an image to extract a difference operation image of two successive difference images, respectively, for each difference operation image of the reference image for each image continuously acquired by the camera unit 100. The information calculating unit 220 may be configured to calculate position information of the moving object from each of the difference calculated images of the two consecutive difference images extracted by the image processing unit.

As shown in FIG. 2, the information calculating unit 220 may be configured to be included in the sensing device, but may be configured to be included in the client 300 without being limited thereto. That is, the sensing device performs a function of extracting a moving object by acquiring an image and processing the image, and transmits the extracted information to the information calculating unit of the client to obtain the location information of the extracted moving object and various information using the same. It is also possible to make calculations.

Meanwhile, a sensing method of a sensing device according to an embodiment of the present invention will be described with reference to the flowchart shown in FIG. 3.

First, the camera unit shoots a golf swing at a golf club at a predetermined angle of view to acquire an image continuously (S10).

The continuously acquired image is transferred to the image processing unit of the sensing processing unit, and the image processing unit selects the reference image for the difference calculation according to a preset item among the images acquired continuously (S11).

The reference image is an image for removing a background part on an image to be used in a difference image for extracting a moving object (object of interest) to be detected from the image.

The reference image may be selected as an image having only a background portion except for a moving object of interest, and at least a moving object of interest may be selected as an image appearing at a position that does not overlap with an object appearing in another image. It is preferable.

For example, when images of moving objects are acquired continuously at a time of ta <... <t1 <t2 <... <tn, viewpoints such as t1, t2 ..., etc. are extracted to extract an object of interest. If images acquired at are used, an image at time ta or an image at time tn may be selected as a reference image for performing a difference operation on each of the images acquired at times t1, t2...

In this case, images of viewpoints t1, t2, etc., in which the difference operation is to be performed using the reference image will be referred to as 'target images' for convenience.

Whether to select an image acquired at a certain point as a reference image can be determined individually from what the object is being exercised and how it is to be exercised. For example, when a golf club exercising during golf swing is to be extracted from the target images, the target is selected. Since the position where the golf club appears on the images is preferably not overlapped with the position where the golf club appears on the reference image, in the above example, the image of the tn viewpoint, which is a point far from the target images, is selected as the reference image. It can be set in advance in the sensing processing unit.

On the other hand, when the reference image is selected as described above, the object of interest is extracted through the difference operation of the reference image for each target image. However, the difference operation according to the present invention is not a general difference operation described in the background art, but a method using a difference operation of a difference operation in which two difference operations are performed, and more specifically, an absolute value difference operation. 'Difference operation' is again performed on 'absolute difference calculation images' generated by performing 'to generate a difference difference image', thereby providing a method of accurately extracting an object of interest.

This is illustrated in steps S12 and S13 of FIG. 3. As described above, when the image processor of the sensing processor selects the reference image (S11), the image processor determines the reference image of each of the images (target images) that are continuously acquired. An absolute value difference image is generated by using the absolute value of the difference value as the pixel value (S12), and a difference-difference image is generated through the difference operation of two consecutive absolute value difference images (S13).

That is, the absolute value difference operation is an operation that takes an absolute value of a difference between a pixel value (brightness value) of a pixel on a target image and a pixel value (brightness value) of a pixel at a corresponding position on a reference image. An image whose absolute value of the difference in values is a pixel value will be referred to as an "absolute value difference operation image".

The image generated by performing the difference operation again using the absolute difference operation images described above will be referred to as a 'difference difference image'.

As described above, the difference-difference image generated by performing two difference calculations excludes an interfering element, that is, an interference element, from the object of interest, that is, a moving object to be extracted, and the object of interest appears fairly accurately. Therefore, the object of interest may be easily extracted from the difference-difference image.

The portion corresponding to the moving object on each difference-difference image generated as described above is used to determine the contour using edge information, and to find a specific position such as the center point or the center of gravity of the determined contour and then move the position. It can calculate as positional information of (S14).

It is most important to extract the part corresponding to the moving object in the image, and if the part is extracted, calculating the position information in the extracted part may be made by a technique used in the related art.

Hereinafter, referring to FIGS. 4 to 6, a specific example of generation of an absolute value difference image and generation of a difference difference image by the sensing method of the sensing device according to an embodiment of the present invention will be described. Explain.

4 illustrates an example of images continuously acquired by a camera unit of a sensing apparatus according to an embodiment of the present invention, and FIG. 5 illustrates a process of generating an absolute value difference image using the images shown in FIG. 4. FIG. 6 illustrates a process of generating a difference-difference image by using the absolute difference images shown in FIG. 5.

The images shown in FIGS. 4 to 6 are not images acquired by photographing, but are images arbitrarily created to more effectively explain the sensing method of the sensing apparatus according to the present invention. It was planned.

As shown in FIG. 4, the images continuously acquired by the camera unit are images acquired at the times t1, t2, t3 ... tn, respectively. In FIG. 4A, the images acquired at the time t1 (Frame_t1). ), (B) an image (Frame_t2) acquired at time t2, (c) an image (Frame_t3) acquired at time t3, and (d) somewhat far from t1, t2, t3 ... Each of the images Frame_tn acquired at the viewpoint tn viewpoint is shown.

Here, the images (Frame_t1, Frame_t2, Frame_t3, etc.) shown in (a) to (c) of FIG. 4 are a target image, and Frame_tn, which is the image shown in (d), is defined as an image selected as a reference image. do.

As shown in (a) to (d) of FIG. 4, the same background portion BG is present on the image in the images acquired continuously.

The objects of interest (M1, M2, M3 ... Mn) corresponding to the moving object to be detected are present at different positions on each image while moving over the time t1 to tn when the image is acquired (slightly some of them). Overlapping may occur, but it is desirable to exist in different positions every frame, depending on the movement speed of the object and the shooting speed of the camera.

Then, the movement position overlapping parts (V1, V2, V3, ... Vn), which move over the time t1 to tn when the image is acquired, but do not move significantly, shake, or move in place, indicating overlapping movement changes. Suppose that exists in

4 (a) to (d), it can be seen that the movement position overlapping portions V1, V2, V3, ... Vn represent overlapping movement changes.

For example, in the case where the user continuously acquires an image of golf swinging to a golf club, the user's body part, golf mat part, rubber tee part, etc. correspond to the exercise position overlapping part that indicates the overlapping movement change, and the golf Clubs correspond to objects of interest that represent positional changes that do not overlap every frame.

A process of generating an absolute value difference operation image corresponding to step S12 of FIG. 3 using the target images and reference images illustrated in FIGS. 4A to 4D is illustrated in FIG. 5.

FIG. 5A illustrates a case where an absolute value difference image AbsDiff (t1, tn) is generated by the absolute value difference operation on the reference image Frame_tn of the target image Frame_t1 at the time point t1. (b) illustrates a case in which the absolute value difference image AbsDiff (t2, tn) is generated by the absolute value difference operation on the reference image Frame_tn of the target image Frame_t2 at the time t2, and FIG. 5 (c) Denotes a case where the absolute value difference image AbsDiff (t3, tn) is generated by the absolute value difference operation on the reference image Frame_tn of the target image Frame_t3 at the time t3.

As described above, the absolute difference calculation takes an absolute value for the difference between the pixel value (brightness value) of the target image and the reference image, so the higher the pixel value, the higher the pixel value (brightness value) in the absolute difference calculation image. ).

Accordingly, the background portion BG of which the pixel values of the target image and the reference image are substantially the same as the pixel values can all be removed as shown in FIGS. Objects of interest (M1, M2, M3), motion position overlaps (V1, V2, V3, Vn), and unnecessary position shift objects (Mn) on the reference image appear fairly intact on the absolute difference image.

In the example shown in FIG. 5A, an absolute value difference image AbsDiff (t1, tn) is generated by an absolute value difference operation on the reference image Frame_tn of the target image Frame_t1, and the absolute value difference image AbsDiff (t1, tn) is generated. ), The background portion BG is mostly erased and the object of interest M1 and the unnecessary position movement object Mn, which exist at different positions, appear almost intact by the absolute value of the difference operation, and the movement position overlapping portions VO1 (V1 and Vn). A substantial part of this overlapping part) also exists.

In this case, the moving position overlapping part VO1 may be erased by the absolute value difference operation when the pixel value is the same in the part where V1 and Vn overlap each other. That is, although the drawings are shown as appearing intact, some of them appear to be erased.

The object of interest M2 and the unnecessary position shift object Mn in the generation of the absolute value difference image AbsDiff (t2, tn) by the absolute value difference operation on the reference image Frame_t2 of the target image Frame_t2 shown in FIG. And in the generation of the absolute difference image AbsDiff (t3, tn) by the matter relating to the movement position overlapping part VO2 and the absolute difference operation on the reference image Frame_tn of the target image Frame_t3 shown in FIG. Matters relating to the target object M3, the unnecessary position moving object Mn, and the movement position overlapping portion VO3 are also the same as those in FIG.

Therefore, unlike the general difference calculation, the absolute difference calculation as described above, whether the portion having the high pixel value or the low pixel value of the pixels constituting the object of interest, appears completely in the absolute difference calculation image. Not only that, but the low pixel values can have a high pixel value so that they can be significantly distinguished from the surroundings (i.e., the dark parts of the target image appear bright in the absolute difference image).

Here, as the background part is erased by the absolute difference operation, most of the remaining parts except the object of interest (M1, M2, M3), unnecessary moving object (Mn), and the movement position overlapping part (VO1, VO2, VO3) It has a pixel value of 0 and appears black, but for convenience of explanation and understanding, all parts having a pixel value of 0 are shown in white.

AbsDiff (t1, tn), AbsDiff (t2, tn), AbsDiff (t3, tn), and the like, which are the absolute value difference images respectively generated in FIGS. An image is generated, which will be described with reference to FIG. 6.

As described with reference to FIG. 5, an absolute value difference operation generates absolute difference operations images such as AbsDiff (t1, tn), AbsDiff (t2, tn), and AbsDiff (t3, tn). By performing a difference operation (where a difference operation means a conventional difference operation), a difference difference image can be obtained as shown in FIG. 6.

In the AbsDiff (t1, tn), AbsDiff (t2, tn), AbsDiff (t3, tn), etc., which are absolute difference calculation images shown in FIG. 5, the objects of interest are M1, M2, M3, and the movement position overlapping part (VO1). , VO2, VO3) and the unnecessary displacement object Mn are all parts to be removed as an interference element.

Such interferences can be eliminated by the difference operation of two consecutive absolute value difference images. As shown in FIG. 6 (a), AbsDiff (t2, tn) of the absolute value difference image AbsDiff (t1, tn) is shown. Through the difference operation on, the moving position overlapping portions VO1 and VO2 which are almost identical to the moving object Mn, which are the same portions on the two images, are almost eliminated, and the positions are not overlapped. The object of interest M1 is preserved, and the portion of M2 that has a negative value by the difference operation disappears on the difference-difference image (DOD (t1, t2 | tn)) by replacing the pixel value with zero.

Similarly, as shown in (b) of FIG. 6, the position shift object, which is a part that exists equally on both images through the difference operation on AbsDiff (t3, tn) of the absolute value difference image AbsDiff (t2, tn), The movement position overlapping portions VO2 and VO3, which are almost the same as Mn), are almost all erased, the object of interest M2 whose position does not overlap is preserved, and the M3 portion having a negative value by the difference operation. The pixel value is replaced by 0 and disappears on the difference-difference image (DOD (t2, t3 | tn)).

Here, the motion position overlapping portions VO1 and VO2, VO2 and VO3, which are not completely overlapped by the difference operations of two successive absolute difference images, the difference-difference images DOD (t1, t2 | tn) and DOD (t2, A small portion of each of t3 | tn) may remain and become noise (NZ).

Here, by the difference-difference operation, the movement position overlapping portions VO1 and VO2 or VO2 and VO3 are not removed, leaving only the portions exactly indicated by NZ.

For example, the motion position overlap portion VO1 on the absolute difference image AbsDiff (t1, tn) and the motion position overlap portion VO2 on the absolute difference image AbsDiff (t2, tn) are not completely identical except for the NZ portion, Since there may be portions in which the pixel values of pixels at positions corresponding to each other in the position overlapping portions are different from each other, in reality, there are a number of large and unidentifiable large portions other than the NZ portions shown on each difference-difference image shown in FIG. You can leave small noises.

That is, in the present specification, for the sake of simplicity, only NZ portion is shown as noise in FIG. 6, but in fact, other noises may exist in addition to the NZ portion, and such noise (NZ portion and other noise portions) is generally used. It can be removed by applying the noise reduction technique used.

Therefore, if the difference-difference image is obtained through the difference operation of two consecutive absolute difference images as described above, the object of interest appears on the difference-difference image fairly intactly, and thus the object of interest is processed through appropriate image processing. You will be able to specify the object correctly.

As described above, after extracting the object of interest, i.e., the moving object, from the difference-difference image, the contour is specified using the edge information of the corresponding part, and the center point or the center of gravity of the part specified by the contour. The feature point of can be specified and computed as positional information.

Therefore, even an object having a shape, size, material, reflectivity, etc. which cannot be specified, such as a golf club, can be extracted quite accurately by the generation of the absolute difference image and the difference difference image according to the present invention as described above. Therefore, the positional information of the moving object can be calculated accurately, stably and in a simple manner.

Sensing device for calculating the position information of a moving object and a sensing method using the same according to the present invention, by allowing the golf simulation based on the industrial field and virtual reality related to the golf practice, such as hit analysis according to the golf swing It can be used in the so-called screen golf industry, etc., where the user can enjoy a virtual golf game.

Claims

A sensing method of a sensing device for calculating position information of a moving object,

Acquiring an image continuously at an angle of view looking at the moving object;

Generating a difference operation image of two successive difference images for each difference operation image of the reference image for each of the successively acquired images; And

Calculating position information of the moving object from each of the successive computed images of the two successive computed images;

Sensing method of the sensing device for calculating the position information of the moving object comprising a.
The method of claim 1, wherein the generating of the difference operation images of the two consecutive difference images, respectively,

And determining an image of one frame among the continuously acquired images as the reference image.
The method of claim 1, wherein the generating of the difference operation images of the two consecutive difference images, respectively,

A difference operation image including both pixels having a lighter pixel value than the background part on each image and pixels having a darker pixel value than the background part by the difference calculation of the reference image for each of the successively acquired images Generating each of the

Sensing apparatus for calculating the position information of the moving object comprising the step of generating a difference-difference image by performing a difference operation on each of two consecutive difference image of the generated difference image Sensing method.
The method of claim 1, wherein the generating of the difference operation images of the two consecutive difference images, respectively,

Generating an absolute value difference operation image having an absolute value for the difference between the pixel value of each pixel and the pixel value of each pixel at a corresponding position of the reference image in each of the successively acquired images;

Difference-difference processing is performed for two consecutive absolute value difference images, where the difference between the pixel values of each corresponding pixel is greater than zero, and the value is a pixel value; Sensing method of the sensing device for calculating the position information of the moving object comprising the step of generating an image.
A sensing method of a sensing device for calculating position information of a moving object,

Acquiring an image continuously at an angle of view looking at the moving object;

Each successively acquired image includes a background portion, a position moving object appearing at a new position every frame, and a movement position overlapping portion in which a motion change position overlaps in a successive image. Performing image processing to extract the position moving object and the movement position overlapping portion and remove the background portion in each case;

Extracting the moving object of the moving object by removing the moving position overlapping portion through the difference operation of two consecutive images with respect to the image processed images; And

Calculating position information of the extracted exercise object;

Sensing method of the sensing device for calculating the position information of the moving object comprising a.
The method of claim 5, wherein performing image processing to remove the background portion comprises:

Image processing to include both pixels having a lighter pixel value than the background portion of each image and pixels having a darker pixel value than the background portion through the absolute value of the difference operation of the reference image for each of the successively acquired images Sensing method of the sensing device for calculating the position information of the moving object, characterized in that it comprises a step of doing.
A sensing method of the sensing device for calculating the position information of the golf club exercising according to the user's golf swing,

Acquiring an image continuously at an angle of view looking at the golf swing of the user;

Generating an absolute difference image by taking an absolute value of a difference operation of a reference image for each of the successively acquired images;

Generating a difference-difference image by performing a difference operation on two consecutive absolute difference images among the generated absolute difference images; And

Calculating position information of the exercise golf club from each of the generated difference-calculated images;

Sensing method of the sensing device for calculating the position information of the golf club exercising in accordance with the user's golf swing.
A sensing device for calculating position information of a moving object,

A camera unit which continuously acquires an image at an angle of view looking at the moving object;

An image processing unit which performs image processing for extracting a difference operation image of two successive difference images for each difference operation image of a reference image for each successively acquired image; And

An information calculation unit for calculating position information of the moving object from each of the difference operation images of the two successive difference operations images extracted by the image processing unit;

Sensing device for calculating the position information of the moving object including a.
The method of claim 9, wherein the image processing unit,

An image of one frame among the images acquired continuously is determined as the reference image, and an absolute value difference calculation image is generated in which an absolute value of a value obtained by the difference operation of the reference image with respect to each of the images acquired continuously is a pixel value. And generate a difference-difference image through difference operations of two successive absolute difference images. 2.
A sensing device for calculating the position information of the golf club exercising according to the user's golf swing,

A camera unit which continuously acquires an image at an angle of view looking at the golf swing of the user;

Absolute difference image is generated by taking the absolute value of the difference operation of the reference image with respect to each of the successively acquired images, and performs a difference operation on two consecutive absolute value difference images among the generated absolute value difference images, respectively. An image processing unit which performs image processing to generate a difference-difference image; And

An information calculation unit for calculating position information of the golf club to be exercised from each of the generated difference-calculated images;

Sensing device for calculating the position information of the golf club to exercise in accordance with the user's golf swing.