Determination system, determination method, and program

Abstract

To provide a determination system capable of more accurately determining an ineffective hit in table tennis.SOLUTION: A determination system 1 includes a sensor 30 for detecting contact of a ball in table tennis, and a determination part for determining whether a hit ball is an effective hit on the basis of the result of the detection by the sensor 30. The sensor 30 is installed on a side face of a table tennis table 3 with a prescribed interval from an edge where the top face and the side face of the table tennis table 3 contact each other, and the sensor 30 is stuck on the side face in the long side direction of the table tennis table 3 with an interval of 0.8-1.2 mm from the edge.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a determination system, a determination method, and a program.

Background Art

[0002] For example, in Patent Document 1, in a table tennis game progress system installed on a table tennis table 10 having a first area 11 and a second area 12 partitioned around a net part 50, a first detection sensor 41 installed in the first area 11 to detect an impact applied to the first area 11; a second detection sensor 42 installed in the second area 12 of the table tennis table 10 to detect an impact applied to the second area 12; a net sensor 51 to detect an impact applied to the net part 50; a score calculation terminal 100 installed on one side of the table tennis table 10 to receive an impact detection signal transmitted from the sensor and determine whether players on both sides have scored, and display the scores of the players on both sides according to the determination result; is disclosed, which is characterized by being configured to include.

[0003] Further, Patent Document 2 discloses an object position estimation system that parallel - processes images from a plurality of cameras to enable tracking of a specific object in the image. The system includes: an object candidate region extraction means that inputs continuous images from each camera in frame units to generate an inter - frame difference image, and determines an object candidate region in the current frame based on low - level features in a region divided by the difference image; an object candidate region recognition means that, among the object candidate regions obtained from each camera, specifies a region assumed to contain the object to be tracked based on high - level features based on machine learning; an object trajectory generation means that generates a trajectory of a ball candidate region associated between frames for the object candidate region specified by the high - level features obtained from each camera; an object trajectory selection means that selects, as the trajectory of the object to be tracked, the trajectory having the highest value based on a predetermined index for each camera among the trajectories of the object candidate regions obtained from each camera; a projective transformation means that converts and integrates the image coordinates of the trajectory of the object candidate region selected for each camera and the image of the current frame obtained from each camera into a predetermined projective transformation image; a block determination means that divides the projective transformation image into blocks, calculates and compares feature amounts represented by objects other than the object to be tracked that are predetermined to have a high correlation with the object to be tracked within each block, excludes blocks where the feature amount is below a predetermined value, and selects object candidate coordinates located in the non - excluded blocks; and an object position estimation means that determines, as the final ball position coordinates, the ball candidate coordinates that have continued for a predetermined time or more for the selected object candidate coordinates. An object position estimation system characterized by including these components is disclosed.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] An object of the present invention is to provide a determination system that more accurately determines an invalid hit in table tennis.

Means for Solving the Problems

[0006] The determination system according to the present invention includes a sensor that detects contact of a table tennis ball, and a determination unit that determines whether the ball is a valid hit based on the detection result of the sensor. The sensor is installed on a side surface of the table tennis table at a prescribed interval from an edge where the upper surface and the side surface of the table tennis table are in contact.

[0007] Preferably, the sensor is attached to a side surface in the longitudinal direction of the table tennis table at an interval of 0.8 mm or more and 1.2 mm or less from the edge.

[0008] Preferably, the sensor detects vibration caused by contact of a table tennis ball, and the determination unit determines an invalid hit when the vibration detected by the sensor exceeds a threshold value.

[0009] Preferably, the determination system further includes a learning unit that learns the vibration detected by the sensor and the determination corresponding to the vibration as a data set, and the threshold value is adjusted by learning by the learning unit.

[0010] Preferably, the determination system further includes a photographing unit that photographs the upper surface and the side surface of the table tennis table as a photographing range, an image selection unit that selects an image at the time when the vibration determined to exceed the threshold value by the determination unit is detected from the image photographed by the photographing unit, and a contact position specifying unit that specifies a contact position of the ball on the table tennis table in the image selected by the image selection unit. The determination unit determines whether it is a valid hit based on the vibration caused by contact of the table tennis ball and the contact position specified by the contact position specifying unit.

[0011] Preferably, the apparatus further includes a probability calculation unit that calculates the probability of the determination by the determination unit, and a display unit that associates and displays the determination result by the determination unit and the probability calculated by the probability calculation unit. The probability calculation unit calculates the probability of the determination based on the difference between the vibration detected by the sensor and the threshold value, and the contact position specified by the contact position specifying unit.

[0012] The determination method according to the present invention includes a step of a sensor for detecting contact of a table tennis ball detecting contact of the ball, and a step of a computer determining whether the ball is a valid hit based on the detection result of the sensor detected in the detection step. The sensor is installed on a side surface of a table tennis table at a specified interval from an edge where an upper surface and a side surface of the table tennis table are in contact.

[0013] The program according to the present invention includes a determination step of determining whether a ball is a valid hit based on a detection result of contact of the ball by a sensor. The sensor is a sensor for detecting contact of a table tennis ball, and the sensor is installed on a side surface of a table tennis table at a specified interval from an edge where an upper surface and a side surface of the table tennis table are in contact.

Effects of the Invention

[0014] According to the present invention, an invalid hit in table tennis can be determined more accurately.

Brief Description of the Drawings

[0015]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Figure 6

Figure 7

Mode for Carrying Out the Invention

[0016] First, the background of the present invention will be explained. In the rules of table tennis, a ball that hits the corner or edge of the opponent's court on the table tennis table is called an edge ball. An edge ball is a valid hit, and if it cannot be returned, it results in a point. On the other hand, a ball that hits the side of the opponent's court on the table tennis table is called a side and is an invalid hit, resulting in a point loss. So far, the determination of edge balls or sides, which can affect the result depending on the situation of the game, has been made by the umpire's determination or video determination. It is difficult to judge the contact position of a table tennis ball, which can reach a maximum speed of 120 km / h and rotate up to 170 times per second, on the table tennis table only by the umpire's vision. Therefore, in situations where there is confusion in the determination, video determination is adopted, but there are problems such as the need for multiple cameras for one table tennis table, the possibility that the contact position between the ball and the table tennis table may not be shown due to the player blocking the camera, and the time-consuming determination. Therefore, in view of these circumstances, the present invention determines an edge ball or a side by a determination method different from human vision and video determination in situations where determination of an edge ball or a side is required, and outputs the determination result more accurately and quickly, thereby assisting the umpire's determination.

[0017] FIG. 1 is a diagram illustrating a determination system 1. The determination system 1 includes a sensor 30 attached to the table tennis table 3, a photographing device 5, a determination device 7, and a result display device 9. The determination device 7 and the sensor 30, the determination device 7 and the photographing device 5, and the determination device 7 and the result display device 9 are connected to each other via a network. The sensor 30 is a device that is attached to the ping - pong table 3 and detects the contact of the ball with the side surface of the ping - pong table 3. Specifically, the sensor 30 detects the vibration caused by the contact of the ball. The sensor 30 is, for example, a thin piezoelectric film sensor. The photographing device 5 is a camera that photographs the upper surface and the side surfaces of the ping - pong table 3 as the photographing range. Specifically, the photographing device 5 is arranged on the left and right in the longitudinal direction (the long - side direction) of the ping - pong table 3 so as to be able to photograph the side surfaces in the longitudinal direction (the long - side direction) and the upper surface of the ping - pong table 3. The photographed image by the photographing device 5 is managed by the determination device 7. The determination device 7 is a computer terminal that determines the presence or absence of the contact of the ball with the side surface of the ping - pong table 3 based on the vibration detected by the sensor 30. The determination device 5 displays the determination result on the display device 7. The result display device 9 is, for example, a liquid crystal panel, and displays the determination result by the determination device 7.

[0018] FIG. 2(a) is a diagram illustrating the sensor 30 attached to the ping - pong table 3. The sensor 30 is a piezoelectric film sensor, and its size is approximately 5 mm to 15 mm in width, approximately 4 cm to 6 cm in length, and 0.2 mm to 0.5 mm in thickness. The sensor 30 detects the pressing caused by the contact of the ball, generates electric charges due to the piezoelectric effect, the generated electric charges cause a voltage fluctuation, and this voltage fluctuation is transmitted as a voltage signal to the determination device 7 via the sensor cable 32. As illustrated in FIG. 2(b), a plurality of sensors 30 are attached along the long side direction of the side surface of the ping-pong table 1. Specifically, the sensor 30 is attached to the side surface of the ping-pong table 3 with an adhesive or tape that does not inhibit the detection of vibrations caused by the contact of the ball. The sensor 30 is installed on the side surface of the ping-pong table 3 at a prescribed interval from the edge where the upper surface and the side surface of the ping-pong table meet. Preferably, the sensor 30 is attached to the side surface in the long side direction of the ping-pong table 3 at an interval of 0.8 mm or more and 1.2 mm or less from the edge. Further preferably, the sensor 30 is attached at an interval of about 1 mm from the edge of the ping-pong table 3. By attaching it by lowering it by about 1 mm from the edge, vibrations caused by a ball contacting the edge are not directly detected. That is, edge balls are not directly detected. Note that the prescribed interval can be appropriately changed according to the detection accuracy of the vibration of the sensor 30. Also, between the sensors 30, an interval is provided such that the contact of the ball with the side surface of the ping-pong table 3 can be detected and the sensor cable 32 can be installed. About 20 sensors 30 are attached to one side surface in the half surface of the ping-pong table 3 divided by the center line of the table.

[0019] FIG. 3 is a diagram illustrating the hardware configuration of the determination device 7. As illustrated in FIG. 3, the determination device 7 includes a CPU 700, a memory 702, an HDD 704, a network interface 706 (network IF 706), a display device 708, and an input device 710, and these components are connected to each other via a bus 712. The CPU 700 is, for example, a central processing unit. The memory 702 is, for example, a volatile memory and functions as a main memory device. The HDD 704 is, for example, a hard disk drive device and stores computer programs and other data files as a non-volatile recording device. The network IF 706 is an interface for communicating wired or wirelessly, and realizes communication with, for example, the sensor 30, the imaging device 5, and the result display device 9. The display device 708 is, for example, a liquid crystal display. The input device 710 is, for example, a keyboard and a mouse.

[0020] FIG. 4 is a diagram illustrating the functional configuration of the determination device 7. As illustrated in FIG. 4, a determination program 72 is installed in the determination device 7, and a threshold storage database 600 (threshold storage DB 600) and an image data database 610 (image data DB 610) are configured. The determination program 72 includes a vibration data acquisition unit 720, a determination unit 722, a time specifying unit 724, an image selection unit 726, an image data acquisition unit 728, a contact position specifying unit 730, an accuracy calculation unit 732, an output unit 734, and a learning unit 736. Note that part or all of the determination program 72 may be realized by hardware such as an ASIC. Further, the determination program 72 is stored in a recording medium such as a CD-ROM, for example, and is installed in the determination device 7 via this recording medium. The vibration data acquisition unit 720 acquires the vibration data detected by the sensor 30. Specifically, the vibration data includes a vibration value and the time when the vibration is detected, and the vibration data acquisition unit 720 outputs the vibration value acquired from the sensor 30 to the display device 708 as a waveform in association with the detected time.

[0021] The determination unit 722 determines whether the ball is a valid hit based on the detection result of the sensor 30. Specifically, the determination unit 722 makes a determination of an edge ball or a side with respect to the ball hitting the table tennis table 3 based on the detection result of the sensor 30. More specifically, the determination unit 722 determines an invalid hit when the vibration detected by the sensor 30 exceeds a threshold value. Even more specifically, the determination unit 722 compares the detected vibration with the threshold value, determines that it is a side when the vibration value is equal to or greater than the threshold value, and determines that it is an edge ball when the vibration value is less than the threshold value. In addition, the determination unit 722 determines an edge ball or a side based on the vibration detected by the contact of the table tennis ball and the contact position specified by the contact position specifying unit 730.

[0022] The time specifying unit 724 specifies the time when the sensor 30 detects the vibration determined by the threshold determination unit 722 to have exceeded the threshold. Also, it specifies the time when the vibration value designated by the user is detected. The image selection unit 726 selects, from the images taken by the imaging device 5, the image at the time when the vibration determined by the threshold determination unit 722 to have exceeded the threshold is detected. Specifically, the image selection unit 726 selects the image data taken by the imaging device 5 at the time specified by the time specifying unit 724 from the image data DB 610. The image data acquisition unit 728 acquires the images taken by the imaging device 5 and stores them in the image data DB 610. Specifically, the image data acquisition unit 728 stores the acquisition time of the acquired image and the image data in association with each other.

[0023] The contact position specifying unit 730 specifies the contact position of the ball on the table tennis table 3 in the image selected by the image selection unit 726. Specifically, the contact position specifying unit 730 extracts the regions indicating the table tennis table 3 and the ball from the selected image, and specifies the contact position between the table tennis table 3 and the ball. More specifically, the contact position specifying unit 730 specifies whether the contact position is the edge of the table tennis table 3 or the side surface of the table tennis table 3. The contact position specifying unit 730 specifies the contact position based on a learned learning model. The learning model is learned using the image data for specifying the contact position between the table tennis table 3 and the ball as teacher data. The accuracy calculation unit 732 calculates the accuracy of the determination by the determination unit 722. The accuracy becomes lower as the vibration value detected by the sensor 30 is closer to the threshold, and higher as it is farther from the threshold. The accuracy calculation unit 732 calculates the accuracy of the determination based on the difference between the vibration value detected by the sensor 30 and the threshold, and the position specified by the contact position specifying unit 730.

[0024] The output unit 734 displays the determination result by the determination unit 722 on the result display device 9. Specifically, the output unit 734 outputs the possibility of an edge ball and the possibility of a side in percentage to the result display device 9 based on the accuracy calculated by the accuracy calculation unit 732. The learning unit 736 performs machine learning using the detection result of the sensor 30 as input data. Specifically, the learning unit 736 learns the vibration caused by the contact of the ping-pong ball and the determination corresponding to the vibration as a data set. Further, the learning unit 736 adjusts a threshold value that serves as a reference value for determining an invalid hit (side) through learning. Thereby, the accuracy of determining an edge ball or a side is improved. The learning unit 736 stores the adjusted threshold value in the threshold storage DB600.

[0025] FIG. 5(a) is a graph showing the vibration result by the sensor 30. Specifically, FIG. 5(a) shows how the vibration generated by contacting the ping-pong table 3 varies with the passage of time, and is time-series data with the vibration value on the vertical axis and the time on the horizontal axis. As illustrated in FIG. 5(a), when the sensor 30 detects the contact of the ball, it is shown that the voltage signal indicating the vibration appears as a waveform. The vibration value becomes lower as the contact position of the ball moves away from the sensor 30, and the vibration value becomes larger when the ball directly hits the sensor 30. The sensor 30 also detects the vibration caused by the contact with the upper surface of the ping-pong table 3 even if the ball does not directly contact the sensor 30. FIG. 5(b) is a table associating the vibration value detected by the sensor 30 with the determination. As illustrated in FIG. 5(b), when the hit ball is an edge ball, the vibration value detected by the sensor 30 is 0 to 50, and when it is a side, the vibration value is 90 to 100. Note that the vibration value of the ball on the upper surface excluding the edge of the ping-pong table 3 is 10 to 20. In this example, the determination system 1 sets the threshold value to 50, and when the vibration value is closer to the threshold value, in the determination based on the determination result by the vibration value and the determination result by the image data, the image determination result may be weighted more heavily to determine the determination.

[0026] FIG. 6 is a display example of the display device 708 that displays the vibration value and the image data at time t. The vibration value detected by the sensor 30 is associated with the image data captured by the imaging device 5. As illustrated in FIG. 6, in a graph representing the vibration value at a time, a slider is displayed, and the slider specifies the time. The image data at the time specified by the slider is displayed on the display device 708 together with the graph. That is, the time specifying unit 724 specifies the time based on the position of the slider specified by the user, and the image selection unit 726 selects the image data at the time specified by the time specifying unit 724 from the image data DB 610 and outputs it to the display device 708 for display. By displaying the graph of the vibration value and the image data together on the display device 708, it is possible to compare the vibration value necessary for the determination with the image data at the time when the vibration value was detected. Note that the vibration value and the image data may be displayed on the result display device 9.

[0027] FIG. 7 is a flowchart for explaining the determination process (S10) in the determination system 1. As illustrated in FIG. 7, in step 100 (S100), the sensor 30 detects the contact of the ball with the table tennis table 3. The sensor 30 converts the pressure due to the contact into a voltage signal and outputs it to the determination device 7 as vibration data. The vibration data acquisition unit 720 of the determination device 7 acquires the vibration data. The vibration data acquisition unit 720 of the determination device 7 graphs the vibration data and displays it on the display device 708. In step 105 (S105), the determination unit 722 of the determination device 7 compares the threshold value stored in the threshold value storage DB 600 with the vibration value acquired by the vibration data acquisition unit 720. When the vibration value is equal to or greater than the threshold value, the determination unit 722 determines that it is an invalid hit (side) and proceeds to S110. When the vibration value is less than the threshold value, the determination unit 722 determines that it is an edge ball and proceeds to S130. In step 110 (S110), the time specifying unit 724 of the determination device 7 specifies the time when the vibration value exceeding the threshold value was detected.

[0028] In step 115 (S115), the image selection unit 726 of the determination device 7 selects an image captured at the time specified by the time specifying unit 724 from the image data DB610. In step 120 (S120), the contact position specifying unit 730 of the determination device 7 extracts the ball area and the table tennis table 3 area from the image data selected by the image selection unit 726. The contact position specifying unit 726 specifies the contact position with the ball on the extracted table tennis table 3. In step 125 (S125), the determination unit 722 specifies whether the contact position specified by the contact position specifying unit 730 is the edge of the table tennis table 3 or the side surface of the table tennis table 3. In step 130 (S130), the accuracy calculation unit 732 calculates the accuracy of the determination based on the difference between the vibration value and the threshold value and the contact position between the ball and the table tennis table specified by the contact position specifying unit 730. Also, in S105, when it is determined that it is an edge ball, the accuracy calculation unit 732 calculates the accuracy of the determination based on the difference between the vibration value and the threshold value. In step 135 (S135), the output unit 734 of the determination device 7 outputs the determination result to the result display device 9.

[0029] As described above, the determination system 1 detects the vibration of the ball by the sensor 30 attached to the side surface of the table tennis table 3, and compares the vibration value with the threshold value to determine whether the hit ball is an edge ball or a side ball. Since the detection of vibration is performed in real time, the determination result can be obtained more quickly compared to video determination. Also, since the determination system 1 performs determination based on a learning model instead of human vision, the determination accuracy is high. Furthermore, since the determination system 1 performs machine learning, data can be accumulated and the determination accuracy can be improved.

[0030] [Modification Example] In the above embodiment, the sensor 30 is installed on the side surface of the ping-pong table 3, but it is not limited thereto. For example, it may be further installed along the net tape of the net of the ping-pong table 3. When the hit ball contacts the net and then enters the opponent's court during a serve (net-in serve), it is not a valid hit and a re-serve is required. Therefore, in such a situation, by determining the contact with the net not only visually but also by the determination system 1, the accuracy of the determination can be improved.

[0031] Also, in the above embodiment, the sensor 30 is installed on the side surface in the longitudinal direction of the ping-pong table 3, but it is not limited thereto. For example, the sensor 30 may also be installed on the side surface in the short side direction of the ping-pong table 3, and based on the vibration value of the sensor 30, the contact with the edge in the short side direction of the ping-pong table 3 may be determined.

Explanation of Reference Numerals

[0032] 1…Determination system 3…Ping-pong table 5…Photographing device 7…Determination device 9…Result display device 30…Sensor 72…Determination program

Claims

1. A sensor for detecting contact of a table tennis ball, and a determination unit that determines whether the ball is a valid hit based on the detection result of the sensor and having, The sensor is installed on the side surface of the table tennis table with a specified interval from the edge where the upper surface and the side surface of the table tennis table are in contact. Determination system.

2. The sensor is attached to the side surface in the longitudinal direction of the table tennis table with an interval of 0.8 mm or more and 1.2 mm or less from the edge. The determination system according to claim 1.

3. The sensor detects vibrations caused by contact of the table tennis ball, When the vibration detected by the sensor exceeds a threshold value, the determination unit determines it as an invalid hit The determination system according to claim 1.

4. A learning unit that learns the vibration detected by the sensor and the determination corresponding to the vibration as a data set further having, The threshold value is adjusted by learning by the learning unit The determination system according to claim 3.

5. An imaging unit that images the upper surface and side surface of the table tennis table as an imaging range, An image selection unit that selects an image at the time of detecting a vibration determined to have exceeded a threshold value by the determination unit from the images taken by the imaging unit, A contact position specifying unit that specifies the contact position of the ball on the table tennis table in the image selected by the image selection unit further having, Based on the vibration caused by contact of the table tennis ball and the contact position specified by the contact position specifying unit, the determination unit determines whether it is a valid hit The determination system according to claim 3.

6. An accuracy calculation unit that calculates the accuracy of the determination by the determination unit, A display unit that associates and displays the determination result by the determination unit and the accuracy calculated by the accuracy calculation unit further having, The accuracy calculation unit calculates the accuracy of the determination based on the difference between the vibration detected by the sensor and the threshold value, and the contact position specified by the contact position specifying unit The determination system according to claim 5.

7. A sensor for detecting contact of a table tennis ball has a detection step of detecting contact of the ball, A determination step in which a computer determines whether the ball is a valid hit based on the detection result of the sensor detected in the detection step and having, The sensor is installed on the side surface of the table tennis table with a specified interval from the edge where the upper surface and the side surface of the table tennis table are in contact. Determination method.

8. A determination step in which a computer determines whether a ball is a valid hit based on a detection result of contact of the ball by a sensor having wherein the sensor is a sensor that detects contact of a ping-pong ball wherein the sensor is installed on a side surface of a ping-pong table with a prescribed interval from an edge where an upper surface and a side surface of the ping-pong table are in contact Program

Images