JP4234407B2 - Stroke information measuring device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a stroke information measuring apparatus for measuring a time required for each stroke of a swimmer, a distance traveled by each stroke, and the like by imaging a swimmer during swimming.
[0002]
[Prior art]
In swimming, the pace distribution of swimmers is one of the important factors for improving competitiveness. Therefore, it is required to measure stroke information such as the time required for each stroke of the swimmer (hereinafter referred to as “stroke time”) and the distance traveled by each stroke (hereinafter referred to as “stroke distance”).
[0003]
Traditionally, measurement of such stroke information has been done manually, such as taking a picture of a swimmer with a video and using a stopwatch while watching the video. .
[0004]
As a technique for measuring the stroke information, the techniques of Patent Document 1 and Patent Document 2 are known. Patent Document 1 describes that a motion of a swimmer during swimming is captured by motion capture. Further, Patent Document 2 describes that measurement is performed by attaching measurement targets to specific parts of a swimmer and detecting these images with a camera.
[0005]
[Patent Document 1]
Japanese Patent Laid-Open No. 11-159173 (paragraphs 0014 to 0018, FIG. 2)
[Patent Document 2]
Japanese Patent Laid-Open No. 2002-277481 (paragraphs 0035 to 0049, FIGS. 2 to 6)
[0006]
[Problems to be solved by the invention]
However, in a swimmer who is swimming, the part exposed on the water surface changes depending on the movement of the swimmer, and the part also differs depending on the swimming method. For this reason, it is difficult to continuously track a predetermined target site using either a camera installed on the water surface or an underwater camera. Furthermore, tracking of the target site is made more difficult due to the influence of water splashes and waves.
[0007]
Then, this invention makes it a subject to provide the stroke information measuring device which made it possible to measure stroke information with high precision, without tracking a swimmer's specific site | part.
[0008]
In order to solve the above problems, a stroke information measuring device according to the present invention is a stroke information measuring device that images a swimmer during swimming with an imaging means and measures stroke information of the swimmer based on the acquired image data. Pixel extraction means for extracting pixel data at a position corresponding to a predetermined measurement line in each acquired image data, integration means for binarizing the pixel values of the extracted pixels and integrating along the measurement line, and storage means based on the accumulated data outputted from the one in which is provided and an output transit time determining means determines that the measurement line passing time of the swimmer the rise time of the peak at time series change.
[0009]
By extracting and binarizing the pixel data on the set measurement line, the pixel data corresponding to the body image of the swimmer is extracted from the course background data (pool floor, water surface, course rope, etc.). It can be accurately extracted regardless of swimming style and posture. Then, by accumulating the binarized data in this way, it becomes possible to determine with high accuracy when the swimmer passes the predetermined measurement line on the course. According to this apparatus, since it is not necessary to attach a target to the swimmer, it is possible to perform measurement that is simple and independent of the swimming method, posture, and the like. In addition, since it does not interfere with swimming, it can be measured during competitions and during regular practice, which is useful for training swimmers.
[0010]
It is preferable that a plurality of the imaging means are arranged along the course of the swimmer and a plurality of predetermined measurement lines are arranged along the course. By arranging one or more measurement lines for each imaging means, it becomes possible to easily measure the lap time for each section.
[0011]
It is preferable that the apparatus further includes a stroke information calculation unit that calculates information regarding the stroke including the stroke cycle based on the measurement line passing time data of the swimmer measured for each measurement line and the position data of the measurement line. The measurement line passage time data is the data that indicates when a part of the swimmer's body has passed the measurement line, and the position of the body that first passes through the measurement line depends on the posture of the swimmer when passing through the measurement line. Come different. And because parts such as hands and feet are periodically moved relative to the center of the body such as the head and torso, the position of the swimmer's body must move periodically according to the stroke of the swimming method. become. However, within a distance of several meters shorter than one stroke, the central part of the body such as the head and the torso moves at a substantially constant speed. Moreover, since the head is located at the forefront of the body of a swimmer who moves periodically in any swimming method, it is possible to estimate the movement of the body center from the measurement line passing data. From the movement of the body center and the periodicity in the measurement line passage time data, stroke information such as the movement distance, movement speed, average speed, and stroke time for each stroke can be calculated.
[0012]
DETAILED DESCRIPTION OF THE INVENTION
DESCRIPTION OF EXEMPLARY EMBODIMENTS Hereinafter, preferred embodiments of the invention will be described in detail with reference to the accompanying drawings. In order to facilitate the understanding of the description, the same reference numerals are given to the same components in the drawings as much as possible, and duplicate descriptions are omitted.
[0013]
FIG. 1 is a schematic configuration diagram showing a configuration of a stroke information measuring apparatus according to the present invention, and FIGS. 2 to 4 are detailed partial configuration diagrams thereof. The stroke information measuring apparatus 1 captures an image of the course 120 divided by the course rope 110 in the pool 100 by a plurality of video cameras (imaging means) 2 that can output a swimmer 150 that is swimming, and that captures image data. Are processed by the arithmetic unit 3 provided for each video camera 2, the processing data is sent to the server unit 5 connected by the network unit 4 such as a LAN, and the stroke information is obtained. To the display processing unit 6 connected to, and displays desired information on the display monitor 7.
[0014]
A plurality of video cameras 2 are arranged along the course 120 direction of the pool 100. For example, the video cameras 2 are installed along a ceiling, a wall surface, or a course side overlooking the pool 100. Each video camera 2 captures all the courses 120 provided in 8 courses at the same time, and by overlapping a part of the imaging range of the adjacent video camera 2, it is possible to capture from one end of the course 120 to the other end. It is said. The video camera 2 has a function of outputting an image signal obtained by A / D converting a NTSC (National Television System Committee) composite video signal. In this image signal, for example, each frame size is 640 × 480, and each pixel information is RGB 24 bits. Note that the video camera 2 itself outputs an analog video signal, and the video signal is converted to A by each arithmetic unit 3 or by an A / D converter connected between each arithmetic unit 3 and the video camera 2. A / D conversion may be adopted.
[0015]
Each calculation unit 3, server unit 5, and display processing unit 6 are each composed of a CPU, a memory, a storage unit such as a hard disk, a communication unit, and an input / output unit. A system based on a computer, EWS, or the like may be used. Each functional unit, which will be described later, includes a case of being divided by software as well as a case of being divided by hardware.
[0016]
As shown in FIG. 2, each arithmetic unit 3 includes an image memory unit 30 for temporarily storing image data transferred from the video camera 2, an area setting unit 31 for setting a measurement range, and a measurement range. A pixel extracting unit (pixel extracting unit) 32 for extracting pixel data on the measurement line, an integrating unit (integrating unit) 33 for integrating the extracted pixels, and time series data for integrating the data in time series The conversion unit 34, the passage time determination unit (passage time determination means) 35 that determines the passage time of the swimmer from the data arranged in time series, the transmission / reception unit 36 that communicates with the network unit 4, and the overall control. The control part 37 to perform is comprised. As shown in FIG. 1, the calculation unit 2 may be provided one by one corresponding to each video camera 2, or a plurality of measurement lines (measurement positions) described later are arranged per one video camera 2. In order to facilitate high-speed processing, a plurality of calculation units 3 may be arranged for each video camera 2. On the contrary, when the arithmetic processing performance of the arithmetic unit 3 is high, the image data sent from the plurality of video cameras 2 may be processed by one arithmetic unit 3. When the video camera 2 and the calculation unit 3 do not correspond one-to-one, the video camera 2 may be connected to the network unit 4.
[0017]
The network unit 4 is, for example, a line network connected by a wired line such as an electric wire or an optical cable, a wireless line such as infrared ray, radio wave, or light, or a combination thereof, TCP / IP (Transmission Control Protocol / Internet Protocol), It is controlled by various protocols such as Netware (trademark of Novell) and NETBEUI (NetBIOS Extended User Interface).
[0018]
As shown in FIG. 3, the server unit 5 includes a control unit 50 that performs overall control, a transmission / reception unit 51 that performs communication with the network unit 4, and a section speed based on data received from each calculation unit 3. A section speed detecting unit 52 for detecting the above, a passing time information determining unit 53, a passing time information memory unit 54 for temporarily storing time series information, and a database unit 55 for accumulating and storing measurement data and analysis data, And a cycle information calculation unit (stroke information calculation means) 56 for calculating various types of stroke information. The cycle information calculation unit 56 includes a difference processing unit 560, a theoretical value calculation unit 561, a cycle calculation unit 562, and a stroke information calculation unit 563. Further, an external input device 8 such as a keyboard and a mouse is connected to the server unit 5, and various data necessary for period information calculation and input settings can be set.
[0019]
As shown in FIG. 4, the display processing unit 6 includes a control unit 60 that performs overall control, a transmission / reception unit 61 that performs communication with the network unit 4, and a numerical value display unit 62 that performs processing of text data to be displayed. And a graph display unit 63 for processing graphic data. The display processing unit 6 may be configured as a part of the server unit 5. Although not shown, when the display content displayed on the monitor 7 can be changed, it is preferable to connect an external input device to the display processing unit 6 so that the change of the display content can be selected. .
[0020]
Next, the operation of the stroke information measuring apparatus 1, that is, the stroke information measuring method using this apparatus will be specifically described.
[0021]
First, processing in the video camera 2 and the calculation unit 3 will be described. FIG. 5 is an example of image data acquired by the video camera 2 (FIG. 5A) and a diagram (FIG. 5B) for explaining a processing target range in the calculation unit 3. FIG. FIG. 7 is a graph showing time-series data generated by the calculation unit 3.
[0022]
Each video camera 2 outputs image data of the pool 100 in which the swimmer 150 is swimming every frame (that is, every 1/30 seconds). This image is, for example, the image shown in FIG. This image data is stored in the image memory unit 30 of the calculation unit 3 and is taken into the pixel extraction unit 32 for each frame (may be for each field) (step S1). In the area setting unit 31, the position data of the measurement range 140 in the course 120 and the measurement line 140 that is the measurement position is set in advance for each video camera 2 and stored as pixel position data on the image data. Based on the pixel position data of the measurement line 140 stored in the region setting unit 31, the pixel extraction unit 32 performs pixel extraction for extracting image data at the pixel position (step S2). The integration unit 33 binarizes the extracted pixel image data by threshold determination based on the hue value (step S3), and performs integration processing for integration along the measurement line 140 (step S4). By appropriately setting the threshold value, the image data of the swimmer 150 is accurately extracted from the background image of the pool 100 (including the water surface, floor surface and waves, water splash, course rope shadow, swimmer 150 shadow, etc.). can do. In this binarization, an image of the pool 100 in the absence of the swimmer 150, that is, background data is taken in advance, and a difference between the obtained image and the obtained image is calculated, or a threshold value is set for each RGB color. A determination may be made.
[0023]
The accumulated data is sent to the time-series data conversion unit 34, converted into predetermined time data and time-series data associated with the measurement position 140, and written into a memory or a recording medium (step S5). This time data may be generated by the number of frames from the video camera 2 or may be generated by a timer or the like in the accumulating unit 3. However, when the server unit 5 generates the time data based on the official recording clock, the time data It is preferable because the accuracy of data can be ensured and the data of each integrating unit 33 can be aligned. The processing of steps S2 to S5 is performed for each measurement position 140 existing in one image data, but this processing may be executed in parallel or by time division processing.
[0024]
The time variation of the integrated value in the specific measurement line 140 obtained in this way is as shown in FIG. The larger the integrated value, the longer the body length of the swimmer 150 that crosses the measurement line 140. Therefore, it is possible to determine the timing of the swimmer 150 passing through the measurement line 140 by determining the rising point of the peak of the integrated value. FIG. 7 shows the forward and backward passage timings as peak values when the 50-meter pool is reciprocated.
[0025]
The passage time determination unit 35 performs this peak detection for each measurement line 140 (step S6). If the peak is not detected, that is, if the swimmer 150 has not yet reached the measurement line 140 or has already passed, the process is terminated immediately after step S7, so that the next frame process can be performed. Transition. When a peak is detected, after step S7, the passage time of the measurement line 140 is determined from the rising point of this peak, and passage time data associated with the measurement line 140 is output. The transmission / reception unit 36 transfers this data to the server unit 5 via the network unit 4. The control unit 37 controls the start and end of processing in the calculation unit 3 and the setting / change of the measurement line 140 based on instructions from the server unit 5.
[0026]
Next, processing in the server unit 5 will be described. FIGS. 8 and 9 are flowcharts showing processing in the server unit, FIGS. 10 and 11 are graphs for explaining automatic extraction of stroke points, and FIGS. 12 and 13 are graphs showing passage time data. FIG. 14 is a graph showing an example of stroke information.
[0027]
First, the control unit 50 establishes a connection with the calculation unit 3 via the network unit 4 via the transmission / reception unit 51 (step S11). Next, an analysis start command is transmitted to each calculation unit 3 by an analysis start button input from the external input device 8 (step S12) (step S13). Thereby, each calculating part 3 starts the analysis and output of the passage time data mentioned above. Here, the analysis may be started by operating the external input device 8 by the operator. However, the analysis may be started in conjunction with the start signal of the recording staff, the start of the official recorder, or the like.
[0028]
Next, transit time data collection / analysis processing is performed (step S14). Specifically, the transmission / reception unit 51 receives the above-described passage time data from the transmission / reception unit 36 of each calculation unit 3 via the network unit 4 (step S21). And the section speed detection part 52 reads each information of the course of the measurement line 140, a position, and time from the received data (step S22). The read course and position information of the measurement line 140 is converted into distance information in the course 120 (step S23), and the section speed is calculated (step S24). Specifically, it indicates that the latest passing time data in the same course 120 has passed the position of L _n from one end of the pool 100 at t _n seconds, and the previous passing time data is at t _n-1 seconds. If it indicates that the position of L _n-1 has passed from the same end of the pool 100, the section speed V _n is V _n = (L _n -L _n-1 ) / (t _n -t _n-1 (In this case, when V _n is positive, the swimmer 150 moves in a direction away from this one end of the pool 100, and in the negative case, the swimmer 150 moves in a direction approaching this one end of the pool 100). Show). Next, the passage time information determination unit 53 determines whether or not the section speed V _n thus obtained matches a predetermined speed range (step S25). If the speed is out of the speed range, the section speed data is discarded by skipping the next process as error data (noise). When the speed range is met, the elapsed time information is written in the passage time information memory unit 54 and the database unit 55 for each course 120 (that is, for each swimmer 150).
[0029]
The elapsed time information is sent to the display processing unit 6 via the network unit 4 by the transmission / reception unit 51. Further, predetermined stroke information is obtained by processing by the period information calculation unit 56. This process will be described later.
[0030]
When the swimmer 150 finishes, an analysis stop button (signal) is input by the operator operating the external input device 8 or by a signal from the official recorder (step S15). And the control part 50 transmits an analysis stop command to the calculating part 3 by the transmission / reception part 51, and complete | finishes a process.
[0031]
The cycle information calculation unit 56 calculates the stroke cycle, the stroke time, and the stroke length based on the elapsed time information for each course 120 in real time or after all measurements are completed. The data can be calculated later, but if the calculation is performed in real time or immediately after the measurement is completed and provided to the swimmer, the swimmer can easily match the data with his / her sense of swimming and more effective use is possible. There is an advantage that can be achieved.
[0032]
The theoretical value calculator 561 calculates the estimated position L _sn at time t _n when the swimmer 150 is assumed to move at the average speed V _sn . Specifically, L _sn = L _sn−1 + (t _n −t _n−1 ) × V _sn The average velocity V _sn takes a long unit time t _x than the stroke time from the time t _n-1 -t _x may be used the average speed in the time t _n-1. Alternatively, it may be set as the average swimmer 150 speed.
[0033]
The difference processing unit 560 calculates a difference between the estimated position Lsn thus obtained and the actually obtained measurement data Ls, and transfers the difference to the cycle calculation unit 562. The period calculation unit 562 plots the transmitted difference data for each movement distance, and calculates a period point by the following method. In this case, the difference data takes a positive value when the movement distance of the swimmer is larger than the estimated value, and takes a negative value when the distance is small.
[0034]
First, the minimum point within the calculation range is obtained ((1) in FIG. 10). Next, the section where the difference data has the largest change in the minus direction before the minimum point (2 in the figure), and the section where the difference data has the largest change in the plus direction after the minimum point (3 in the figure). ) Then, an intersection point (4) in the figure obtained by extending the sections obtained in (2) and (3) is obtained, and the position pointed to by this intersection point is set as the distance of the periodic point. The time passing through this distance is calculated by interpolating from the previous and next points. This is the extraction stroke point. FIG. 12 shows a calculation example of the extracted stroke point. FIG. 12 is a plot of the position and time of the stroke point, and FIG. 13 is a plot of this plotted against the distance. Plot data of transit time against distance represents the movement of the most advanced part of the swimmer's body, and thus fluctuates periodically. Thus, when compared with the extracted stroke points, the swimmer's body excluding the difference in posture Can be compared to provide useful information for training.
[0035]
Furthermore, the cycle calculation unit 562 calculates the stroke length from the distance between the adjacent stroke points, and calculates the stroke time from the time difference between the adjacent stroke points. FIG. 14 shows changes in the stroke length, stroke time, and stroke speed (average speed during the stroke) obtained as described above. The reason why the stroke length and stroke time data are cut before and after 50 m is because accurate stroke information cannot be obtained because the swimmer 150 is around the turn.
[0036]
The control unit 60 of the display processing unit 6 receives measurement results and analysis results from the transmission / reception unit 61 via the network unit 4. Then, it is processed as text data in the numerical display unit 62 or processed as graphic data in the graph display unit 63 and displayed as text data as shown in FIG. 15 (a), or as shown in FIG. 15 (b). Display as a graph. Further, an overlay display may be performed on an image of an arbitrary video camera 2.
[0037]
In order for the period calculation unit 562 to accurately determine the stroke information by the above-described method, the measurement lines 140 are preferably arranged at intervals of about 1/8 of the stroke length. In the case of a Japanese champion level swimmer, the stroke length is 0.7 to 1.3 m for freestyle, backstroke, 1.8 to 2.2 m for butterfly, and 1.5 to 2.8 m for breaststroke. Therefore, in order to calculate the stroke length, it is preferable that the measurement lines 140 are set at intervals of about 11 cm, 25 cm, and 30 cm, respectively. On the other hand, even if the interval is shortened, the effect of the increase in processing becomes greater for improving accuracy, so the lower limit of the interval is considered to be about 1/16 of the stroke length. Further, the passage time can be widely used when the measurement lines 140 are arranged at predetermined intervals of 1 m. Therefore, it can be said that the measurement lines 140 are preferably set at intervals of 5 to 25 cm, more preferably at intervals of 10 cm or 20 cm in the case of the Japanese championship level. Note that the optimum measurement line 140 interval may vary depending on the level of the swimmer 150 and the stroke length.
[0038]
In the above description, the example in which the stroke point is extracted from the minimum point has been described, but the extraction of the stroke point is not limited to this. According to the present invention, as the passing time data, the time when the most advanced part of the body of the swimmer 150 passes through the measurement line 140 is obtained. On the other hand, in any swimming style, there may be a time when the head is temporarily positioned at the forefront of the body during the stroke. FIG. 16 is a diagram schematically showing this state. The tip part moves periodically according to the stroke by the stroke motion of the hand of the swimmer 150. In contrast, the head moves at a relatively constant speed. Therefore, for each stroke, find the time when the head is located at the tip of the swimmer (usually the tip exposed on the surface of the water), and interpolate the obtained time and distance with a straight line, spline function, etc. May be used to estimate the continuous head position. FIG. 17 is a graph showing the head position thus obtained together with the change in the tip position.
[0039]
In this way, by replacing the movement of the head position, it is possible to evaluate an actual swimming method that is not only a change in the movement of the hand, and the training effect can be evaluated accurately and stably.
[0040]
【The invention's effect】
As described above, according to the present invention, it is not necessary to attach a specific target to the swimmer, and the lap time at an arbitrary distance can be automatically measured, so that detailed stroke information can be provided to the swimmer. Furthermore, it can be used for training swimmers because it can be measured in competitions and the like. In addition, since various kinds of stroke information can be provided in real time or at a time shortly after the end of the competition, it is easy for swimmers and coaches to match the stroke information data with their own senses, and data utilization is enhanced.
[Brief description of the drawings]
FIG. 1 is a schematic configuration diagram showing a configuration of a stroke information measuring apparatus according to the present invention.
FIG. 2 is a detailed configuration diagram of a calculation unit of the apparatus of FIG. 1;
FIG. 3 is a schematic configuration diagram of a server unit of the apparatus of FIG. 1;
4 is a detailed configuration diagram of a display processing unit of the apparatus of FIG. 1;
5A is a diagram for explaining an example (a) of image data acquired by the video camera 2 and a processing target range in the calculation unit 3. FIG.
6 is a flowchart of processing in a calculation unit 3. FIG.
7 is a graph showing time-series data generated by the calculation unit 3. FIG.
FIG. 8 is a flowchart of main processing in the server unit 5;
FIG. 9 is a flowchart of the collection process of FIG.
FIG. 10 is a graph illustrating automatic extraction of stroke points.
FIG. 11 is another graph illustrating automatic extraction of stroke points.
FIG. 12 is a graph showing an example of passage time data.
FIG. 13 is a graph showing another example of passage time data.
FIG. 14 is a graph showing an example of stroke information.
FIG. 15 is a diagram showing a display example of the monitor 7;
FIG. 16 is a schematic graph showing the movement of the position of the tip and head of a body while a swimmer is swimming.
FIG. 17 is a graph corresponding to FIG. 16 using actual data.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 1 ... Stroke information measuring device, 2 ... Video camera, 3 ... Operation part, 4 ... Network part, 5 ... Server part, 6 ... Display processing part, 7 ... Monitor, 8 ... External input device, 30 ... Image memory part, 31 ... Area setting section, 32 ... Pixel extraction section (pixel extraction means), 33 ... Integration section (integration means), 34 ... Time-series data conversion section, 35 ... Passing time determination section (passing time determination means), 36 ... Transmission / reception section , 37 ... control unit, 50 ... control unit, 51 ... transmission / reception unit, 52 ... section speed detection unit, 53 ... passage time information determination unit, 54 ... passage time information memory unit, 55 ... database unit, 56 ... period information calculation unit (Stroke information calculation means), 60 ... control unit, 61 ... transmission / reception unit, 62 ... numerical value display unit, 63 ... graph display unit, 100 ... pool, 110 ... course rope, 120 ... course, 130 ... measurement range, 140 ... measurement Inn, 150 ... swimmer, 560 ... differential unit, 561 ... theoretical value calculating section, 562 ... period calculating section, 563 ... stroke information calculation unit.

Claims (3)

In a stroke information measuring device that images a swimmer during swimming with an imaging means and measures stroke information of the swimmer based on the acquired image data,
Pixel extraction means for extracting pixel data at a position corresponding to a predetermined measurement line in each acquired image data;
Integration means for binarizing the pixel values of the extracted pixels and integrating them along the measurement line ;
Based on the integrated data outputted from said integrating means, a transit time determining means for determining and outputting the swimmer the measurement line passing time of the rising edge of the peak at time series change,
Stroke information measuring device equipped with. 2. The stroke information measuring device according to claim 1, wherein a plurality of the imaging means are arranged along a course where a swimmer swims, and one or more predetermined measurement lines are arranged for each imaging means along the course.   The stroke information calculation means which calculates the information regarding the stroke containing a stroke period based on the swimmer's measurement line passage time data measured for every measurement line and the position data of the measurement line is further provided. Stroke information measuring device.

Images