WO2017193591A1 - 俯卧撑测试的计数方法及系统 - Google Patents
俯卧撑测试的计数方法及系统 Download PDFInfo
- Publication number
- WO2017193591A1 WO2017193591A1 PCT/CN2016/112312 CN2016112312W WO2017193591A1 WO 2017193591 A1 WO2017193591 A1 WO 2017193591A1 CN 2016112312 W CN2016112312 W CN 2016112312W WO 2017193591 A1 WO2017193591 A1 WO 2017193591A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- texture
- test
- frame
- texture image
- tester
- Prior art date
Links
- 238000012360 testing method Methods 0.000 title claims abstract description 158
- 238000000034 method Methods 0.000 title claims abstract description 46
- 230000008569 process Effects 0.000 claims abstract description 20
- 230000003247 decreasing effect Effects 0.000 claims description 14
- 239000000284 extract Substances 0.000 claims description 14
- 238000012545 processing Methods 0.000 claims description 13
- 230000004913 activation Effects 0.000 claims description 8
- 230000009471 action Effects 0.000 description 4
- 230000008859 change Effects 0.000 description 3
- 238000009825 accumulation Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 230000006698 induction Effects 0.000 description 2
- 238000005259 measurement Methods 0.000 description 2
- 238000005452 bending Methods 0.000 description 1
- 230000000903 blocking effect Effects 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 210000004744 fore-foot Anatomy 0.000 description 1
- 230000036541 health Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 230000001788 irregular Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 210000003205 muscle Anatomy 0.000 description 1
- 230000036961 partial effect Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000036962 time dependent Effects 0.000 description 1
- 210000001364 upper extremity Anatomy 0.000 description 1
Images
Classifications
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0003—Analysing the course of a movement or motion sequences during an exercise or trainings sequence, e.g. swing for golf or tennis
- A63B24/0006—Computerised comparison for qualitative assessment of motion sequences or the course of a movement
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0062—Monitoring athletic performances, e.g. for determining the work of a user on an exercise apparatus, the completed jogging or cycling distance
-
- A—HUMAN NECESSITIES
- A63—SPORTS; GAMES; AMUSEMENTS
- A63B—APPARATUS FOR PHYSICAL TRAINING, GYMNASTICS, SWIMMING, CLIMBING, OR FENCING; BALL GAMES; TRAINING EQUIPMENT
- A63B24/00—Electric or electronic controls for exercising apparatus of preceding groups; Controlling or monitoring of exercises, sportive games, training or athletic performances
- A63B24/0087—Electric or electronic controls for exercising apparatus of groups A63B21/00 - A63B23/00, e.g. controlling load
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
- G06V40/23—Recognition of whole body movements, e.g. for sport training
Definitions
- the present invention relates to the field of image recognition processing technologies, and more particularly to the application of push-up test counting.
- Infrared sensing counting method Although the disadvantages of manual operation are reduced, in the specific operation, it is necessary to manually set the highest point and the lowest point, and manually turn on the counting function, and the human-computer interaction experience is still not strong.
- Radio frequency identification counting method The tester must carry the proximity card and use the distance sensed by the induction card to achieve counting. However, this method requires the tester to force the induction card, which is extremely inconvenient for the tester.
- the present invention provides a counting method for a push-up test, which includes:
- a first identification area and a second identification area each having a specific texture are set in the test area, where the first identification area is an area where the upper body of the tester is in the test process, and the second identification area is an area where the tester's legs are located during the test; Setting the camera obliquely above the test area;
- step S3 Analyze the first texture image of each frame from the start frame, and determine the minimum pixel of the texture in the first texture image when it is found that the pixel value of the texture in the first texture image changes from gradually decreasing to gradually increasing. Whether the value is less than a preset low threshold, if yes, step S4 is performed; otherwise, the minimum of the texture pixel value of the first texture image is set as the start frame, and then step S3 is repeatedly executed;
- step S4 Continue to analyze the first texture image of each subsequent frame. When it is found that the pixel value of the texture in the first texture image changes from gradually increasing to decreasing gradually, determine whether the maximum pixel value of the texture in the first texture image is greater than one.
- the preset high-order threshold is performed in step S5, otherwise the frame with the largest grain pixel value of the first texture image is set as the start frame, and then returns to step S3;
- the following steps are further included between the steps S2 and S3, and the first frame of the self-timer is opened. Extracting and analyzing the first texture image, when it is first found that the pixel value of the texture in the first texture image of the current frame is less than a preset activation threshold, the current frame is set as the start frame, and the timing starts; in steps S3 and S4, the extraction is performed. Before the first texture image of the frame, it is first determined whether the time relative to the timing of the frame has reached the preset duration, and the test is ended, otherwise the test is continued.
- steps S3 and S4 if the pixel value of the texture in the first texture image of the current frame is found to be greater than a preset startup threshold, the test is terminated.
- the specific texture disposed in the first identification zone is at least one long line, and the extension direction of the long line is substantially perpendicular to the prone direction of the tester; the pixel value of the texture in the first texture image analyzed in each step
- the length of the long line; the low threshold is the maximum length of the long line that the camera can capture when the tester maintains the standard push-up position; the high threshold is the tester's body rises when the tester maintains the standard push-up position.
- the second identification zone is spaced apart from two positioning lines extending substantially along the prone direction of the tester. During testing, the tester's legs are respectively placed along two positioning lines, and the second identification zone is set to be specific.
- the texture is at least one short line, and the short line is disposed between the two positioning lines and the extending direction is substantially perpendicular to the prone direction of the tester.
- the pixel value of the texture in the second texture image determined in step S5 is a short line length, which is greater than a preset.
- the kneeling threshold is any short line length greater than zero.
- the invention also provides a counting system for a push-up test, comprising:
- a test area comprising a first identification area and a second identification area having a specific texture, wherein the first identification area is an area where the upper body of the tester is in the test process, and the second identification area is an area where the tester's legs are located during the test;
- a camera which is placed obliquely above the test area, and the camera continues to shoot at a certain frame rate. a test area image of the tester, the first identification area, and the second identification area, obtaining a first texture image of the first identification area and a second texture image of the second identification area;
- the low-position attitude analysis module extracts and analyzes the first texture image of each frame from the start frame, and determines the first texture image when it is found that the pixel value of the texture in the first texture image changes from gradually decreasing to gradually increasing. Whether the minimum pixel value of the middle grain path is less than a preset low level threshold, the high position attitude analysis module is activated, otherwise the frame with the smallest grain pixel value of the first texture image is set as the start frame and the first texture is continued for the subsequent frames. Whether the minimum pixel value of the texture in the image is less than the low threshold value for judging processing;
- the high-position attitude analysis module further extracts and analyzes the first texture image of each subsequent frame after starting, and determines the texture of the first texture image when it is found that the pixel value of the texture in the first texture image changes gradually and gradually decreases. Whether the maximum pixel value is greater than a preset high-order threshold, the posture analysis module is activated, otherwise the frame with the largest grain value of the first texture image is set as the start frame, and the low-level attitude analysis module is started;
- the posture analysis module extracts all the second texture images during the frame period from the start frame to the first texture image of the first texture image, and determines whether the pixel values of the textures in all the second texture images are greater than one.
- the preset kneeling threshold is issued by the counting instruction; otherwise, the frame with the largest grain value of the first texture image is set as the starting frame, and the low-post posture analysis module is started;
- the counting module counts the push-up test once after receiving the counting instruction issued by the kneeling analysis module.
- the system further includes: a timing module that extracts and analyzes the first texture image from the first frame of the self-shooting before the low-position attitude analysis module starts working, when the first frame of the first texture image is first found in the current frame
- a timing module that extracts and analyzes the first texture image from the first frame of the self-shooting before the low-position attitude analysis module starts working, when the first frame of the first texture image is first found in the current frame
- the current frame is set as the start frame, the timing starts, and the time duration data of the current frame relative to the timing start time is recorded in real time
- the test end determination module The time length data recorded by the timing module is retrieved, and before the low-position attitude analysis module and the high-position attitude analysis module extract the first texture image of a certain frame, it is determined whether the time length data corresponding to the frame has reached the preset duration, and the end is This test, otherwise indicates that the low position analysis module or the high position analysis module continues the test.
- system further includes a temporary termination determination module that determines whether the pixel value of the texture in the first texture image of the current frame is greater than a pre-determination before the low-position attitude analysis module and the high-level attitude analysis module analyze the current frame first texture image. If the startup threshold is set, the test is ended, otherwise the low-level attitude analysis module or the high-level attitude analysis module is instructed to continue the test.
- the specific texture disposed in the first identification area is at least one long line, and the extension direction of the long line is substantially perpendicular to the prone direction of the tester; the first texture analyzed by the low position analysis module and the high position analysis module
- the pixel value of the image in the image is the length of the long line; the low threshold is the maximum length of the long line that the camera can capture when the tester maintains the normal push-up position; the high threshold is the tester's standard push-up posture. When the tester's body rises to the highest position, the minimum length of the long line can be captured by the camera.
- the second identification zone is spaced apart from two positioning lines extending substantially along the prone direction of the tester. During testing, the tester's legs are respectively placed along two positioning lines, and the second identification zone is set to be specific.
- the texture is at least one short line, the short line is disposed between the two positioning lines and the extending direction is substantially perpendicular to the prone direction of the tester, and the pixel value of the texture in the second texture image determined by the kneeling analysis module is a short line length, greater than a pre- Set the kneeling threshold to any short line length greater than zero.
- the push-up test counting system provided by the present invention can also prevent user cheating.
- the push-up test process is thus automated, reducing manual intervention, significantly improving test efficiency, and a good test experience.
- FIG. 1 is a schematic flow chart of a push-up test counting method provided by the present invention.
- FIG. 2 is a schematic view of a test area in an embodiment of the present invention.
- FIG. 3 is a partial schematic view of a frame test image taken by a camera
- FIG. 4 is a line image of an untested person after processing the test image of FIG. 3;
- FIG. 5 is a schematic diagram of a push-up test counting system provided by the present invention.
- the implementation and application of the present invention is primarily, but not exclusively, limited to the National Physique Measurement Program.
- the push-up test reflects the upper limbs, shoulder and back muscle strength and continuous working ability.
- the tester is required to hold the ground with both hands, the fingers forward, the distance between the hands and the shoulder width, the body straight, the arm bends the body straight down to the same level of the shoulder and elbow, then the body is straight up and restored to In the preparatory posture, one is completed, and the number of consecutive push-ups is obtained by such a combination.
- the body is not kept straight (such as bending or touching the legs) or the body is not at the same level as the shoulders and elbows, this situation will not be counted.
- the present invention uses the image recognition processing means to realize automatic recognition and counting of push-ups without manual intervention.
- FIG. 1 A schematic flow chart of a counting method for a push-up test as shown in FIG.
- step S1 a first identification area and a second identification area each having a specific texture are set in the test area.
- the first identification area is the area where the upper body of the tester is in the test process, and the second identification area is the tester's legs in the test process. Area; set the camera diagonally above the test area.
- Test area 210 includes, but is not limited to, providing a push-up test for the tester.
- the second identification area 212 is not limited to be in contact with the first identification area 211.
- the second identification area 212 is narrower than the first identification area 211 in the direction in which the width W is extended, in order to distinguish the position of the upper body and the legs, and to indicate the prone direction.
- the second identification area 212 is spaced apart from the positioning line 2122a and the positioning line 2122b extending substantially in the prone direction of the tester, and the tester's legs are placed along the positioning line 2122a and the positioning line 2122b, respectively, for testing.
- the first identification area 211 and the second identification area 212 further include strip-like specific lines in the same straight line.
- the specific texture in the first identification area 211 is marked as the first texture 2111, and is located in the second identification area.
- the specific texture of 212 is labeled as second texture 2121.
- the first grain 2111 is at least one long line
- the second grain 2121 is at least one short line disposed between the positioning line 2122a and the positioning line 2122b, and the extending direction of the first grain 2111 and the second grain 2121 are in the prone direction of the tester. Basically vertical, the length and length of the line are relative.
- the specific texture of the present invention is not limited in its pattern. In an alternative embodiment, the specific texture may also be a linear pattern of uniformly distributed curves, broken lines, arcs, or the like, or may include a pattern composed of regular geometric figures, or Combination of light and dark geometric figures, etc.
- image capturing is performed on the first recognition area 211 and the second recognition area 212 by setting a camera (not shown in FIG. 2).
- the position of the camera is preferably set to a side substantially perpendicular to the prone direction of the tester, as illustrated by the field of view of FIG.
- step S2 the image of the test area including the tester, the first identification area and the second identification area is continuously captured by the camera at a certain frame rate, and the first texture image in the first identification area and the second identification area are obtained.
- the second grain image is obtained.
- the tester can continue to perform image shooting with the aforementioned camera. Since the camera has a specific frame rate, a number of time-dependent testers can be obtained.
- FIG. 3 partially illustrates a test image of the tester in the range of the first identification area 211.
- the change in the length value of the first texture 2111 and the second texture 2121 in the test image will be analyzed, and the number of tester's push-ups is identified, which will be detailed in later steps.
- the length is specifically represented by a plurality of pixels, so the fingerprint path length of the present invention can be equivalently interpreted as a texture pixel value.
- the accuracy of the test of the method of the present invention can be improved by effectively determining the starting frame of the push-up test. Since the upper body is the main implementation site of the push-up operation, the following description will be understood by referring to FIGS. 2 and 3 in the following description.
- the first texture 2111 in the acquired first recognition area 211 is analyzed in real time from the time when the camera starts to acquire the test image frame by frame, specifically by analyzing the change of the length value of the first texture 2111. Because the upper body of the tester will block the first texture 2111 to a different extent during the lowering of the flexing arm or the raising of the arm, the first texture 2111 will exhibit different lengths in the obtained frame test images.
- the present invention selects one of the first lines 2111a for a brief description.
- a start threshold is preset for the pixel value of the first texture 2111a (characterized as a length value in this embodiment).
- the upper part of the tester is in the first identification area 211, the legs are placed in the second identification area 212, the body is flat, the hands are supported, the legs are stretched back and forth, the forefoot is on the ground, the body is straight, and the first frame of the camera self-timer is taken.
- the pixel value of the first texture 2111a in the first texture image of the current frame (that is, the length value in the real-time example) is less than a preset activation threshold, it is regarded as a tester.
- Formally enter the test state set the current frame as the start frame, the test starts, and the timing starts.
- step S3 the first texture image of each frame from the start frame is analyzed, and when the pixel value of the texture in the first texture image is changed from gradually decreasing to gradually increasing, the minimum texture in the first texture image is determined. If the pixel value is less than a preset low threshold, step S4 is performed, otherwise step S3 is repeated after setting the minimum of the texture pixel value of the first texture image as the start frame.
- a complete push-up includes the process of lowering the arm and raising the arm.
- the pixel value of the first texture 2111a continuously acquired by the camera exhibits a gradually decreasing state (the extent to which the tester upper body blocks the first texture 2111a gradually increases);
- the pixel value of the first texture 2111a continuously acquired by the camera is gradually increased (the degree of the upper body blocking the first texture 2111a is gradually reduced).
- the analysis of the first texture 2111a in this step and subsequent steps is an analysis of its length value, but should not be limited thereto.
- the present invention presets a low threshold, which is regarded as the judgment point at which the tester reaches the same horizontal plane of the shoulder and the elbow. That is, the low threshold is the maximum length of the aforementioned first texture 2111a that can be captured by the camera when the body is lowered to the lowest position when the tester maintains the normal push-up position. Then, the first texture 2111a is image-processed from the start frame, and when the length value of the first texture 2111a is changed from gradually decreasing to gradually increasing, it is determined whether the minimum pixel value is smaller than the preset. If the lower threshold is reached, the process proceeds to step S4. If not, the frame having the smallest length value is set as the start frame, and then the operation of step S3 is repeated.
- step S4 the first texture image of each subsequent frame is continuously analyzed.
- step S5 it is determined whether the maximum pixel value of the texture in the first texture image is greater than a preset high threshold, then step S5 is performed, otherwise the first texture map is executed.
- the frame is set to the start frame and then returns to step S3.
- the present invention presets a high threshold, which is regarded as the high position that the upper body can reach when the tester follows the standard push-up posture. point. That is to say, the high position threshold is the minimum value of the length of the first texture 2111a that the tester's body is raised to the highest position when the tester maintains the normalized push-up posture.
- the high position threshold is the minimum value of the length of the first texture 2111a that the tester's body is raised to the highest position when the tester maintains the normalized push-up posture.
- step S3 in the image processing process of the subsequent first texture 2111a, when it is found that the length value of the first texture 2111a changes from gradually increasing to decreasing gradually, it is further determined whether the maximum pixel value is greater than a preset. If the high level threshold is reached, the process proceeds to step S5. If not, the frame of the maximum pixel value is set as the start frame and the operation returns to step S3.
- the tester in the push-up test, the tester is required to keep a certain distance between the legs and the ground during the test, and cannot bend or touch the ground.
- the present invention is specifically determined by the change in the length value of the second texture during the test, and the length value referred to herein is still equivalently understood as the pixel value of the second texture 2121 in the test image.
- the ⁇ The attitude threshold is regarded as a situation in which the tester has bent or touched the legs, and in the obtained test image, at least one second texture 2121 has a length value of zero.
- one of the second lines 2121 is selected for brief description. Specifically, during the determination of the foregoing steps S3 to S4, that is, the maximum length from the start frame to the first texture 2111a During the frame period, only the length value of the second texture 2121 is allowed to be greater than the kneeling threshold (that is, the length of any one of the short lines is greater than zero), so that the tester can be regarded as the push-up action during the period, and counted once; If the pixel value of the second texture 2121 is less than the preset kneeling threshold, that is, the legs are bent or touched, the push-ups during this period are invalid and are not counted. Based on this, the present invention can prevent cheating in the test process.
- the kneeling threshold that is, the length of any one of the short lines is greater than zero
- the present invention enables automatic activation, recognition, and counting of a push-up.
- the accumulation of the number of times is a repeated reproduction of the steps described in the method.
- the present invention is described in two scenarios, which may be applied separately or in combination:
- step S3 and S4 before extracting the first texture image of a certain frame, it is first determined whether the time at which the frame is located has reached a preset one minute duration. If yes, stop extracting the first texture image of the current frame, the timing is terminated, the test ends, and the accumulated number of push-ups is saved; otherwise, the test is continued.
- Test for unspecified time In this case, the tester judges by himself whether the physical limit is reached and decides whether to continue the test. In this regard, during the test from the start of the frame, it is continuously determined whether the pixel value of the first texture of the current frame is greater than the activation threshold, and if so, the tester is regarded as having a lifting action, the timing is over, and the accumulated number of push-ups is saved.
- the embodiment of the present invention adopts the following method: First, the camera shoots a test area where no tester is present before the test, and obtains a template image. The coordinate information of the first recognition area and the second identification area is located; then, the camera continuously captures the test area where the tester is present during the test, and obtains each frame test image. Then, in the image processing of the current frame, the current frame test image and the template image obtained in advance are respectively extracted according to the coordinate information to obtain two edge information, and the two edge information are obtained and compared. The edge information of the occluded line is finally obtained.
- the length information of the line in the current frame test image can be obtained by filling the edge information of the line.
- the processing image in FIG. 3 can be processed as shown in the figure. An image of the line shown by 4 that is not obscured by the tester.
- the present invention also provides a push-up test counting system using the foregoing method, as shown in FIG. 5, which includes:
- Test area 210 includes, but is not limited to, providing a push-up test for the tester.
- the test area 210 is provided with a first identification area 211 and a second identification area 212, respectively, to indicate the area where the tester's upper body and its legs should be placed.
- the layout of the first identification area 211 and the second identification area 212 is the same as that shown in FIG. 2, and details are not described herein again.
- the second identification area 212 is spaced apart from the positioning line 2122a and the positioning line 2122b extending substantially along the prone direction of the tester.
- the tester's legs are respectively along the positioning line 2122a and the test.
- the positioning line 2122b is placed.
- the first identification area 211 includes at least one long line inside
- the second identification area 212 includes at least one short line between the two positioning lines, and the extending direction of the long line and the short line are substantially perpendicular to the tester's prone direction.
- the length and length of the lines are relative.
- the pattern of the specific texture is not limited.
- the specific texture may be a linear pattern of uniformly distributed curves, broken lines, arcs, or the like, or may include a pattern composed of a geometric shape such as a rectangle. Or a combination of light and dark geometric patterns.
- Camera 220 its position is as shown in the field of view of FIG. 2, which is disposed obliquely above the test area for image capturing of the first identification area 211 and the second identification area 212, the position of which is preferably set in the prone direction of the tester. Vertical side.
- the camera 220 continuously captures the test image including the tester, the first identification area 211 and the second identification area 212 at a certain frame rate, and obtains the first texture image of the first identification area 211 and the second texture of the second identification area 212. image.
- Low-position attitude analysis module 230 In the analysis of the module, the long line length value of the first texture is still equivalently understood as a pixel value, but is not limited thereto. A low threshold is preset, and its function is as described in the previous method, and will not be described again. In the description of this module, a brief description is made by taking one of the first lines. The module receives the first texture image, and analyzes the first texture image of each frame from the start frame.
- the high-level attitude analysis module is activated, otherwise the minimum texture pixel value of the first texture image is set as the start frame and then continues to the subsequent frames. Whether the minimum pixel value of the texture in the texture image is smaller than the low threshold is used for judging processing.
- the high-position attitude analysis module 240 receives the start signal sent by the low-level attitude analysis module, and continues to extract and analyze the first texture image of the subsequent frames after startup.
- a high threshold is preset, and its function is as described in the previous method, and will not be described again.
- the pose analysis module is activated. Otherwise, the frame with the largest grain value of the first texture image is set as the start frame, and the low posture analysis module is started.
- Kneeling analysis module 250 In the analysis of the module, the short line length value of the second texture is still equivalently understood as a pixel value, but is not limited thereto. Firstly, the preset threshold value is preset, and its function is as described in the previous method, and will not be described again. In the description of the kneeling analysis module 250, by briefly describing the second texture, when the tester bends or touches the ground, at least one second texture length value is zero.
- the posture analysis module 250 starts after receiving the activation signal sent by the high-position attitude analysis module, and extracts all the second texture images during the frame from the start frame to the maximum of the texture length value of the first texture image, and determines all the Whether the length value of the texture in the second texture image is greater than a preset threshold position, and Otherwise, the frame with the largest texture length value of the first texture image is set as the start frame, and the low-level attitude analysis module is started to continue analyzing the first texture image of the subsequent frames.
- the counting module 260 after receiving the counting instruction issued by the kneeling analysis module 250, counts the push-up test once.
- the system enables automatic identification and counting of a push-up action.
- the accumulation of the number of times only needs to be repeated by the relevant modules of the system, that is, the repetition of the low-position attitude analysis module, the processing of the first texture by the high-position attitude analysis module, and the processing of the second texture by the posture analysis module, The count of the counting module is accumulated.
- system is designed to automatically start and end the test, further including:
- a timing module before the low-position attitude analysis module starts working, extracting and analyzing the first texture image from the first frame captured by the camera, and first finding that the length value in the first texture image of the current frame is less than a preset startup threshold.
- the timing starts, and the duration data of the current frame relative to the start time of the current frame is recorded in real time.
- the automatic end is divided into two situations, which can be applied separately or in combination:
- the first one is for tests with a specified time, such as a 1-minute test.
- the system further includes a test end determination module, which retrieves the duration data recorded by the timing module, and determines the duration data corresponding to the frame before the low posture analysis module and the high posture analysis module extract the first texture image of the frame. Whether the preset duration has been reached, and the test is ended, otherwise the low posture analysis module or the high posture analysis module is instructed to continue the test.
- the system further includes a temporary termination determination module that determines that the current frame is first before the low-position attitude analysis module and the high-position attitude analysis module analyze the current frame first texture image Whether the pixel value of the texture in the texture image is greater than a preset activation threshold is to end the test, otherwise the low posture analysis module or the high posture analysis module is instructed to continue the test.
- the self-service test of the push-ups realized by the invention can reduce the labor cost, improve the test efficiency, and give the tester a good test experience.
Landscapes
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Physical Education & Sports Medicine (AREA)
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Psychiatry (AREA)
- Social Psychology (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Analysis (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
一种俯卧撑测试的计数方法及对应的计数系统，该计数方法通过在测试区域分别设置具有特定纹路的第一识别区和第二识别区，第一识别区包括第一纹路，第二识别区包括第二纹路（S1）；测试中，利用设置于测试区域斜上方的摄像机以一定帧率持续拍摄获取包括测试者、第一纹路和第二纹路在内的测试图像（S2）；在自起始帧往后的各帧测试图像中，判定是否第一纹路像素值先在由减小变为增大的过程中其最小值小于低位阈值（S3），接续再在由增大变为减小的过程中其最大值大于高位阈值（S4），若是则通过判定第二纹路的像素值在前述判定期间是否均大于跪姿阈值（S5），若是则计数一次。该方案可实现自助式的俯卧撑测试与计数，测试过程无需人工干涉，提高测试效率。
Description
本发明涉及图像识别处理技术领域，尤其涉及俯卧撑测试计数的应用。
目前，随着社会经济发展、人民对健康认识的提高、以及国际形势的变化，国家对整体国民体质的增强已经提升到战略性的高度。在这个大背景下，迫切需要对全民体质进行更高效、准确的监测。
在国家体育总局群体司于2003年发布的《国民体质测定标准手册及标准(成年人部分)》中，俯卧撑被列入国民体质的测试项目之一。
但是，传统的俯卧撑实施计数方法中，通常有人工计数、红外感应计数以及射频识别计数，但各自存在不足：
1、人工计数方法：存在测试效率不高，且误差率大，长时间的反复性操作也容易造成测试员疲惫，对不规范动作不易察觉等不足，现已较少应用于体质检测。
2、红外感应计数方法：虽然减少了人工操作的弊端，但在具体操作中，其需依靠手动设定最高点和最低点，手动开启计数功能，人机交互体验感仍然不强。
3、射频识别计数方法：测试者必须携带感应卡，利用感应卡感应的距离来实现计数，但该方法要求测试者强制携带感应卡，对测试者来说极其不方便。
因此，面对当代人们生活质量的提高，对体质检测的越发关注，对检测项目的测试精度要求越来越高的情形下，上述提及的俯卧撑测试的计数方法尚不具备良好的测试体验，测试效率仍有待提高。
发明内容
为解决上述技术问题，本发明提供一种俯卧撑测试的计数方法，包括：
S1、在测试区域设置各具有特定纹路的第一识别区和第二识别区，第一识别区为测试过程中测试者上半身所在区域,第二识别区为测试过程中测试者双腿所在区域；在测试区域的斜上方设置摄像机；
S2、利用摄像机以一定帧率持续拍摄包括测试者、第一识别区和第二识别区在内的测试区域图像，得到第一识别区内的第一纹路图像以及第二识别区内的第二纹路图像；
S3、分析自起始帧往后的各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐减小变化为逐渐增大时，判定该第一纹路图像中纹路的最小像素值是否小于一预设的低位阈值，是则执行步骤S4，否则将第一纹路图像的纹路像素值最小那帧设置为起始帧后重复执行步骤S3；
S4、继续分析后续各帧的第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐增大又变化为逐渐减小时，判定该第一纹路图像中纹路的最大像素值是否大于一预设的高位阈值，是则执行步骤S5，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后返回执行步骤S3；
S5、提取自起始帧至第一纹路图像的纹路像素值最大那帧期间所有的第二纹路图像，判定该所有的第二纹路图像中纹路的像素值是否均大于一预设的跪姿阈值，是则计数一次，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后返回执行步骤S3。
在可选实施例中，步骤S2与S3之间还包括如下步骤，自拍得的第一帧开
始提取分析第一纹路图像，当首次发现当前帧第一纹路图像中纹路的像素值小于预设的启动阈值时，将当前帧设为起始帧，计时开始；步骤S3和S4中，提取某帧的第一纹路图像之前，先判定该帧所在时刻相对计时开始时刻是否已达到预设时长，是则结束本次测试，否则继续测试。
在可选实施例中，步骤S3和S4中，若发现当前帧第一纹路图像中纹路的像素值大于预设的启动阈值，则结束本次测试。
在可选实施例中，第一识别区内设置的特定纹路为至少一长线条，该长线条的延伸方向与测试者俯卧方向基本垂直；各步骤中分析的第一纹路图像中纹路的像素值为长线条长度；低位阈值为测试者保持规范的俯卧撑姿势时身体降到最低位摄像机所能拍到的长线条长度的最大值；高位阈值为测试者保持规范的俯卧撑姿势时测试者身体升到最高位摄像机所能拍到的长线条长度的最小值。
在可选实施例中，第二识别区内间隔设置有基本沿着测试者俯卧方向延伸的两定位线条，测试时测试者的双腿分别沿着两定位线条放置，第二识别区设置的特定纹路为至少一短线条，短线条设置在两定位线条之间且延伸方向与测试者俯卧方向基本垂直，步骤S5中判定的第二纹路图像中纹路的像素值为短线条长度，大于一预设的跪姿阈值为任意一条短线条长度均大于零。
本发明还提供一种俯卧撑测试的计数系统，包括：
测试区域，其包括具有特定纹路的第一识别区和第二识别区，第一识别区为测试过程中测试者上半身所在区域，第二识别区为测试过程中测试者双腿所在区域；
摄像机，其设置在测试区域的斜上方，该摄像机以一定帧率持续拍摄包括
测试者、第一识别区和第二识别区在内的测试区域图像，得到第一识别区的第一纹路图像以及第二识别区的第二纹路图像；
低位姿态分析模块，其提取并分析自起始帧往后的各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐减小变化为逐渐增大时，判定第一纹路图像中纹路的最小像素值是否小于一预设的低位阈值，是则启动高位姿态分析模块，否则将第一纹路图像的纹路像素值最小那帧设置为起始帧后继续对后续各帧第一纹路图像中纹路的最小像素值是否小于低位阈值进行判断处理；
高位姿态分析模块，其启动后继续提取并分析后续各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐增大又变化为逐渐减小时，判定第一纹路图像中纹路的最大像素值是否大于一预设的高位阈值，是则启动跪姿分析模块，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后启动低位姿态分析模块；
跪姿分析模块，其启动后提取自起始帧至第一纹路图像的纹路像素值最大那帧期间所有的第二纹路图像，判定该所有的第二纹路图像中纹路的像素值是否均大于一预设的跪姿阈值，是则发出计数指令，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后启动低位姿态分析模块；
计数模块，其接收到跪姿分析模块发出的计数指令后，对俯卧撑测试计数一次。
在可选实施例中，该系统进一步包括：计时模块，其在低位姿态分析模块开始工作前，自拍得的第一帧开始提取分析第一纹路图像，当首次发现当前帧第一纹路图像中的像素值小于预设的启动阈值时，将当前帧设为起始帧，计时开始，后续实时记录当前帧相对计时开始时刻的时长数据；测试结束判定模块，
其调取计时模块记录的时长数据，并在低位姿态分析模块和高位姿态分析模块提取某帧的第一纹路图像之前，先判定该帧所对应的时长数据是否已达到预设时长，是则结束本次测试，否则指示低位姿态分析模块或者高位姿态分析模块继续测试。
在可选实施例中，该系统进一步包括临时终止判定模块，其在低位姿态分析模块以及高位姿态分析模块分析当前帧第一纹路图像之前判定当前帧第一纹路图像中纹路的像素值是否大于预设的启动阈值，是则结束本次测试，否则指示低位姿态分析模块或者高位姿态分析模块继续测试。
在可选实施例中，第一识别区内设置的特定纹路为至少一长线条，该长线条的延伸方向与测试者俯卧方向基本垂直；低位姿态分析模块和高位姿态分析模块分析的第一纹路图像中纹路的像素值为长线条长度；低位阈值为测试者保持规范的俯卧撑姿势时身体降到最低位摄像机所能拍到的长线条长度的最大值；高位阈值为测试者保持规范的俯卧撑姿势时测试者身体升到最高位摄像机所能拍到的长线条长度的最小值。
在可选实施例中，第二识别区内间隔设置有基本沿着测试者俯卧方向延伸的两定位线条，测试时测试者的双腿分别沿着两定位线条放置，第二识别区设置的特定纹路为至少一短线条，短线条设置在两定位线条之间且延伸方向与测试者俯卧方向基本垂直，跪姿分析模块判定的第二纹路图像中纹路的像素值为短线条长度，大于一预设的跪姿阈值为任意一条短线条长度均大于零。
实施本发明，能够实现俯卧撑的有效识别与计数，而且，本发明提供给的俯卧撑测试计数系统还能防止用户作弊。俯卧撑测试过程由此实现自动化，减少人工干预，测试效率明显提高，测试者体验感好。
图1是本发明提供的一种俯卧撑测试计数方法的流程示意图；
图2是本发明实施例中测试区域示意图；
图3是摄像机拍摄的某帧测试图像局部示意图；
图4是对图3测试图像进行处理后的未被测试者遮挡的线条图像；
图5是本发明提供的一种俯卧撑测试计数系统示意图。
本发明的实施和应用主要但不限于是在国民体质测定项目中。
在国民体质测定中，俯卧撑测试反映人体上肢、肩背部肌肉力量及持续工作能力。测试时，要求测试者双手撑地，手指向前，双手间距与肩同宽，身体挺直，屈臂使身体平直下降至肩与肘处于同一水平面，然后将身体平直撑起，恢复至预备姿势则为完成1个，并以此类推累计得出俯卧撑的连续个数。其中需注意的是，如果身体未保持平直(如双腿弯曲或触地)或身体未将至肩与肘处于同一水平面，该情况不予计数。
基于对动作的规范定义，本发明运用图像识别处理手段实现俯卧撑的自动识别和计数，无需人工干预。
如图1所示的一种俯卧撑测试的计数方法流程示意图。
步骤S1中，在测试区域设置各具有特定纹路的第一识别区和第二识别区，第一识别区为测试过程中测试者上半身所在区域,第二识别区为测试过程中测试者双腿所在区域；在测试区域的斜上方设置摄像机。
请结合参照如图2，图2是本发明所提供的测试区域示意图。测试区域210包括但并不限于为测试者提供俯卧撑测试。测试区域210内通过设置第一识别区211和第二识别区212，如图中右左所示宽度不同的两矩形区域，分别标示出测试者在测试时上半身及双腿所应放置的位置，第二识别区212不限于与第一识别区211相接。图中，第二识别区212在其宽度W延长方向上较第一识别区211窄，是为区别上半身与双腿的位置，示意俯卧方向。第二识别区212内间隔设置有基本沿着测试者俯卧方向延伸的定位线条2122a和定位线条2122b，示意测试时测试者的双腿分别沿着定位线条2122a和定位线条2122b放置。第一识别区211和第二识别区212内部还包括呈同样直线的条状特定纹路，为方便后续的描述，位于第一识别区211的特定纹路标记为第一纹路2111，位于第二识别区212的特定纹路标记为第二纹路2121。第一纹路2111为至少一长线条，第二纹路2121为至少一设置在定位线条2122a和定位线条2122b之间的短线条，第一纹路2111以及第二纹路2121的延伸方向均与测试者俯卧方向基本垂直，线条的长与短是相对而言。本发明特定纹路其图案不受限制，在可选实施例中，特定纹路还可为均匀分布的曲线、折线、弧线等线状图案，也可以是包括由规则几何图形组成的图案、或是明暗相间的几何图形组合等。
另外，在测试区域210的斜上方一侧，通过设置摄像机(图2中未示出)对第一识别区211和第二识别区212进行图像拍摄。该摄像机的位置优选设置在与测试者俯卧方向基本垂直的一侧，如图2视野所示意的位置。
步骤S2中，利用摄像机以一定帧率持续拍摄包括测试者、第一识别区和第二识别区在内的测试区域图像，得到第一识别区内的第一纹路图像以及第二识别区内的第二纹路图像。
当测试者在图2所示的测试区域210做好俯卧撑测试准备时，可用前述的摄像机对测试者持续做图像拍摄，由于摄像机的拍摄具有特定帧率，因此可得到依时序的若干包括测试者、第一识别区211和第二识别区212的测试图像。图3局部示意出测试者在第一识别区211范围的测试图像。在后续的图像处理步骤中将分析第一纹路2111和第二纹路2121在测试图像中的长度值变化情况，对此识别出测试者的俯卧撑完成个数，将在后面步骤中详述。需说明的是，在测试图像中，长度具体是由若干像素体现，所以本发明所指纹路长度，可等同理解为纹路像素值。
另外，能够有效判断出俯卧撑测试的起始帧，则可提高本发明方法测试的准度。由于上半身是俯卧撑动作的主要实施部位，因此在以下说明中仍请通过参考图2和图3来理解以下说明。
自摄像机开始逐帧拍摄获取测试图像起，实时对获取的第一识别区211中的第一纹路2111进行分析，具体通过分析第一纹路2111的长度值变化。因为测试者的上半身在屈臂下降或撑臂上升过程中，会以不同程度遮挡第一纹路2111，在获得的各帧测试图像中，第一纹路2111将呈现不同长度。在接下来的步骤中，本发明选取其中一条第一纹路2111a作简要说明。首先针对该条第一纹路2111a的像素值(本实施例中表征为长度值)预设一启动阈值。测试者上半身处于第一识别区211，双腿放置在第二识别区212，身体平落，双手撑地，两腿并拢向后伸，前脚掌着地，身体挺直，摄像机自拍得的第一帧开始提取分析所述第一纹路图像,当首次发现当前帧第一纹路图像中第一纹路2111a的像素值(即本实时例中的长度值)小于预设的启动阈值时，即视作测试者正式进入测试状态，将当前帧设为起始帧，测试开始，同时计时开始。
步骤S3中，分析自起始帧往后的各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐减小变化为逐渐增大时，判定第一纹路图像中纹路的最小像素值是否小于一预设的低位阈值，是则执行步骤S4，否则将第一纹路图像的纹路像素值最小那帧设置为起始帧后重复执行步骤S3。
我们知道，一个完整的俯卧撑包括屈臂下降和撑臂上升的过程。而本发明中，当测试者屈臂下降时，对应地，摄像机持续获取的第一纹路2111a的像素值会呈现逐渐减小态(测试者上半身遮挡第一纹路2111a的程度逐渐增大)；同样，当测试者撑臂上升时，对应地，摄像机持续获取的第一纹路2111a的像素值则会呈现逐渐增大态(测试者上半身遮挡第一纹路2111a的程度逐渐减小)。在本步骤及后续步骤中对第一纹路2111a所作的分析，是对其长度值的分析，但并不应仅限于此。
进一步，在俯卧撑测试中，其屈臂下降要求身体平直下降至肩与肘处于同一水平面，因此本发明预设一个低位阈值，视作测试者达到肩与肘处于同一水平面的判断点。也就是说，低位阈值为测试者保持规范的俯卧撑姿势时身体降到最低位摄像机所能拍到的前述第一纹路2111a的长度最大值。再自起始帧依序对第一纹路2111a做图像处理，当发现第一纹路2111a的长度值存在由逐渐减小变化为逐渐增大的过程时，则进而判定其最小像素值是否小于预设的低位阈值，是则进入步骤S4，若不是，则将长度值最小的那帧设置为起始帧后再重复步骤S3的操作。
步骤S4中，继续分析后续各帧的第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐增大又变化为逐渐减小时，判定第一纹路图像中纹路的最大像素值是否大于一预设的高位阈值，是则执行步骤S5，否则将第一纹路图
像的纹路像素值最大那帧设置为起始帧后返回执行步骤S3。
在俯卧撑测试中，其撑臂上升要求平直撑起至预备姿态为最佳，因此本发明再预设一个高位阈值，视作测试者按照规范俯卧撑姿势时，上半身撑起所能达到的高位判断点。也就是说，高位阈值为测试者保持规范的俯卧撑姿势时测试者身体升到最高位摄像机所能拍到的第一纹路2111a长度的最小值。接续步骤S3，在依序对后续第一纹路2111a做图像处理过程中，当发现第一纹路2111a的长度值存在由逐渐增大变化为逐渐减小时，则进而判定其最大像素值是否大于预设的高位阈值，是则进入步骤S5，若不是，则将最大像素值那帧设置为起始帧后返回步骤S3的操作。
S5、提取自起始帧至第一纹路图像的纹路像素值最大那帧期间所有的第二纹路图像，判定该所有的第二纹路图像中纹路的像素值是否均大于一预设的跪姿阈值，是则计数一次，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后返回执行步骤S3。
此外，在俯卧撑测试中，其要求测试者在测试时，其双腿与地面始终保持一定距离，不能弯曲或是触地。对此，本发明具体通过第二纹路在测试过程中其长度值的变化来判断，此处所指长度值仍等同理解为第二纹路2121在测试图像中的像素值。本发明所述第二纹路2121，在测试者双腿弯曲或触地时，则会至少有一条纹路被双腿完全遮挡住，无法被摄像机获取到，因此对应预设一跪姿阈值，该跪姿阈值视作测试者发生双腿弯曲或触地的情形，在获得的测试图像中则是至少一条第二纹路2121长度值为零。
本步骤中，通过选取其中一条第二纹路2121用以简要说明。具体为，在前述步骤S3至S4的判断期间，即自起始帧至前述第一纹路2111a的最大长
度值那帧期间，只允许第二纹路2121的长度值均大于跪姿阈值(即任意一条短线条的长度均大于零)，才能视作测试者在该期间的俯卧撑动作合格，予以计数一次；如果第二纹路2121的像素值小于预设的跪姿阈值，即自动判定双腿弯曲或触地，此期间的俯卧撑判为无效，不予计数。基于此，本发明可防止测试过程中的作弊现象。
如上所述，本发明能够实现对一个俯卧撑实现自动启动、识别和计数。而测试中，对次数的累计则是本方法所述步骤的重复再现。针对俯卧撑测试的自动结束，本发明分两种情形说明，该两种情形可单独应用也可结合起来应用：
1、针对有规定时间的测试，比如1分钟测试。自起始帧开始计时后，步骤S3和S4中，提取某帧的第一纹路图像之前，先判定该帧所在时刻相对计时开始时刻是否已达到预设的1分钟时长。若是，则停止提取当前帧的第一纹路图像，计时终止，测试结束，并保存已累计的俯卧撑次数；否则继续测试。
2、针对未规定时间的测试。此情况由测试者自行判断是否达到体能极限，并决定是否继续测试。对此，自起始帧开始的测试过程中，持续判定当前帧第一纹路的像素值是否大于启动阈值，若是，则视作测试者有起身动作，计时结束，保存已累计的俯卧撑次数。
在图像处理过程中，有关上述各测试图像中特定纹路的线条长度值的获取，本发明实施例采用如下方法：首先，摄像机在测试前对无测试者在场的测试区域拍摄，获得一模板图像，定位出第一识别区和第二识别区的坐标信息；接着，摄像机在测试过程中持续拍摄测试者在场的测试区域，获得各帧测试图像。然后，在当前帧的图像处理中，根据坐标信息分别提取当前帧测试图像和提前获得的模板图像中线条获得两边缘信息，并将两边缘信息作与运算获得没
有被遮挡的线条的边缘信息；最后，对线条的边缘信息填充即可获取当前帧测试图像中线条的长度值，具体实施例中，如对图3中的测试图像进行处理后可得到如图4所示的的未被测试者遮挡的线条图像。
本发明还提供一种采用前述方法的俯卧撑测试计数系统，如图5所示，其包括：
测试区域210：测试区域210包括但并不限于为测试者提供俯卧撑测试。测试区域210内通过设置第一识别区211和第二识别区212，分别标示出测试者上半身及其双腿所应放置的区域。第一识别区211和第二识别区212其布局与图2所示相同，在此不再赘述。请结合参阅图2所示，第二识别区212内间隔设置有基本沿着测试者俯卧方向延伸的定位线条2122a和定位线条2122b，测试时测试者的双腿分别沿着所述定位线条2122a和定位线条2122b放置。其中，第一识别区211内部包括至少一长线条，第二识别区212内部包括位于两定位线条之间的至少一短线条，长线条与短线条的延伸方向均与测试者俯卧方向基本垂直，线条的长与短是相对而言。特定纹路其图案不受限制，在本发明的可选实施例中，特定纹路还可为均匀分布的曲线、折线、弧线等线状图案，也可以是包括由矩形等几何形状组成的图案、或是明暗相间的几何形状图案组合。
摄像机220：其位置如图2视野所示意，其设置在测试区域的斜上方，用于对第一识别区211和第二识别区212进行图像拍摄，其位置优选设置在与测试者俯卧方向基本垂直的一侧。摄像机220以一定帧率持续拍摄包括测试者、第一识别区211和第二识别区212在内的测试图像，得到第一识别区211的第一纹路图像以及第二识别区212的第二纹路图像。
低位姿态分析模块230：在本模块的分析中，第一纹路的长线条长度值仍可等同理解为像素值，但并不限于此。预设一低位阈值，其作用如前方法所述，不再赘述。在本模块的说明中，通过以其中一条第一纹路来做简要说明。本模块接收第一纹路图像，分析自起始帧往后的各帧第一纹路图像，当发现第一纹路图像中纹路的长度值由逐渐减小变化为逐渐增大时，判定该第一纹路图像中纹路的最小长度值是否小于一预设的低位阈值，是则启动高位姿态分析模块，否则将第一纹路图像的纹路像素值最小那帧设置为起始帧后继续对后续各帧第一纹路图像中纹路的最小像素值是否小于低位阈值进行判断处理。
高位姿态分析模块240，其接收低位姿态分析模块发出的启动信号，并在启动后继续提取并分析后续各帧第一纹路图像。预设一高位阈值，其作用如前方法所述，不再赘述。在继续分析后续各帧第一纹路图像过程中，当发现第一纹路图像中纹路的长度值由逐渐增大又变化为逐渐减小时，判定第一纹路图像中纹路的最大长度值是否大于一预设的高位阈值，是则启动跪姿分析模块，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后启动低位姿态分析模块。
跪姿分析模块250：在本模块的分析中，第二纹路的短线条长度值仍等同理解为像素值，但并不限于此。首先预设跪姿阈值，其作用如前方法所述，不再赘述。在跪姿分析模块250的说明中，通过以第二纹路来做简要说明，当测试者发生双腿弯曲或触地，至少一条第二纹路长度值为零。跪姿分析模块250接收高位姿态分析模块发来的启动信号后启动，其启动后提取自起始帧至前述第一纹路图像的纹路长度值最大那帧期间所有的第二纹路图像，判定该所有的第二纹路图像中纹路的长度值是否均大于一预设的跪姿阈值，是则发出计数指
令，否则将第一纹路图像的纹路长度值最大那帧设置为起始帧后启动低位姿态分析模块对后续各帧第一纹路图像继续分析。
计数模块260：其接收由跪姿分析模块250发出的计数指令后，对俯卧撑测试计数一次。
同样，本系统能够实现对一个俯卧撑动作实现自动识别和计数。测试中，对次数的累计则仅需本系统相关模块重复运行即可，即重复低位姿态分析模块、高位姿态分析模块对第一纹路的处理，以及跪姿分析模块对第二纹路的处理，对计数模块的计数进行累计。
另外，本系统针对测试的自动启动和结束，进一步包括：
计时模块：其在低位姿态分析模块开始工作前，自摄像机拍得的第一帧开始提取分析所述第一纹路图像，当首次发现当前帧第一纹路图像中的长度值小于预设的启动阈值时，将当前帧设为起始帧，计时开始，后续实时记录当前帧相对计时开始时刻的时长数据。
自动结束分两种情形，该两种情形可单独应用也可结合起来应用：
第一种，针对有规定时间的测试，比如1分钟测试。系统还包括一测试结束判定模块，其调取所述计时模块记录的时长数据，并在低位姿态分析模块和高位姿态分析模块提取某帧第一纹路图像之前，先判定该帧所对应的时长数据是否已达到预设时长，是则结束本次测试，否则指示所述低位姿态分析模块或者所述高位姿态分析模块继续测试。
第二种，针对未规定时间的测试。此情况由测试者自行判断是否达到体能极限，并决定是否继续测试。系统还包括一临时终止判定模块，其在低位姿态分析模块和高位姿态分析模块分析当前帧第一纹路图像之前判定当前帧第一
纹路图像中纹路的像素值是否大于预设的启动阈值，是则结束本次测试，否则指示所述低位姿态分析模块或者所述高位姿态分析模块继续测试。
基于上述方法和系统，本发明所实现的俯卧撑的自助测试，可减少人工成本，提高测试效率，给测试者良好的测试体验。
以上所述仅为本发明的较佳实施例而已，并不用以限制本发明，凡在本发明的精神和原则之内所作的任何修改、等同替换和改进等，均应包含在本发明的保护范围之内。
Claims (10)
- 一种俯卧撑测试的计数方法，其特征在于，包括步骤：S1、在测试区域设置各具有特定纹路的第一识别区和第二识别区，所述第一识别区为测试过程中测试者上半身所在区域,所述第二识别区为测试过程中测试者双腿所在区域；在所述测试区域的斜上方设置摄像机；S2、利用所述摄像机以一定帧率持续拍摄包括测试者、第一识别区和第二识别区在内的测试区域图像，得到第一识别区内的第一纹路图像以及第二识别区内的第二纹路图像；S3、分析自起始帧往后的各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐减小变化为逐渐增大时，判定所述第一纹路图像中纹路的最小像素值是否小于一预设的低位阈值，是则执行步骤S4，否则将第一纹路图像的纹路像素值最小那帧设置为起始帧后重复执行步骤S3；S4、继续分析后续各帧的第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐增大又变化为逐渐减小时，判定所述第一纹路图像中纹路的最大像素值是否大于一预设的高位阈值，是则执行步骤S5，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后返回执行步骤S3；S5、提取自起始帧至第一纹路图像的纹路像素值最大那帧期间所有的第二纹路图像，判定该所有的第二纹路图像中纹路的像素值是否均大于一预设的跪姿阈值，是则计数一次，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后返回执行步骤S3。
- 如权利要求1所述的俯卧撑测试的计数方法，其特征在于，步骤S2与S3之间还包括如下步骤，自拍得的第一帧开始提取分析所述第一纹路图像，当首次发现当前帧第一纹路图像中纹路的像素值小于预设的启动阈值时，将当 前帧设为起始帧，计时开始；步骤S3和S4中，提取某帧的第一纹路图像之前，先判定该帧所在时刻相对计时开始时刻是否已达到预设时长，是则结束本次测试，否则继续测试。
- 如权利要求1所述的俯卧撑测试的计数方法，其特征在于，步骤S3和S4中，若发现当前帧第一纹路图像中纹路的像素值大于预设的启动阈值，则结束本次测试。
- 如权利要求1所述的俯卧撑测试的计数方法，其特征在于，所述第一识别区内设置的特定纹路为至少一长线条，该长线条的延伸方向与测试者俯卧方向基本垂直；各步骤中分析的第一纹路图像中纹路的像素值为长线条长度；所述低位阈值为测试者保持规范的俯卧撑姿势时身体降到最低位摄像机所能拍到的长线条长度的最大值；所述高位阈值为测试者保持规范的俯卧撑姿势时测试者身体升到最高位摄像机所能拍到的长线条长度的最小值。
- 如权利要求1所述的俯卧撑测试的计数方法，其特征在于，所述第二识别区内间隔设置有基本沿着测试者俯卧方向延伸的两定位线条，测试时测试者的双腿分别沿着所述两定位线条放置，所述第二识别区设置的特定纹路为至少一短线条，所述短线条设置在两定位线条之间且延伸方向与测试者俯卧方向基本垂直，步骤S5中判定的第二纹路图像中纹路的像素值为短线条长度，所述大于一预设的跪姿阈值为任意一条短线条的长度均大于零。
- 一种俯卧撑测试的计数系统，其特征在于，包括：测试区域，其包括具有特定纹路的第一识别区和第二识别区，所述第一识别区为测试过程中测试者上半身所在区域，所述第二识别区为测试过程中测试者双腿所在区域；摄像机，其设置在所述测试区域的斜上方，该摄像机以一定帧率持续拍摄包括测试者、第一识别区和第二识别区在内的测试区域图像，得到第一识别区的第一纹路图像以及第二识别区的第二纹路图像；低位姿态分析模块，其提取并分析自起始帧往后的各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐减小变化为逐渐增大时，判定所述第一纹路图像中纹路的最小像素值是否小于一预设的低位阈值，是则启动高位姿态分析模块,否则将第一纹路图像的纹路像素值最小那帧设置为起始帧后继续对后续各帧第一纹路图像中纹路的最小像素值是否小于低位阈值进行判断处理；高位姿态分析模块，其启动后继续提取并分析后续各帧第一纹路图像，当发现第一纹路图像中纹路的像素值由逐渐增大又变化为逐渐减小时，判定所述第一纹路图像中纹路的最大像素值是否大于一预设的高位阈值，是则启动跪姿分析模块，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后启动低位姿态分析模块；跪姿分析模块，其启动后提取自起始帧至第一纹路图像的纹路像素值最大那帧期间所有的第二纹路图像，判定该所有的第二纹路图像中纹路的像素值是否均大于一预设的跪姿阈值，是则发出计数指令，否则将第一纹路图像的纹路像素值最大那帧设置为起始帧后启动低位姿态分析模块；计数模块，其接收到所述跪姿分析模块发出的计数指令后，对俯卧撑测试计数一次。
- 如权利要求6所述的俯卧撑测试的计数系统，其特征在于，该系统进一步包括：计时模块，其在所述低位姿态分析模块开始工作前，自拍得的第一帧开始提取分析所述第一纹路图像，当首次发现当前帧第一纹路图像中的像素值小于预设的启动阈值时，将当前帧设为起始帧，计时开始，后续实时记录当前帧相对计时开始时刻的时长数据；测试结束判定模块，其调取所述计时模块记录的时长数据，并在所述低位姿态分析模块和高位姿态分析模块提取某帧第一纹路图像之前，先判定该帧所对应的时长数据是否已达到预设时长，是则结束本次测试，否则指示所述低位姿态分析模块或者所述高位姿态分析模块继续测试。
- 如权利要求6所述的俯卧撑测试的计数系统，其特征在于，该系统进一步包括临时终止判定模块，其在所述低位姿态分析模块以及所述高位姿态分析模块分析当前帧第一纹路图像之前判定当前帧第一纹路图像中纹路的像素值是否大于预设的启动阈值，是则结束本次测试,否则指示所述低位姿态分析模块或者所述高位姿态分析模块继续测试。
- 如权利要求6所述的俯卧撑测试的计数系统，其特征在于，所述第一识别区内设置的特定纹路为至少一长线条，该长线条的延伸方向与测试者俯卧方向基本垂直；所述低位姿态分析模块和高位姿态分析模块分析的第一纹路图像中纹路的像素值为长线条长度；所述低位阈值为测试者保持规范的俯卧撑姿势时身体降到最低位摄像机所能拍到的长线条长度的最大值；所述高位阈值为测试者保持规范的俯卧撑姿势时测试者身体升到最高位摄像机所能拍到的长线条长度的最小值。
- 如权利要求6所述的俯卧撑测试的计数系统，其特征在于，所述第二识别区内间隔设置有基本沿着测试者俯卧方向延伸的两定位线条，测试时测试 者的双腿分别沿着所述两定位线条放置，所述第二识别区设置的特定纹路为至少一短线条，所述短线条设置在两定位线条之间且延伸方向与测试者俯卧方向基本垂直，所述跪姿分析模块判定的第二纹路图像中纹路的像素值为短线条长度，所述大于一预设的跪姿阈值为任意一条短线条长度均大于零。
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610303349.1 | 2016-05-09 | ||
| CN201610303349.1A CN105999655B (zh) | 2016-05-09 | 2016-05-09 | 俯卧撑测试的计数方法及系统 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2017193591A1 true WO2017193591A1 (zh) | 2017-11-16 |
Family
ID=57099951
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/CN2016/112312 WO2017193591A1 (zh) | 2016-05-09 | 2016-12-27 | 俯卧撑测试的计数方法及系统 |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN105999655B (zh) |
| WO (1) | WO2017193591A1 (zh) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105999655B (zh) * | 2016-05-09 | 2018-04-27 | 深圳泰山体育科技股份有限公司 | 俯卧撑测试的计数方法及系统 |
| CN109200565A (zh) * | 2017-06-30 | 2019-01-15 | 深圳泰山体育科技股份有限公司 | 俯卧撑测试的光学标定及测量方法 |
| CN109200563A (zh) * | 2017-06-30 | 2019-01-15 | 深圳泰山体育科技股份有限公司 | 仰卧起坐测试的光学标定及测量方法 |
| CN113158731A (zh) * | 2020-12-31 | 2021-07-23 | 杭州拓深科技有限公司 | 一种俯卧撑计数方法、装置、电子装置和存储介质 |
| CN113599776A (zh) * | 2021-08-05 | 2021-11-05 | 北京理工大学 | 一种实时俯卧撑计数、标准判断方法及系统 |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101468256A (zh) * | 2008-08-13 | 2009-07-01 | 孙学川 | 俯卧撑智能测试仪 |
| KR20140015779A (ko) * | 2012-07-24 | 2014-02-07 | (주)웰텍 | 윗몸일으키기 측정장치 |
| CN104190068A (zh) * | 2014-09-01 | 2014-12-10 | 刘敏 | 俯卧撑测试仪 |
| CN104667510A (zh) * | 2015-02-09 | 2015-06-03 | 深圳泰山在线科技有限公司 | 一种人体动作测试系统 |
| CN104665838A (zh) * | 2015-02-11 | 2015-06-03 | 深圳泰山在线科技有限公司 | 体质检测方法及系统 |
| CN104688233A (zh) * | 2015-02-11 | 2015-06-10 | 深圳泰山在线科技有限公司 | 体质测试机 |
| CN105999655A (zh) * | 2016-05-09 | 2016-10-12 | 深圳泰山体育科技股份有限公司 | 俯卧撑测试的计数方法及系统 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN201519448U (zh) * | 2009-05-09 | 2010-07-07 | 韩洪波 | 一种俯卧撑测试仪 |
| KR101269561B1 (ko) * | 2011-11-30 | 2013-06-04 | 주식회사 바이오스페이스 | 초음파를 이용한 팔굽혀펴기 횟수 측정 장치 및 측정 시스템 |
| CN103426025B (zh) * | 2013-08-23 | 2015-10-28 | 华南理工大学 | 一种基于智能手机平台的非接触式俯卧撑计数方法 |
| CN104537670B (zh) * | 2014-12-31 | 2017-08-04 | 深圳泰山体育科技股份有限公司 | 一种柔韧性测试方法及系统 |
-
2016
- 2016-05-09 CN CN201610303349.1A patent/CN105999655B/zh active Active
- 2016-12-27 WO PCT/CN2016/112312 patent/WO2017193591A1/zh active Application Filing
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101468256A (zh) * | 2008-08-13 | 2009-07-01 | 孙学川 | 俯卧撑智能测试仪 |
| KR20140015779A (ko) * | 2012-07-24 | 2014-02-07 | (주)웰텍 | 윗몸일으키기 측정장치 |
| CN104190068A (zh) * | 2014-09-01 | 2014-12-10 | 刘敏 | 俯卧撑测试仪 |
| CN104667510A (zh) * | 2015-02-09 | 2015-06-03 | 深圳泰山在线科技有限公司 | 一种人体动作测试系统 |
| CN104665838A (zh) * | 2015-02-11 | 2015-06-03 | 深圳泰山在线科技有限公司 | 体质检测方法及系统 |
| CN104688233A (zh) * | 2015-02-11 | 2015-06-10 | 深圳泰山在线科技有限公司 | 体质测试机 |
| CN105999655A (zh) * | 2016-05-09 | 2016-10-12 | 深圳泰山体育科技股份有限公司 | 俯卧撑测试的计数方法及系统 |
Also Published As
| Publication number | Publication date |
|---|---|
| CN105999655A (zh) | 2016-10-12 |
| CN105999655B (zh) | 2018-04-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2017193591A1 (zh) | 俯卧撑测试的计数方法及系统 | |
| WO2017193590A1 (zh) | 仰卧起坐测试的计数方法及系统 | |
| CN102915372B (zh) | 图像检索方法、装置及系统 | |
| JP4770936B2 (ja) | 静脈認証装置およびテンプレート登録方法 | |
| CN105144228B (zh) | 图像处理方法以及信息终端 | |
| WO2020134527A1 (zh) | 人脸识别的方法及装置 | |
| CN106022209A (zh) | 一种基于人脸检测的距离估算和处理的方法及装置 | |
| US9642521B2 (en) | Automatic pupillary distance measurement system and measuring method | |
| TWI733168B (zh) | 基於滑動式採樣的指紋範本註冊方法 | |
| WO2016165107A1 (zh) | 一种采集指纹的方法、指纹采集器以及终端 | |
| CN109002786A (zh) | 人脸检测方法、设备及计算机可读存储介质 | |
| KR20180093917A (ko) | 지문이미지를 형성하기 위한 방법 및 지문감지 시스템 | |
| CN112001334B (zh) | 人像识别装置 | |
| CN110363087A (zh) | 一种长基线双目人脸活体检测方法及系统 | |
| CN105654531B (zh) | 一种图像轮廓的绘制方法及装置 | |
| CN110121377A (zh) | 推移板计分系统 | |
| JP5163281B2 (ja) | 静脈認証装置および静脈認証方法 | |
| CN109729268B (zh) | 一种人脸拍摄方法、装置、设备及介质 | |
| US20230237682A1 (en) | Method and system for binocular ranging for children | |
| CN109492585B (zh) | 一种活体检测方法和电子设备 | |
| KR20130075470A (ko) | 전치태반 판정 장치 및 방법 | |
| CN106056080A (zh) | 一种可视化的生物计量信息采集装置和方法 | |
| CN105956550B (zh) | 一种视频鉴别的方法和装置 | |
| CN112132011B (zh) | 一种面部识别方法、装置、设备及存储介质 | |
| EP3540630A1 (en) | Glint detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 16901546 Country of ref document: EP Kind code of ref document: A1 |
|
| 122 | Ep: pct application non-entry in european phase |
Ref document number: 16901546 Country of ref document: EP Kind code of ref document: A1 |