CN115937967A - Body-building action recognition and correction method - Google Patents
Body-building action recognition and correction method Download PDFInfo
- Publication number
- CN115937967A CN115937967A CN202211274395.5A CN202211274395A CN115937967A CN 115937967 A CN115937967 A CN 115937967A CN 202211274395 A CN202211274395 A CN 202211274395A CN 115937967 A CN115937967 A CN 115937967A
- Authority
- CN
- China
- Prior art keywords
- action
- building
- camera
- standard
- data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Image Analysis (AREA)
Abstract
The invention provides a body-building action recognition and correction method, which belongs to the technical field of body-building action correction and comprises the following steps of setting a body-building video recognition camera unit, acquiring body-building action video data of a standard body-building coach, obtaining a body-building data source through image processing, enabling a body-building person to do body-building actions under the camera unit, acquiring body-building video data through the camera unit, obtaining an outer frame diagram of each action through image processing, comparing an outer frame diagram of each action of the body-building person with the body-building data source to obtain difference data, outputting and displaying the difference data on an upper outer frame of the standard action to perform difference comparison, and outputting a correction suggestion. The camera unit through setting up spatial structure discerns body-builder's action in real time, then compares the action of discernment with the action of standard, has realized automatic correction, and when the comparison, use polar coordinates's contrast mode simultaneously, great reduction because fat etc. of health cause the influence to the action discernment.
Description
Technical Field
The invention relates to the technical field of body-building action correction, in particular to a body-building action recognition correction method.
Background
The body-building field has certain particularity as a branch field of the sports field, in the body-building field, the posture in the body-building process is very important, in the body-building process, the assistance of sports equipment is often needed, and if the posture is not standard, the body is easily damaged. In a traditional mode, an acceleration sensor can be used for acquiring acceleration data acquired by a user in a movement process, a processor performs data processing through the acquired acceleration data acquired by the acceleration sensor and angular velocity data acquired by the angular velocity to determine current movement data, and then the movement data is matched with action types in action input and output equipment to determine whether a current posture is standard or not.
In the above manner, the data acquired by the acceleration sensor can only determine the movement of the exerciser, such as upward and downward movement of the body of the exerciser, but the standard of the movement posture of the exerciser in the exercise process using the exercise equipment cannot be identified, and the exercise guidance value for the exercise of the user is not great. Therefore, it is necessary to design a fitness gesture that can identify whether the gesture during the fitness process meets the standard, and if not, which gesture needs to be corrected for notification.
Disclosure of Invention
The invention aims to provide a body-building action recognition and correction method, which solves the technical problems in the background technology.
In order to achieve the purpose, the technical scheme adopted by the invention is as follows:
a method for identifying and correcting exercise motions, the method comprising the steps of:
step 1: setting a body-building video identification camera unit;
and 2, step: the standard fitness coach does fitness exercise under the camera unit, and the camera unit acquires fitness action video data of the standard fitness coach;
and step 3: carrying out image processing on the acquired body-building action video data of the standard body-building coach to obtain a standard outer frame diagram of each action, and summarizing the standard outer frame diagrams of each action to obtain a body-building data source;
and 4, step 4: the body builder does body building actions under the camera unit, the camera unit collects body building video data, and then image processing is carried out to obtain an outer frame diagram of each action;
and 5: comparing the external block diagram of each action of the exerciser with the exercise data source to obtain difference data;
step 6: and outputting and displaying the difference data, displaying the difference data on an outer frame on the standard action for difference comparison, and then outputting a correction suggestion.
Further, the specific process of step 1 is: the front camera, the back camera, the left camera, the right camera and the upper camera are arranged on a frame type frame, the front camera level is right opposite to the front end of a body builder, the back camera level is right opposite to the back end of the body builder, the left camera level is right opposite to the left end of the body builder, the right camera level is right opposite to the right end of the body builder, the front camera, the back camera, the left camera and the right camera are arranged in a mode capable of sliding up and down, a height sensor of the body builder is arranged on the top end of the frame type frame, after the body builder enters the frame type frame, the height sensor senses the height of the body builder firstly, and then the front camera, the back camera, the left camera and the right camera are controlled to move to half of the height of the body builder.
Further, in step 2, the camera unit collects video data of five directions of the body-building action of the body-building trainer, then the video data is coded and labeled, and meanwhile, time is embedded into the video data.
Further, the specific process of step 3 is to extract the fitness action video data according to the fact that time is a standard symbol, obtain the extracted image data with one time as an extraction condition, perform frame recognition on the images in five directions to obtain frame maps of the standard action in the five directions at each moment, then summarize the frame maps at all times to obtain a total action outer frame map, place each frame map on a two-dimensional coordinate axis, select the outline center point of the action outer frame map to correspond to the origin of the two-dimensional coordinate axis, then set the outline of the action outer frame map to k points, and if k is an integer multiple value greater than 64, obtain the polar coordinates of the points set on the outline as the polar coordinates of the points
n is a positive integer and is greater than or equal to k, r in all polar coordinates n Extracting to obtain polar coordinate function l (k) = (r) 1 ,r 2 ,r 3 …r n ) Then obtaining the polar coordinate function l of all the coded contour line graph sources t (k)＝(r 1 ,r 2 ,r 3 …r n ) T is the number of the coded contour line graphs and is a positive integer, and all polar coordinate functions l are combined t (k)＝(r 1 ,r 2 ,r 3 …r n ) And summarizing to obtain a fitness data source.
Further, the specific process of step 5 is: putting the outline of the outer frame diagram of all the actions of the exerciser on a two-dimensional coordinate axis, corresponding the central point of the outline of the closed loop to the origin of the two-dimensional coordinate axis, setting the outline of the closed loop to be k points, and obtaining the polar coordinates of the points arranged on the outline as k points when k is an integer multiple value larger than 64
n is a positive integer and is greater than or equal to k, r in all polar coordinates n Extracting to obtain a polar coordinate function l' (k) = (r) 1 '，r 2 '，r 3 '…r n ') then l' (k) = (r) 1 '，r 2 '，r 3 '…r n ') and all the source polar coordinate functions l of the encoded contour line pattern t (k)＝(r 1 ,r 2 ,r 3 …r n ) Fourier transform to obtain discrete function l "(k), inverse Fourier transform of discrete function l" (k) to obtain contour acquaintance function, finding out maximum value of function, and when the maximum value is lower than set fixed value, explaining deviation of motion and standard motion, and calculating polar coordinate r 'of motion of body-building person' n Sorting, then standard fitness action polar coordinate r n Sorting and then sorting r of the sorted corresponding serial numbers n And r' n And comparing the difference values to obtain data which is the difference data.
Further, the specific process of step 6 is: outputting the difference data and the corresponding action on a display screen, and summarizing the difference data in five directions to obtain oneA three-dimensional model of motion, when r is in five directions n And r' n When the difference value is positive, it indicates that the exerciser's action is not in place, and the action needs to be extended outwards, if r in five directions is positive n And r' n When the difference comparison value is a negative value, the action exceeds the frame, and the action needs to be moved to the center to realize automatic correction of the action.
Due to the adoption of the technical scheme, the invention has the following beneficial effects:
the invention recognizes the action of the exerciser in real time by arranging the camera unit with the three-dimensional structure, and then compares the recognized action with the standard action, thereby realizing automatic correction, and simultaneously, when the comparison is carried out, the polar coordinate comparison mode is used, thereby greatly reducing the influence on the action recognition caused by obesity of the body and the like.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be described in further detail below with reference to the accompanying drawings by way of examples of preferred embodiments. However, it should be noted that the numerous details set forth in the description are merely intended to provide a thorough understanding of one or more aspects of the present invention, even though such aspects of the invention may be practiced without these specific details.
As shown in fig. 1, a method for recognizing and correcting exercise motions includes the following steps:
step 1: and a body-building video identification camera unit is arranged. The front camera, the back camera, the left camera, the right camera and the upper camera are arranged on a frame type frame, the front camera level is right opposite to the front end of a body builder, the back camera level is right opposite to the back end of the body builder, the left camera level is right opposite to the left end of the body builder, the right camera level is right opposite to the right end of the body builder, the front camera, the back camera, the left camera and the right camera are arranged in a mode capable of sliding up and down, a height sensor of the body builder is arranged on the top end of the frame type frame, after the body builder enters the frame type frame, the height sensor senses the height of the body builder firstly, and then the front camera, the back camera, the left camera and the right camera are controlled to move to half of the height of the body builder. The height sensor is an ultrasonic array unit, and the height of the body builder can be obtained by acquiring the distance from the top of the head of the body builder to ultrasonic waves and then subtracting the measured distance from the frame type frame.
Step 2: the standard fitness coach does fitness exercise under the camera unit, and the camera unit acquires fitness action video data of the standard fitness coach. The camera unit collects video data of five directions of body-building actions of a body-building coach, then the video data are coded and labeled, and time is embedded into the video data. For the collected video data, direction marking is firstly carried out, namely the video collected in the direction is the video collected in the direction, in the later comparison, each direction needs to be compared with the corresponding direction, and then the total difference data is obtained.
And 3, step 3: and carrying out image processing on the acquired body-building action video data of the standard body-building coach to obtain a standard outer frame diagram of each action, and summarizing the standard outer frame diagrams of each action to obtain a body-building data source. Performing frame extraction on body-building action video data according to the condition that time is a standard symbol, obtaining extracted image data by taking the time as an extraction condition, performing frame identification on images in five directions to obtain frame images of standard actions in five directions at each moment, summarizing the frame images at all times to obtain a total action frame diagram, putting each frame image on a two-dimensional coordinate axis, selecting a contour center point of the action frame diagram to correspond to an origin of the two-dimensional coordinate axis, setting the contour of the action frame diagram to be k points, and obtaining a polar coordinate of a point set on the contour as an integral multiple value of more than 64 if the k is an integral multiple value of more than 64
n is a positive integer and is greater than or equal to k, r in all polar coordinates n Extracting to obtain polar coordinate function l (k) = (r) 1 ,r 2 ,r 3 …r n ) Then obtaining the polar coordinate function l of all the coded contour line graph sources t (k)＝(r 1 ,r 2 ,r 3 …r n ) T is the number of the coded contour line graphs and is a positive integer, and all polar coordinate functions l are combined t (k)＝(r 1 ,r 2 ,r 3 …r n ) And summarizing to obtain a fitness data source. Through the memorial polar coordinate processing, the whole processing is realized, namely, the recognized action can not deviate due to the fact that the human body is fat or tall and short.
And 4, step 4: the body builder does body building actions under the camera unit, the camera unit collects body building video data, and then image processing is carried out to obtain an outer frame diagram of each action. The data collected by the process is the same as the specific process of step 2.
And 5: the external block diagram of each action of the exerciser is compared with the exercise data source to obtain difference data. Putting the outline of the outer frame diagram of all the actions of the exerciser on a two-dimensional coordinate axis, corresponding the central point of the outline of the closed loop to the origin of the two-dimensional coordinate axis, setting the outline of the closed loop to be k points, and obtaining the polar coordinates of the points arranged on the outline as k points when k is an integer multiple value larger than 64
n is a positive integer and is greater than or equal to k, r in all polar coordinates n Extracting to obtain polar coordinate function l' (k) = (r) 1 '，r 2 '，r 3 '…r n ') and then k' (k) = (r) 1 '，r 2 '，r 3 '…r n ') and all source polar coordinate functions of the encoded contour line map l t (k)＝(r 1 ,r 2 ,r 3 …r n ) Fourier transform to obtain discrete function l "(k), inverse Fourier transform to obtain profile identification function, finding out maximum value of the function, and describing action and target when the maximum value is lower than set fixed valueThe quasi-motion is deviated, and the polar coordinate r 'of the motion of the exerciser is determined' n Sorted and then standard fitness action polar r n Sorting and then sorting r of the sorted corresponding serial numbers n And r' n And comparing the difference values to obtain data which is the difference data.
After Fourier transform and inverse Fourier transform are carried out on a polar coordinate function, similarity comparison is carried out on two contours, similarity is compared, no matter the size of a human body is fat, identification of actions cannot be influenced, namely, the slope between two points of an action external wheel bank is identified, namely, the contour similarity, whether the actions are in place or not can be identified more accurately, then the maximum value found through the recognition function is that the maximum value is after rotation contrast in all directions, the contrast in the direction is the highest, then the difference value in the direction is taken for comparison, accurate action position positioning contrast is achieved, and action contrast correction is more accurate.
Step 6: and outputting and displaying the difference data, displaying the difference data on an outer frame on the standard action for difference comparison, and then outputting a correction suggestion. Outputting the difference data and the corresponding action on a display screen, summarizing the difference data in five directions to obtain a three-dimensional model of the action, and calculating the difference data in the five directions as r n And r' n When the difference value is positive, it indicates that the exerciser's action is not in place, and the action needs to be extended outwards, if r in five directions is positive n And r' n When the difference comparison value is a negative value, the action exceeds the frame, and the action needs to be moved to the center to realize automatic correction of the action.
The foregoing is only a preferred embodiment of the present invention, and it should be noted that those skilled in the art can make various improvements and modifications without departing from the principle of the present invention, and these improvements and modifications should also be construed as the protection scope of the present invention.
Claims (6)
1. A body-building action recognition and correction method is characterized by comprising the following steps:
step 1: setting a body-building video identification camera unit;
step 2: the standard fitness coach does fitness exercise under the camera unit, and the camera unit acquires fitness action video data of the standard fitness coach;
and step 3: carrying out image processing on the acquired body-building action video data of the standard body-building coach to obtain a standard outer frame diagram of each action, and summarizing the standard outer frame diagrams of each action to obtain a body-building data source;
and 4, step 4: the body builder does body building actions under the camera unit, the camera unit collects body building video data, and then image processing is carried out to obtain an outer frame diagram of each action;
and 5: comparing the external block diagram of each action of the exerciser with the exercise data source to obtain difference data;
step 6: and outputting and displaying the difference data, displaying the difference data on an outer frame on the standard action for difference comparison, and then outputting a correction suggestion.
2. The method for recognizing and correcting exercise motions of claim 1, wherein the method comprises the following steps: the specific process of the step 1 is as follows: the front camera, the rear camera, the left camera, the right camera and the upper camera are installed on a frame-shaped frame, the front camera is horizontally right opposite to the front end of a body builder, the rear camera is horizontally right opposite to the rear end of the body builder, the left camera is horizontally right opposite to the left end of the body builder, the right camera is horizontally right opposite to the right end of the body builder, the front camera, the rear camera, the left camera and the right camera are arranged in a vertically sliding mode, a body builder height sensor is arranged at the top end of the frame-shaped frame, after the body builder enters the frame-shaped frame, the height sensor firstly senses the height of the body builder, and then the front camera, the rear camera, the left camera and the right camera are controlled to move to a half of the height of the body builder.
3. A method of fitness motion recognition and correction according to claim 2, wherein: in step 2, the camera unit collects video data of five directions of body-building actions of the body-building coach, then the video data are coded and labeled, and time is embedded into the video data.
4. A method of fitness motion recognition and correction according to claim 3, wherein: the specific process of the step 3 is that the body-building action video data is subjected to extraction according to the fact that time is a standard symbol, one time is an extraction condition, extracted image data is obtained, images in five directions are subjected to frame recognition, frame diagrams of standard actions in five directions at all times are obtained, then the frame diagrams at all times are collected to obtain a total action outer frame diagram, each frame diagram is placed on a two-dimensional coordinate axis, the outline center point of the action outer frame diagram corresponds to the origin of the two-dimensional coordinate axis, then the outline of the action outer frame diagram is set to be k points, the k is an integral multiple value larger than 64, and the polar coordinates of the points arranged on the outline are obtained as
n is a positive integer and is greater than or equal to k, r in all polar coordinates n Extracting to obtain polar coordinate function l (k) = (r) 1 ,r 2 ,r 3 ···r n ) Then, the polar coordinate function l of all the coding contour line graph sources is obtained t (k)＝(r 1 ,r 2 ,r 3 ···r n ) T is the number of the coded contour line graphs and is a positive integer, and all polar coordinate functions l are combined t (k)＝(r 1 ,r 2 ,r 3 ···r n ) And summarizing to obtain a fitness data source.
5. The method for correcting exercise movement recognition according to claim 4, wherein: the specific process of the step 5 is as follows: putting the outline of the outer frame diagram of all actions of the exerciser on a two-dimensional coordinate axis, corresponding the central point of the outline of the closed loop to the origin of the two-dimensional coordinate axis, setting the outline of the closed loop to be k points, and if k is an integral multiple value greater than 64, obtaining the polar coordinates of the points arranged on the outline as
n is a positive integer and is greater than or equal to k, r in all polar coordinates n Extracting to obtain polar coordinate function l' (k) = (r) 1 ’，r 2 ’，r 3 ’…r n ') then l' (k) = (r) 1 ’，r 2 ’，r 3 ’…r n ') and all source polar coordinate functions of the encoded contour line map l t (k)＝(r 1 ,r 2 ,r 3 ···r n ) Fourier transform to obtain discrete function l ' (k), inverse Fourier transform to obtain contour degree of acquaintance function, finding out maximum value of function, and describing deviation between motion and standard motion when maximum value is lower than set fixed value, and calculating polar coordinate r ' of motion of exerciser ' n Sorted and then standard fitness action polar r n Sorting and then sorting r of the sorted corresponding serial numbers n And r' n And comparing the difference values to obtain data which is the difference data.
6. A method of fitness motion recognition and correction according to claim 5, wherein: the specific process of step 6 is as follows: outputting the difference data and the corresponding action on a display screen, summarizing the difference data in five directions to obtain a three-dimensional model of the action, and calculating the difference data in the five directions as r n And r' n When the difference value is positive, it indicates that the exerciser's action is not in place, and the action needs to be extended outwards, if r in five directions is positive n And r' n When the value of the difference comparison is a negative value, the action exceeds the frame, and the action needs to be moved to the center to realize the automatic correction of the action.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274395.5A CN115937967A (en) | 2022-10-18 | 2022-10-18 | Body-building action recognition and correction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211274395.5A CN115937967A (en) | 2022-10-18 | 2022-10-18 | Body-building action recognition and correction method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115937967A true CN115937967A (en) | 2023-04-07 |
Family
ID=86552977
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211274395.5A Pending CN115937967A (en) | 2022-10-18 | 2022-10-18 | Body-building action recognition and correction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115937967A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116631054A (en) * | 2023-04-27 | 2023-08-22 | 成都体育学院 | Motion correction method, system, electronic equipment and medium |
-
2022
- 2022-10-18 CN CN202211274395.5A patent/CN115937967A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN116631054A (en) * | 2023-04-27 | 2023-08-22 | 成都体育学院 | Motion correction method, system, electronic equipment and medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109863535B (en) | Motion recognition device, storage medium, and motion recognition method | |
| CN103530599B (en) | The detection method and system of a kind of real human face and picture face | |
| Chen et al. | Computer-assisted self-training system for sports exercise using kinects | |
| CN111931701A (en) | Gesture recognition method and device based on artificial intelligence, terminal and storage medium | |
| Chen et al. | Yoga posture recognition for self-training | |
| CN113762133A (en) | Self-weight fitness auxiliary coaching system, method and terminal based on human body posture recognition | |
| WO2017161734A1 (en) | Correction of human body movements via television and motion-sensing accessory and system | |
| CN111401260B (en) | Sit-up test counting method and system based on Quick-OpenPose model | |
| CN105335725A (en) | Gait identification identity authentication method based on feature fusion | |
| CN113516005B (en) | Dance action evaluation system based on deep learning and gesture estimation | |
| CN111368791A (en) | Pull-up test counting method and system based on Quick-OpenPose model | |
| CN109117753A (en) | Position recognition methods, device, terminal and storage medium | |
| CN113398556B (en) | Push-up identification method and system | |
| CN109274883A (en) | Posture antidote, device, terminal and storage medium | |
| JP6381368B2 (en) | Image processing apparatus, image processing method, and program | |
| CN115482580A (en) | Multi-person evaluation system based on machine vision skeletal tracking technology | |
| CN115937967A (en) | Body-building action recognition and correction method | |
| CN103310193B (en) | A kind of method that records sportsman's important technology action moment in gymnastics video | |
| Yang et al. | Human exercise posture analysis based on pose estimation | |
| CN109993116B (en) | Pedestrian re-identification method based on mutual learning of human bones | |
| CN113239849B (en) | Body-building action quality assessment method, body-building action quality assessment system, terminal equipment and storage medium | |
| CN116266415A (en) | Action evaluation method, system and device based on body building teaching training and medium | |
| CN115331153B (en) | Posture monitoring method for assisting vestibule rehabilitation training | |
| CN114639168B (en) | Method and system for recognizing running gesture | |
| CN115690902A (en) | Abnormal posture early warning method for body building action |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |