CN115311610A - Method for recognizing abnormity of fitness equipment - Google Patents

Method for recognizing abnormity of fitness equipment Download PDF

Info

Publication number
CN115311610A
CN115311610A CN202211245039.0A CN202211245039A CN115311610A CN 115311610 A CN115311610 A CN 115311610A CN 202211245039 A CN202211245039 A CN 202211245039A CN 115311610 A CN115311610 A CN 115311610A
Authority
CN
China
Prior art keywords
fitness equipment
identified
fitness
degree
video
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.)
Granted
Application number
CN202211245039.0A
Other languages
Chinese (zh)
Other versions
CN115311610B (en
Inventor
季勇康
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiangsu Yatai Fitness Co ltd
Original Assignee
Jiangsu Yatai Fitness Co ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Jiangsu Yatai Fitness Co ltd filed Critical Jiangsu Yatai Fitness Co ltd
Priority to CN202211245039.0A priority Critical patent/CN115311610B/en
Publication of CN115311610A publication Critical patent/CN115311610A/en
Application granted granted Critical
Publication of CN115311610B publication Critical patent/CN115311610B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/40—Scenes; Scene-specific elements in video content
    • G06V20/49—Segmenting video sequences, i.e. computational techniques such as parsing or cutting the sequence, low-level clustering or determining units such as shots or scenes
    • 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
    • G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
    • 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

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Human Computer Interaction (AREA)
  • Health & Medical Sciences (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • General Health & Medical Sciences (AREA)
  • Psychiatry (AREA)
  • Social Psychology (AREA)
  • Computing Systems (AREA)
  • Alarm Systems (AREA)

Abstract

The invention relates to a method for recognizing the abnormity of fitness equipment, belonging to the technical field of abnormity recognition of the fitness equipment. The method comprises the following steps: acquiring an image video of a target area in a set time period, wherein the target area comprises fitness equipment to be identified; dividing the image video according to the time index to obtain K image sub-videos; fitting according to the use heat of the fitness equipment to be identified corresponding to each image sub-video to obtain a use heat change function corresponding to the image video; obtaining the use duration and the nonstandard action degree corresponding to each fitness person using the fitness equipment to be identified according to the image video; calculating the corresponding non-standard use degree of the fitness equipment to be identified; and calculating the abnormal degree of the fitness equipment to be identified according to the use heat change function and the irregular use degree. The invention belongs to an automatic identification method, which does not depend on management personnel to carry out abnormity identification, and effectively improves the efficiency of abnormity detection on fitness equipment.

Description

Method for recognizing abnormity of fitness equipment
Technical Field
The invention relates to the technical field of body-building equipment abnormity identification, in particular to a method for body-building equipment abnormity identification.
Background
With the development of society, more and more people start to go to the gymnasium to build up body, and the exercise using the fitness equipment can be carried out without the restriction of weather, and the fitness equipment has gradually become the exercise mode of most sporters. However, exercise equipment may be at risk of abnormalities for some reason, such as those caused by natural wear due to the exercise equipment being used for too long a period of time; or damage to the exercise equipment caused by irregular motions of the exerciser during the exercise process, such as overstretching the pull rod when the exerciser uses the rowing machine, overstretching the chest-pushing exerciser when the sitting type chest-pushing exerciser is used, and the like.
The safety problem of the fitness equipment of the fitness room is vital for fitness personnel, the fitness equipment of the fitness room is managed by management personnel at present, and related management personnel can only troubleshoot and inform maintenance personnel to repair the fitness equipment when finding that the fitness equipment is abnormal. For various fitness equipment in a gymnasium, a large amount of manpower and material resources are consumed during manual detection, each fitness equipment cannot be detected one by one, the detection effect is poor, the subjectivity of the manual detection is strong, and the conditions of wrong detection and missed detection are easy to occur.
Disclosure of Invention
The invention aims to provide a method for identifying the abnormity of fitness equipment, which is used for solving the problem that the existing method for detecting the abnormity of the fitness equipment by using a manager is low in efficiency.
In order to solve the above problems, the technical solution of the method for identifying an abnormality of a fitness equipment of the present invention includes the following steps:
acquiring an image video of a target area in a set time period, wherein the target area comprises fitness equipment to be identified;
dividing the image video according to the time index to obtain K image sub-videos; fitting according to the using heat of the fitness equipment to be identified corresponding to each image sub-video to obtain a using heat change function corresponding to the image video, wherein K is more than or equal to 2;
obtaining the use duration and the nonstandard action degree corresponding to each fitness person using the fitness equipment to be identified according to the image video; calculating the nonstandard use degree corresponding to the fitness equipment to be identified according to the use duration and the action nonstandard degree corresponding to each fitness person;
and calculating the abnormal degree of the fitness equipment to be identified according to the use heat change function and the irregular use degree.
The beneficial effects of the invention are: the method obtains the use heat change function and the non-standard use degree of the fitness equipment to be identified in the set time period based on the acquired image video, and the use heat change function can reflect whether the use heat of the fitness equipment to be identified is obviously reduced by a user, and the obvious reduction is probably caused by the abnormity of the fitness equipment to be identified; the unnormalized use degree can reflect the damage degree of the fitness equipment to be identified by the fitness person, and the higher the unnormalized use degree is, the higher the damage degree of the fitness equipment to be identified is; the abnormal degree of the fitness equipment to be identified is calculated based on the heat change function and the irregular use degree, the method belongs to an automatic identification method, and the abnormal identification of the fitness equipment is not carried out by a manager, so that the efficiency of abnormal detection of the fitness equipment is effectively improved.
Further, the method for calculating the use heat of the fitness equipment to be identified corresponding to each image sub-video comprises the following steps:
marking the fitness person and the fitness equipment to be identified in each image sub-video by using the surrounding frame to obtain the surrounding frame of the fitness person and the surrounding frame of the fitness equipment to be identified;
judging whether the exerciser uses the fitness equipment to be identified or not according to the intersection ratio of the exerciser enclosure frame and the fitness equipment enclosure frame to be identified;
counting the use times and the single use duration of the fitness equipment to be identified in each image sub-video;
and calculating the use heat of the fitness equipment to be identified corresponding to each image sub-video according to the use times and the single use duration of the fitness equipment to be identified corresponding to each image sub-video.
Further, the use heat of the fitness equipment to be identified corresponding to each image sub-video is calculated by the following formula:
Figure 100002_DEST_PATH_IMAGE001
wherein k is the kth image sub-video,
Figure 533253DEST_PATH_IMAGE002
the using heat of the fitness equipment to be identified corresponding to the k image sub-video,
Figure 63460DEST_PATH_IMAGE003
the number of uses corresponding to the k-th image sub-video,
Figure 331893DEST_PATH_IMAGE004
the time length average value of the single use corresponding to the k image sub-video,
Figure 404891DEST_PATH_IMAGE005
in order to use the corresponding weight of the times,
Figure 70052DEST_PATH_IMAGE006
the weight is corresponding to the average single-use duration.
Further, the method for calculating the degree of the action unnormality comprises the following steps:
identifying key points of a body builder using fitness equipment to be identified in the image video;
obtaining a fitness action sequence of a fitness person using the fitness equipment to be identified according to the human body key points;
and comparing the fitness action sequence of the exerciser using the fitness equipment to be identified with the standard fitness action sequence to obtain the nonstandard action degree corresponding to the exerciser using the fitness equipment to be identified.
Calculating the corresponding non-standard use degree of the fitness equipment to be identified by using the following formula:
Figure 962921DEST_PATH_IMAGE007
)
Figure 736842DEST_PATH_IMAGE008
wherein the content of the first and second substances,
Figure 100002_DEST_PATH_IMAGE009
the body-building equipment to be identified corresponds to the nonstandard use degree, R is the total number of the body-building persons using the body-building equipment to be identified, R is the R th body-building person using the body-building equipment to be identified, J is the length of a body-building action sequence, I is the number of key points of the human body, I is the key point of the ith human body, J is the jth body-building action in the body-building action sequence,
Figure 929926DEST_PATH_IMAGE010
the abscissa of the ith personal key point in the jth body-building action corresponding to the body-building person,
Figure 537625DEST_PATH_IMAGE011
is the abscissa of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 215993DEST_PATH_IMAGE012
is the ordinate of the ith personal key point in the jth body-building action corresponding to the body-building person,
Figure 668972DEST_PATH_IMAGE013
is the ordinate of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 247720DEST_PATH_IMAGE014
is the vertical coordinate of the ith personal body key point in the jth body-building action corresponding to the body-building person,
Figure 26321DEST_PATH_IMAGE015
is the vertical coordinate of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 424941DEST_PATH_IMAGE016
to the extent that the motion of the r-th exerciser using the exercise apparatus to be identified is not standardized,
Figure 71823DEST_PATH_IMAGE017
is the r < th >The length of time that an exerciser uses the exercise equipment to be identified.
Further, the method for calculating the abnormal degree of the fitness equipment to be identified according to the usage heat change function and the irregular usage degree comprises the following steps:
judging the change trend of the use heat of the fitness equipment to be identified according to the use heat change function; if the change trend of the use heat of the fitness equipment to be identified is increased, calculating the abnormal degree of the fitness equipment to be identified according to the following formula:
Figure 380444DEST_PATH_IMAGE018
wherein the content of the first and second substances,
Figure 687536DEST_PATH_IMAGE019
to identify the degree of abnormality of the exercise apparatus,
Figure 448818DEST_PATH_IMAGE009
for the nonstandard use degree corresponding to the fitness equipment to be identified,
Figure 899391DEST_PATH_IMAGE020
for the length of time the exercise apparatus has been used to be identified,
Figure 921574DEST_PATH_IMAGE021
in order not to specify the weight corresponding to the degree of use,
Figure 41977DEST_PATH_IMAGE022
the weight corresponding to the used time length.
Further, if the variation trend of the use heat of the fitness equipment to be identified is decreased progressively and the difference between the use heat of the fitness equipment of the same type and the use heat of the fitness equipment to be identified is greater than the set use heat threshold, calculating the abnormal degree of the fitness equipment to be identified according to the following formula:
Figure 680768DEST_PATH_IMAGE023
wherein the content of the first and second substances,
Figure 544819DEST_PATH_IMAGE024
the difference between the use heat of the fitness equipment of the same type and the use heat of the fitness equipment to be identified,
Figure 922973DEST_PATH_IMAGE025
is composed of
Figure 479856DEST_PATH_IMAGE024
The corresponding weight.
Drawings
FIG. 1 is a flow chart of a method for fitness equipment anomaly identification of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be described below with reference to the drawings in the embodiments of the present invention.
The embodiment aims to solve the problem that the efficiency of detecting the abnormality of the fitness equipment by using a manager is low, and as shown in fig. 1, the method for identifying the abnormality of the fitness equipment comprises the following steps:
(1) Acquiring an image video of a target area in a set time period, wherein the target area comprises fitness equipment to be identified;
in order to identify the abnormal conditions of all the fitness equipments in the gymnasium, a plurality of cameras are installed in the gymnasium to acquire the image videos related to all the fitness equipments in the gymnasium. The arrangement of the number of the cameras and the positions of the cameras can be adjusted according to the actual situation of the gymnasium, and in order to realize the abnormal recognition of each fitness equipment in the gymnasium, the acquisition range of the cameras needs to contain each fitness equipment area in the gymnasium.
After the camera is installed, the image video of each fitness equipment area is collected, and therefore the image video of a certain time duration corresponding to each fitness equipment can be obtained. The embodiment of the present invention will next describe the identification method of the embodiment by taking the length of the image video of one of the fitness apparatuses as K days as an example.
(2) Dividing the image video according to the time index to obtain K image sub-videos; fitting according to the using heat of the fitness equipment to be identified corresponding to each image sub-video to obtain a using heat change function corresponding to the image video, wherein K is more than or equal to 2;
in consideration of the fact that under normal conditions, when the fitness equipment of the fitness room is abnormal, the use heat of the exerciser will be obviously reduced, and therefore, the use heat condition of the fitness equipment to be identified is taken as one of the bases for identification in the embodiment. The process of obtaining the usage heat variation function of the fitness equipment to be identified in the embodiment is as follows:
the method comprises the steps of detecting fitness equipment and a fitness person by adopting a target detection network, marking the bounding boxes of the fitness equipment and the fitness person in an image by using label data of the target detection network, marking coordinates, width, height and type of the center point of the bounding box, namely (x, y, w, h, class), wherein x is the horizontal coordinate of the center of the bounding box, y is the vertical coordinate of the center of the bounding box, w is the width of the bounding box, h is the height of the bounding box, and class is the category of the bounding box. And the target detection network adopts a mean square error loss function to carry out iterative training.
After the target in each frame of image is detected, the position relation of the fitness equipment and the exerciser is further analyzed: when the intersection ratio of the surrounding frame of the fitness equipment and the surrounding frame of the exerciser in the image
Figure 340365DEST_PATH_IMAGE026
When the user needs to use the fitness equipment, the fitness equipment is considered to be in a used state; when the intersection ratio of the surrounding frame of the fitness equipment and the surrounding frame of the exerciser in the image
Figure 601582DEST_PATH_IMAGE027
And if so, the fitness equipment is considered to be in an idle state. In practical application, the comparison value of the intersection ratio of the surrounding frames can be adjusted according to the type of the fitness equipment to be identified.
For body-building to be identifiedThe K-day image video of the equipment is divided according to the time sequence to obtain K image sub-videos, each image sub-video corresponds to the image video of the fitness equipment to be identified in one day, and the image sub-video corresponding to the kth day corresponds to the image sub-video of the fitness equipment to be identified in one day. Counting the usage frequency of the k day
Figure 739302DEST_PATH_IMAGE003
And obtaining the duration sequence of single use of the fitness equipment on the k day
Figure 591721DEST_PATH_IMAGE028
Calculate the average of the duration of single use on day k:
Figure 283733DEST_PATH_IMAGE029
(ii) a The formula for calculating the use heat of the fitness equipment based on the use frequency and the single-use duration average is as follows:
Figure 378335DEST_PATH_IMAGE030
wherein k is the kth image sub-video,
Figure 370562DEST_PATH_IMAGE002
the using heat of the fitness equipment to be identified corresponding to the k image sub-video,
Figure 862723DEST_PATH_IMAGE003
the number of uses corresponding to the k-th image sub-video,
Figure 697824DEST_PATH_IMAGE031
is the single-use time length average value corresponding to the kth image sub-video,
Figure 707368DEST_PATH_IMAGE005
in order to use the corresponding weight of the times,
Figure 413156DEST_PATH_IMAGE006
the weight is corresponding to the average single-use duration. Considering that the average value of the duration of a single use of the fitness equipment can reflect the abnormal condition of the fitness equipment better, the setting of the embodiment is
Figure 482743DEST_PATH_IMAGE032
The weights can be modified according to requirements in the actual application process.
Therefore, the use heat sequence within K days corresponding to the fitness equipment to be identified can be obtained
Figure 41025DEST_PATH_IMAGE033
. And fitting the data in the use heat sequence data to obtain a corresponding use heat change function. The fitting process of this embodiment specifically includes:
firstly, randomly selecting M using heat data from a sequence, fitting based on selected M points by taking time as an x axis and using heat as a y axis, marking an obtained curve as a curve 1, then calculating the distance from all using heat data to the curve 1, setting a distance threshold value D, judging that the using heat data belongs to the curve 1 if the distance is less than the threshold value D, and recording the number C1 of the using heat data belonging to the curve 1;
selecting M using heat data from the rest using heat data, similarly fitting according to the M using heat data which are selected again, marking the obtained curve as a curve 2, then calculating the distance from all using heat data to the curve 2, if the distance is less than a threshold value D, judging that the data belongs to the curve 2, and counting the number C2 of the using heat data which belong to the curve 2;
repeating the steps until all the heat characteristic data in the sequence are selected, and fitting the heat characteristic data to obtain the heat characteristic data
Figure 713315DEST_PATH_IMAGE034
The curve is used for counting the number of the use heat data belonging to each curve to obtain a corresponding use heat data number sequence { C1, C2, \ 8230;, C
Figure 680134DEST_PATH_IMAGE034
And (5) taking a curve corresponding to the maximum numerical value in the quantitative sequence as a final fitting function, and taking the curve as a use heat variation function corresponding to the fitness equipment to be identified
Figure 514098DEST_PATH_IMAGE035
。
The embodiment adopts the idea of batch fitting when fitting the data in the heat sequence data to obtain a more accurate fitting curve; as another embodiment, the data in the heat sequence data as a whole may also be fitted in a conventional manner.
Obtaining the corresponding use heat change function of the fitness equipment to be identified
Figure 199157DEST_PATH_IMAGE035
Then, further obtain
Figure 675138DEST_PATH_IMAGE035
First derivative function of
Figure 854053DEST_PATH_IMAGE036
According to
Figure 999864DEST_PATH_IMAGE036
Analyzing the use heat variation trend of the fitness equipment to be identified: when in use
Figure 31274DEST_PATH_IMAGE037
When the user needs to use the fitness equipment to be identified, the user can use the fitness equipment to be identified; when in use
Figure 920732DEST_PATH_IMAGE038
In the time, the using heat of the fitness equipment to be identified is decreased progressively, which indicates that the fitness person does not use the fitness equipment to be identified frequently as before.
The reason why the use heat of the fitness equipment to be identified is gradually decreased is that the interest of the fitness equipment to be identified is possibly reduced besides the abnormality of the fitness equipment to be identified;to eliminate the latter reason, the present embodiment is described in
Figure 721198DEST_PATH_IMAGE038
For example, only when the difference between the use heat corresponding to the last sub-image video of the fitness equipment of the same type and the use heat corresponding to the last image video of the fitness equipment to be identified is larger than a certain value, the reason that the use heat of the fitness equipment to be identified is decreased is judged to be that the fitness equipment to be identified is abnormal, and the larger the difference is, the higher the abnormal degree is.
(3) Obtaining the use duration and the nonstandard action degree corresponding to each fitness person using the fitness equipment to be identified according to the image video; calculating the nonstandard use degree corresponding to the fitness equipment to be identified according to the use duration and the action nonstandard degree corresponding to each fitness person;
considering that the irregular exercise posture, the exercise action and the like of the exerciser may also affect the exercise equipment when the exerciser uses the exercise equipment to perform exercise, in this embodiment, when the abnormal condition of the exercise equipment is identified, the irregular use degree corresponding to the exercise equipment to be identified is further calculated according to the use duration and the irregular action degree corresponding to each exerciser, and the specific process is as follows:
the method comprises the steps of detecting key points of a human body in an image video through a trained key point detection network, wherein the key points comprise head key points, neck key points, left and right shoulder joint points, left and right elbow joints, left and right wrist joints, spine center points, left and right hip joints, left and right knee joints and left and right ankle joints. After the key points of each exerciser are obtained, in order to distinguish the key points among the exercisers, the embodiment matches the key points of the human body by combining the relationship vector spectrums Part Affinity Fields (PAFs) to connect the corresponding key points of each exerciser. Detecting key points of a human body by using a key point detection network and matching key points of the human body by using PAFs are prior art and are not described herein again.
Based on the judgment method for whether the fitness equipment to be identified is used in the step (2), the two-dimensional key point information corresponding to the fitness person using the fitness equipment to be identified can be obtained; in order to facilitate subsequent analysis of the nonstandard degree of the exercise motions of the exerciser using the exercise equipment to be identified, the embodiment employs the TCN network model to obtain the three-dimensional motion sequence corresponding to the two-dimensional key point. The process of obtaining a three-dimensional action sequence by using a TCN network is prior art and will not be described herein.
In order to analyze the non-standard degree of the body-building action of the body-building person using the body-building apparatus to be identified, the three-dimensional action sequence of the body-building person is compared and analyzed with the standard body-building action in the body-building action simulator, so as to obtain the non-standard degree of the body-building action of the body-building person. The formula for specifically calculating the degree of irregularity of the exercise movement in this embodiment is as follows:
Figure 303489DEST_PATH_IMAGE039
wherein r is the r-th exerciser using the fitness equipment to be identified, J is the length of the fitness action sequence, I is the number of key points of the human body, I is the ith key point of the human body, J is the jth fitness action in the fitness action sequence,
Figure 556616DEST_PATH_IMAGE010
is the abscissa of the ith human body key point in the jth body-building action corresponding to the body-building person,
Figure 875864DEST_PATH_IMAGE011
is the abscissa of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 671782DEST_PATH_IMAGE012
is the ordinate of the ith personal key point in the jth body-building action corresponding to the body-building person,
Figure 549608DEST_PATH_IMAGE013
is the ordinate of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 899818DEST_PATH_IMAGE014
is the vertical coordinate of the ith personal body key point in the jth body-building action corresponding to the body-building person,
Figure 521292DEST_PATH_IMAGE015
is the vertical coordinate of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 171716DEST_PATH_IMAGE016
the action of the body-building person using the body-building equipment to be identified is irregular for the r-th person.
It should be noted that, when comparing and analyzing the three-dimensional motion sequence of the exerciser with the standard exercise motions in the exercise motion simulator, not all the human body key points of the exerciser using the exercise equipment to be identified are compared, but the comparison and analysis are performed according to the related human body key points which may cause damage to the exercise equipment to be identified, for example, when the related human body key points which may cause damage to the exercise equipment to be identified are mainly leg key points, only the leg key points need to be compared.
Obtaining the action non-standard degree corresponding to each body-building person using the body-building equipment to be identified in the image video, and obtaining the action non-standard degree sequence corresponding to the body-building equipment to be identified
Figure 954864DEST_PATH_IMAGE040
And R is the total number of the fitness users using the fitness equipment to be identified in the image video. Considering that the longer the exercise time of the exerciser with the high irregular exercise action degree is, the greater the damage to the exercise equipment is, in this embodiment, the irregular use degree corresponding to the exercise equipment to be identified corresponding to the image video is calculated by using the following formula:
Figure 692837DEST_PATH_IMAGE041
)
wherein the content of the first and second substances,
Figure 727789DEST_PATH_IMAGE009
for the corresponding non-standard use degree of the fitness equipment to be identified,
Figure 357354DEST_PATH_IMAGE017
for the length of time the r-th exerciser used the exercise apparatus to be identified.
(4) And calculating the abnormal degree of the fitness equipment to be identified according to the use heat degree change function and the irregular use degree.
Based on the step (2), the use heat change function of the fitness equipment to be identified can be obtained
Figure 452349DEST_PATH_IMAGE035
Based on using heat variation functions
Figure 167364DEST_PATH_IMAGE035
Whether the using heat of the fitness equipment to be identified is reduced by the exerciser can be judged, and whether the fitness equipment to be identified is abnormal or not is judged by combining the using heat corresponding to the similar fitness equipment when the using heat is reduced.
The non-standard use degree of the fitness equipment to be identified can be obtained based on the step (3), and the damage size caused by the fitness equipment to be identified in the process of using the fitness equipment to be identified by the exerciser can be judged based on the non-standard use degree. In addition, in consideration of the influence of natural normal wear of the fitness equipment to be identified on the equipment, the embodiment also refers to the used time length of the fitness equipment to be identified, namely the time interval from the production date or the date of starting use of the fitness equipment to be identified to the identification time.
In view of the above considerations, the method of the present embodiment for calculating the degree of abnormality of the fitness equipment to be identified is as follows:
this embodiment is described in
Figure 6007DEST_PATH_IMAGE037
When the change trend of the use heat of the fitness equipment to be identified is judged to be increased, the change trend is calculated according to the following formulaDegree of abnormality of the fitness equipment:
Figure 490078DEST_PATH_IMAGE042
wherein the content of the first and second substances,
Figure 382073DEST_PATH_IMAGE019
to identify the degree of abnormality of the exercise apparatus,
Figure 194171DEST_PATH_IMAGE009
for the corresponding non-standard use degree of the fitness equipment to be identified,
Figure 695560DEST_PATH_IMAGE020
for the length of time the exercise apparatus has been used to be identified,
Figure 175082DEST_PATH_IMAGE021
in order not to specify the weight corresponding to the degree of use,
Figure 736514DEST_PATH_IMAGE022
the weight corresponding to the used time length; this example arrangement
Figure 301487DEST_PATH_IMAGE043
，
Figure 606567DEST_PATH_IMAGE044
In practical applications, the weights may be adjusted according to practical situations.
This is implemented in
Figure 563765DEST_PATH_IMAGE038
When the abnormal degree of the fitness equipment to be identified is calculated according to the following formula:
Figure 171464DEST_PATH_IMAGE023
wherein, the first and the second end of the pipe are connected with each other,
Figure 82788DEST_PATH_IMAGE024
the difference between the use heat of the fitness equipment of the same type and the use heat of the fitness equipment to be identified,
Figure 66925DEST_PATH_IMAGE025
is composed of
Figure 380094DEST_PATH_IMAGE024
A corresponding weight; the present embodiment takes into consideration
Figure 158694DEST_PATH_IMAGE038
And the probability of abnormity of the fitness equipment to be identified is higher when the difference value is greater than the set use heat threshold value, and setting
Figure 557315DEST_PATH_IMAGE025
=0.7，
Figure 440082DEST_PATH_IMAGE045
，
Figure 748704DEST_PATH_IMAGE046
. In practical applications, the weights may be adjusted according to practical situations.
Normalizing the calculated abnormal degree of the fitness equipment to be identified, comparing the normalized value with a set abnormal threshold value, and if the normalized value is smaller than the set abnormal threshold value, judging that the fitness equipment to be identified is abnormal temporarily; if the value after the normalization processing is larger than or equal to the set abnormal threshold value, the body-building equipment to be identified is judged to be abnormal or very easy to be abnormal, and managers can be reminded to pay key attention. In this embodiment, the abnormal threshold is set to 0.7, and the set abnormal threshold can be adjusted according to specific requirements in practical application.
In the embodiment, the used time length of the fitness equipment to be identified is also considered when the abnormal degree of the fitness equipment to be identified is calculated, and as other implementation modes, only the use heat and the nonstandard use degree factors of the fitness equipment to be identified can be considered.
The method and the device have the advantages that the using heat change function and the nonstandard using degree of the fitness equipment to be identified in the set time period are obtained based on the obtained image video, whether the using heat of the fitness equipment to be identified is obviously reduced by the using heat change function can be reflected, and the obvious reduction is probably caused by the fact that the fitness equipment to be identified is abnormal; the unnormalized use degree can reflect the damage degree of the fitness equipment to be identified by the exerciser, and the higher the unnormalized use degree is, the greater the damage degree of the fitness equipment to be identified is; the abnormal degree of the fitness equipment to be identified is calculated based on the heat change function and the nonstandard use range, the method belongs to an automatic identification method, abnormal identification of the fitness equipment is not needed by a manager, and the efficiency of abnormal detection of the fitness equipment is effectively improved.
It should be noted that while the preferred embodiments of the present invention have been described, additional variations and modifications to these embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the invention.

Claims (7)

1. A method for fitness equipment anomaly identification, comprising the steps of:
acquiring an image video of a target area in a set time period, wherein the target area comprises fitness equipment to be identified;
dividing the image video according to the time index to obtain K image sub-videos; fitting according to the using heat of the fitness equipment to be identified corresponding to each image sub-video to obtain a using heat change function corresponding to the image video, wherein K is more than or equal to 2;
according to the image video, obtaining the use duration and the nonstandard action degree corresponding to each fitness person using the fitness equipment to be identified; calculating the nonstandard use degree corresponding to the fitness equipment to be identified according to the use duration and the action nonstandard degree corresponding to each fitness person;
and calculating the abnormal degree of the fitness equipment to be identified according to the use heat degree change function and the irregular use degree.
2. The method for anomaly identification of fitness equipment according to claim 1, wherein the method for calculating the use heat of the fitness equipment to be identified corresponding to each image sub-video comprises the following steps:
marking the fitness person and the fitness equipment to be identified in each image sub-video by using the surrounding frame to obtain the surrounding frame of the fitness person and the surrounding frame of the fitness equipment to be identified;
judging whether the exerciser uses the fitness equipment to be identified or not according to the intersection ratio of the exerciser enclosure frame and the fitness equipment enclosure frame to be identified;
counting the use times and the single use duration of the fitness equipment to be identified in each image sub-video;
and calculating the use heat of the fitness equipment to be identified corresponding to each image sub-video according to the use times and the single use duration of the fitness equipment to be identified corresponding to each image sub-video.
3. The method for anomaly identification of fitness equipment according to claim 2, wherein the heat of use of the fitness equipment to be identified corresponding to each image sub-video is calculated by using the following formula:
Figure DEST_PATH_IMAGE001
wherein k is the kth image sub-video,
Figure 885553DEST_PATH_IMAGE002
the using heat of the fitness equipment to be identified corresponding to the k image sub-video,
Figure 535584DEST_PATH_IMAGE003
the number of uses corresponding to the k-th image sub-video,
Figure 595943DEST_PATH_IMAGE004
the time length average value of the single use corresponding to the k image sub-video,
Figure 149285DEST_PATH_IMAGE005
in order to use the corresponding weight of the times,
Figure 269687DEST_PATH_IMAGE006
the weight is corresponding to the average single-use duration.
4. The method for abnormality recognition of fitness equipment according to claim 1, wherein the calculation of the degree of motion irregularity comprises:
identifying key points of a body builder using fitness equipment to be identified in the image video;
obtaining a fitness action sequence of the fitness person using the fitness equipment to be identified according to the human body key points;
and comparing the fitness action sequence of the exerciser using the fitness equipment to be identified with the standard fitness action sequence to obtain the nonstandard action degree corresponding to the exerciser using the fitness equipment to be identified.
5. The method for anomaly identification of fitness equipment according to claim 4, wherein the corresponding non-standard use degree of the fitness equipment to be identified is calculated by using the following formula:
Figure 642900DEST_PATH_IMAGE007
)
Figure 506951DEST_PATH_IMAGE008
wherein, the first and the second end of the pipe are connected with each other,
Figure DEST_PATH_IMAGE009
the non-standard use degree corresponding to the fitness equipment to be identified, R is the total number of the fitness users using the fitness equipment to be identified, R is the R th fitness user using the fitness equipment to be identified, J is the length of the fitness action sequence, I is the number of key points of the human body, I is the key point of the ith human body, J is the jth fitness action in the fitness action sequence,
Figure 383640DEST_PATH_IMAGE010
is the abscissa of the ith human body key point in the jth body-building action corresponding to the body-building person,
Figure 301042DEST_PATH_IMAGE011
is the abscissa of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 36917DEST_PATH_IMAGE012
is the ordinate of the ith personal key point in the jth body-building action corresponding to the body-building person,
Figure 970238DEST_PATH_IMAGE013
is the ordinate of the ith personal key point in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 967013DEST_PATH_IMAGE014
is the vertical coordinate of the ith personal body key point in the jth body-building action corresponding to the body-building person,
Figure 553852DEST_PATH_IMAGE015
is the vertical coordinate of the key point of the ith personal body in the standard body-building action corresponding to the jth body-building action corresponding to the body-building person,
Figure 42602DEST_PATH_IMAGE016
for the degree of irregularity of the motion of the r-th exerciser using the exercise equipment to be recognized,
Figure 514035DEST_PATH_IMAGE017
for the length of time the r-th exerciser used the exercise apparatus to be identified.
6. A method for fitness equipment abnormality identification according to claim 1, wherein the method for calculating the degree of abnormality of the fitness equipment to be identified according to the usage heat variation function and the degree of irregular usage comprises:
judging the change trend of the use heat of the fitness equipment to be identified according to the use heat change function; if the variation trend of the use heat of the fitness equipment to be identified is increased progressively, calculating the abnormal degree of the fitness equipment to be identified according to the following formula:
Figure 630896DEST_PATH_IMAGE018
wherein the content of the first and second substances,
Figure 264002DEST_PATH_IMAGE019
to identify the degree of abnormality of the exercise apparatus,
Figure 343778DEST_PATH_IMAGE009
for the corresponding non-standard use degree of the fitness equipment to be identified,
Figure 618901DEST_PATH_IMAGE020
for the used length of time of the exercise apparatus to be identified,
Figure 590268DEST_PATH_IMAGE021
in order not to specify the weight corresponding to the degree of use,
Figure 659855DEST_PATH_IMAGE022
the weight corresponding to the used time length.
7. The method for abnormality recognition of fitness equipment according to claim 6, wherein if the trend of the usage heat of the fitness equipment to be recognized is decreasing and the difference between the usage heat of the fitness equipment of the same type and the usage heat of the fitness equipment to be recognized is greater than the set usage heat threshold, the abnormality degree of the fitness equipment to be recognized is calculated according to the following formula:
Figure 716673DEST_PATH_IMAGE023
wherein the content of the first and second substances,
Figure 529908DEST_PATH_IMAGE024
the difference between the use heat of the fitness equipment of the same type and the use heat of the fitness equipment to be identified,
Figure 355782DEST_PATH_IMAGE025
is composed of
Figure 596270DEST_PATH_IMAGE024
The corresponding weight.
CN202211245039.0A 2022-10-12 2022-10-12 Method for recognizing abnormity of fitness equipment Active CN115311610B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202211245039.0A CN115311610B (en) 2022-10-12 2022-10-12 Method for recognizing abnormity of fitness equipment

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202211245039.0A CN115311610B (en) 2022-10-12 2022-10-12 Method for recognizing abnormity of fitness equipment

Publications (2)

Publication Number Publication Date
CN115311610A true CN115311610A (en) 2022-11-08
CN115311610B CN115311610B (en) 2023-03-28

Family

ID=83867864

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202211245039.0A Active CN115311610B (en) 2022-10-12 2022-10-12 Method for recognizing abnormity of fitness equipment

Country Status (1)

Country Link
CN (1) CN115311610B (en)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113627409A (en) * 2021-10-13 2021-11-09 南通力人健身器材有限公司 Body-building action recognition monitoring method and system
CN113642896A (en) * 2021-08-16 2021-11-12 江苏动泰运动用品有限公司 Gymnasium safety risk early warning method and system based on artificial intelligence
CN113701825A (en) * 2021-10-27 2021-11-26 南通高桥体育用品有限公司 Body-building facility abnormity detection method and system based on artificial intelligence
CN114870323A (en) * 2022-04-11 2022-08-09 北京觅淘智联科技有限公司 Fitness equipment and exercise evaluation method for fitness equipment

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113642896A (en) * 2021-08-16 2021-11-12 江苏动泰运动用品有限公司 Gymnasium safety risk early warning method and system based on artificial intelligence
CN113627409A (en) * 2021-10-13 2021-11-09 南通力人健身器材有限公司 Body-building action recognition monitoring method and system
CN113701825A (en) * 2021-10-27 2021-11-26 南通高桥体育用品有限公司 Body-building facility abnormity detection method and system based on artificial intelligence
CN114870323A (en) * 2022-04-11 2022-08-09 北京觅淘智联科技有限公司 Fitness equipment and exercise evaluation method for fitness equipment

Also Published As

Publication number Publication date
CN115311610B (en) 2023-03-28

Similar Documents

Publication Publication Date Title
CN110837784B (en) Examination room peeping and cheating detection system based on human head characteristics
CN107133612A (en) Based on image procossing and the intelligent ward of speech recognition technology and its operation method
CN109671068B (en) Abdominal muscle labeling method and device based on deep learning
CN113158752A (en) Intelligent safety management and control system for electric power staff approach operation
CN109544523A (en) Quality of human face image evaluation method and device based on more attribute face alignments
CN106709438A (en) Method for collecting statistics of number of people based on video conference
CN110096945B (en) Indoor monitoring video key frame real-time extraction method based on machine learning
CN115227234B (en) Cardiopulmonary resuscitation pressing action assessment method and system based on camera
CN114973401A (en) Standardized pull-up assessment method based on motion detection and multi-mode learning
CN111709365A (en) Automatic human motion posture detection method based on convolutional neural network
CN116563922A (en) Automatic rope skipping counting method based on artificial intelligence
CN114998986A (en) Computer vision-based pull-up action specification intelligent identification method and system
CN114170686A (en) Elbow bending behavior detection method based on human body key points
CN117612261A (en) Sit-up standard counting method and system based on image judgment
CN115311610B (en) Method for recognizing abnormity of fitness equipment
CN110188718A (en) It is a kind of based on key frame and joint sparse indicate without constraint face identification method
CN117339184A (en) Intelligent easy-to-operate body-building sports equipment
CN112233770A (en) Intelligent gymnasium management decision-making system based on visual perception
CN114639168B (en) Method and system for recognizing running gesture
CN115909400A (en) Identification method for using mobile phone behaviors in low-resolution monitoring scene
CN112597842B (en) Motion detection facial paralysis degree evaluation system based on artificial intelligence
CN110502992B (en) Relation graph based fast face recognition method for fixed scene video
CN110309722B (en) Sports video motion identification method based on motion hotspot graph
CN114220084A (en) Distribution equipment defect identification method based on infrared image
CN112802051A (en) Fitting method and system of basketball shooting curve based on neural network

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
GR01 Patent grant
GR01 Patent grant