Artificial intelligence fitness guidance system based on visual communication and device thereof

Abstract

The present application relates to a kind of artificial intelligence fitness guidance system based on visual communication and its device, including movable observation camera support (1), human observation camera (2), display screen (3), display screen lifting support (4) and movable observation camera support arc track (5), the display screen lifting support (4) one side surface fixedly connected with one end of movable observation camera support arc track (5), display screen (3) is installed on the surface of display screen lifting support (4) by lift, movable observation camera support (1) is clamped on the surface of movable observation camera support arc track (5) by roller, more than two human observation cameras (2) are fixedly installed on the surface of movable observation camera support (1), the action state of the person of real-time capture can be carried out to the correction of not standard action, can effectively improve fitness safety, ensure healthy safety body-building.

Description

TECHNICAL FIELD

[0001] The application relates to a fitness guidance device, in particular to an artificial intelligence fitness guidance system based on visual communication and a device thereof. BACKGROUND

[0002] In a fitness process, if direct training is performed without knowing, muscle strains are easily caused, and if exercise exceeding a threshold value of a person is forcedly used, serious problems such as bone fracture are caused, which greatly violates the original intention of fitness and body building. SUMMARY

[0003] In view of the problems in the prior art, the application provides an artificial intelligence fitness guidance system based on visual communication and a device thereof, which can capture the action state of a fitness person in real time and correct non-standard actions, and can effectively improve fitness safety and ensure healthy body building.

[0004] In order to achieve the above purpose, the application realizes the technical scheme as follows: an artificial intelligence fitness guidance system based on visual communication and a device thereof, comprising a movable observation camera support, a human body observation camera, a display screen, a display screen lifting support and a movable observation camera support arc-shaped track, one end of the movable observation camera support arc-shaped track is fixedly connected to one side surface of the display screen lifting support, the display screen is installed on the surface of the display screen lifting support through an elevator, the movable observation camera support is clamped on the surface of the movable observation camera support arc-shaped track through a roller, two or more human body observation cameras are fixedly installed on the surface of the movable observation camera support, and the display screen lifting support is electrically connected with the movable observation camera support, the human body observation camera and the display screen respectively.

[0005] The human body observation camera is a 3D scanning camera.

[0006] The display screen lifting support is internally provided with an information processing machine.

[0007] The information processing machine internally provided in the display screen lifting support is a PPK2035 type information processing machine.

[0008] The logic steps of the artificial intelligence fitness guidance system are as follows:

[0009] Step one: personnel actions are comprehensively collected through a camera;

[0010] Step two: a rule table is generated according to the collected personnel actions, rules of motion index analysis based on human body joint points are defined, the motion images of a motion person are collected in real time, and the skeleton point information of the motion person is acquired;

[0011] Step three: human body joint points in a standard action in a database are collected;

[0012] Step four: matching the human skeleton node information output by step two and step three;

[0013] Step five: judging whether the error value of the generated skeleton point information and the standard action is less than the predetermined threshold value, if yes, the action is qualified, otherwise, output the error skeleton point information;

[0014] Step six: overlappingly displaying the error skeleton node information and the standard skeleton node information on the screen;

[0015] Step seven: observing whether the personnel action is corrected, if yes, starting to guide the next action, if not, repeating steps one to six.

[0016] When the step one is executed, the human key point information is directly extracted by the depth camera, the specified human skeleton points and positions are obtained, the rule index is generated, and is recorded one by one.

[0017] The human joint node data in step two is identified by a multi-angle mode image recognition system.

[0018] By the above scheme, the present application has at least the following advantages: through comprehensive observation and scientific comparison of the fitness personnel, the action state of the fitness personnel can be captured in real time and the non-standard action can be corrected, the fitness safety can be effectively improved, and the health safety of body building is ensured. BRIEF DESCRIPTION OF DRAWINGS

[0019] Figure 1 It is a device perspective view of an artificial intelligence fitness guidance system based on visual communication and its device;

[0020] Figure 2 It is a logic block diagram of an artificial intelligence fitness guidance system based on visual communication;

[0021] In the figure: 1, movable observation camera support, 2, human observation camera, 3, display screen, 4, display screen lifting support, 5, movable observation camera support arc-shaped track. DETAILED DESCRIPTION

[0022] The present application will be further described below in combination with the drawings.

[0023] In the description of the present application, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer" and the like indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present application and simplifying the description, and do not indicate or imply that the devices or elements referred to must have a particular orientation, be constructed and operated in a particular orientation, and therefore cannot be understood as limiting the present application. In addition, the terms "first", "second", "third" are only for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0024] As shown in Figure 1 The present application discloses an artificial intelligence fitness guidance system based on visual communication and a device thereof, which comprises a movable observation camera support 1, a human body observation camera 2, a display screen 3, a display screen lifting support 4 and a movable observation camera support arc-shaped track 5. One side of the display screen lifting support 4 is fixedly connected with one end of the movable observation camera support arc-shaped track 5, the display screen 3 is installed on the surface of the display screen lifting support 4 through a lifting machine, the movable observation camera support 1 is clamped on the surface of the movable observation camera support arc-shaped track 5 through a roller, two or more human body observation cameras 2 are fixedly installed on the surface of the movable observation camera support 1, and the display screen lifting support 4 is electrically connected with the movable observation camera support 1, the human body observation camera 2 and the display screen 3 respectively.

[0025] The human body observation camera 2 is a 3D scanning camera.

[0026] The display screen lifting support 4 is internally provided with an information processing machine.

[0027] The information processing machine of the display screen lifting support 4 is a PPK2035 type information processing machine.

[0028] The logic steps of the artificial intelligence fitness guidance system are as follows:

[0029] Step one: comprehensively collecting the actions of personnel through the camera;

[0030] Step two: generating a rule table based on the rules of motion index analysis defined by the human body joint nodes according to the collected actions of personnel, collecting the motion images of the motion person in real time, and obtaining the skeleton point information of the motion person;

[0031] Step three: collecting the human body joint nodes of the standard actions in the database;

[0032] Step four: matching the human body skeleton node information output by steps two and three;

[0033] Step five: judge whether the generated skeleton point information and the standard action are less than the predetermined threshold value, if yes, the action is qualified, otherwise, output the error skeleton point information;

[0034] Step six: superimposedly display the error skeleton node information and the standard skeleton node information on the screen;

[0035] Step seven: observe whether the personnel action is corrected, if yes, start to guide the next action, if not, repeat steps one to six.

[0036] When the step one is executed, the human key point information is directly extracted by using the depth camera, the specified human skeleton points and positions are acquired, the rule index is generated, and is recorded one by one.

[0037] The human joint node data in the step two is identified by using a multi-angle mode image recognition system.

[0038] When used, the artificial intelligence fitness guidance system based on visual communication and the device thereof are installed, the power supply is connected and the database equipment is imported, and then the operation can be started

[0039] Beneficial effects: through the comprehensive observation and scientific comparison of the fitness personnel, the action state of the fitness personnel can be captured in real time and the non-standard action can be corrected, the fitness safety can be effectively improved, and the health safety of the body building is ensured.

Claims

1. A visual communication-based artificial intelligence fitness guidance system, comprising a movable observation camera bracket (1), a human body observation camera (2), a display screen (3), a display screen lifting bracket (4), and a movable observation camera bracket arc track (5), characterized in that, The display screen lifting bracket (4) is fixedly connected to one end of the arc-shaped track (5) of the movable observation camera bracket. The display screen (3) is installed on the surface of the display screen lifting bracket (4) via a lifting mechanism. The movable observation camera bracket (1) is clamped on the surface of the arc-shaped track (5) of the movable observation camera bracket via rollers. Two or more human observation cameras (2) are fixedly installed on the surface of the movable observation camera bracket (1). The display screen lifting bracket (4) is electrically connected to the movable observation camera bracket (1), the human observation camera (2), and the display screen (3). The display screen lifting bracket (4) has a built-in information processor. It captures multi-angle images and identifies them through a multi-angle mode image recognition system. It overlays incorrect skeleton node information and standard skeleton node information on the screen. The logical steps of the AI-powered fitness guidance system are as follows: Step 1: Collect comprehensive data on personnel movements using cameras mounted on the curved track of the movable observation camera bracket; Step 2: Based on the collected human movements, define the rules for motion index analysis based on human joint points, generate a rule table, collect motion images of the exerciser in real time, and obtain the skeletal point information of the exerciser; Step 3: Collect human joint points from the standard movements in the database; Step 4: Match the human skeleton node information output from Step 2 and Step 3; Step 5: Determine whether the error value between the generated skeleton point information and the human joint point data of the standard movement is less than the predetermined threshold. If yes, the movement is qualified; otherwise, output the incorrect skeleton point information. Step 6: Overlay the incorrect skeleton node information and the standard skeleton node information on the screen; Step 7: Observe whether the observer's actions have been corrected. If yes, begin instructing the next action. If not, repeat steps 1 to 6.

2. The AI-powered fitness guidance system based on visual communication according to claim 1, characterized in that, The human body observation camera (2) mentioned above is a 3D scanning camera.

3. The AI-powered fitness guidance system based on visual communication according to claim 1, characterized in that, The information processor built into the display screen lifting bracket (4) is a PPK2035 type information processor.

4. The AI-powered fitness guidance system based on visual communication according to claim 1, characterized in that, In the execution step one, the depth camera is used to extract key human body information, obtain the specified human skeleton points and positions, generate rule indicators, and record them one by one.

Images