Visual analysis or active control-based fitness equipment intelligent assistance system driven by motor and action cloud recording method

Abstract

The invention relates to the field of intelligent fitness auxiliary equipment, in particular to a fitness action real-time assisting and analyzing system based on a mobile phone terminal and motor driving, is matched with dumbbell, barbell and other instruments, and is suitable for families, gymnasiums and other scenes. An existing scheme has the defects that a traditional instrument does not have intelligent interaction and is prone to exhaustion, so that action deformation and even injury are caused, and the bodybuilding threshold is high; commercial intelligent equipment is high in cost and poor in portability; an existing fitness APP only records data and does not have a real-time power assisting function, and use of a free weight instrument used by a green hand is easily affected due to being afraid of being injured. The invention aims to overcome the problems of high cost, poor portability and disjunction of analysis and power assistance, a mobile phone is used as a computing power / acquisition terminal, a motor is used as a power assistance execution mechanism, a closed loop of'real-time action analysis-insufficient power judgment-precise power assistance driving 'is realized, and meanwhile, functions of training automatic recording, editing and uploading are integrated to form a personal exercise database, so that the personal exercise database can be used for training. The fitness safety and effect are improved, and the intelligent fitness threshold is lowered.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent fitness assistance devices, in particular to a fitness motion real-time analysis and assistance system based on a mobile phone terminal and a motor drive, which is suitable for various scenes such as home fitness, fitness centers, and fitness studios, and can be adapted to common fitness equipment such as dumbbells, barbells, kettlebells, and other fixed equipment. BACKGROUND

[0002] The existing fitness assistance solutions have significant technical defects, which are manifested in the following three aspects: 1. Traditional fitness equipment lacks intelligent interaction capability: Conventional dumbbells, barbells, and other equipment completely rely on user self-control actions and cannot judge the action standard and user strength state in real time. When the user's strength is insufficient, elbow joint over-bending, shoulder joint compensation, and other action deformation problems are prone to occur, which not only seriously affects the fitness effect, but also may cause muscle strains, joint injuries, and other safety risks. Although professional trainers can assist in avoiding risks, it is difficult for trainers to accurately monitor the action standard during the assistance process; 2. Existing intelligent fitness equipment is high in cost and poor in portability: Although some commercial intelligent fitness equipment has basic motion analysis function, it needs to be equipped with special cameras, independent computing power modules, and customized driving mechanisms, resulting in large equipment size and high cost (generally between several thousand yuan and tens of thousands of yuan), which cannot adapt to light-weight use scenarios such as small family spaces; 3. Insufficient coordination between motion analysis and assistance function: A few fitness equipment with APP can only record basic data such as exercise frequency and duration through the mobile phone, cannot generate accurate assistance instructions combined with real-time motion parameters, and cannot realize equipment assistance through hardware driving. Users still need to overcome the weight of the equipment alone, and need the assistance of others when the strength is insufficient, so the use convenience and safety need to be improved.

[0003] In view of the defects of the above-mentioned prior art, there is an urgent need in the art for a low-cost, high-portability, and real-time motion analysis-accurate assistance closed-loop fitness assistance system, and the present application is proposed accordingly. SUMMARY

[0004] I. Invention purpose The present application aims to overcome the core defects of "high cost, poor portability, disconnection between motion analysis and assistance" in existing fitness assistance solutions, and provides an intelligent system with a mobile phone as the core computing power and data collection terminal, and a motor end as the assistance execution mechanism. The system can realize the integrated function of "real-time motion analysis → power deficiency determination → precise power driving", and at the same time, integrate the whole process recording function of automatic recording, intelligent editing and cloud uploading of motion video, and build a personal exclusive exercise database. The present application can effectively reduce the threshold of intelligent fitness, improve the safety and training effect of the fitness process; at the same time, the active control switch is added, which allows the user to judge the power opportunity and trigger the control independently, and enhances the psychological certainty and operation autonomy in the use process.

[0005] Technical scheme

[0006] The fitness motion assistance system disclosed by the present application is mainly composed of software end, Bluetooth control end and motor end. Among them, the software end realizes data interaction with the motor end through WiFi communication, and the Bluetooth control end realizes control signal interaction with the motor end through Bluetooth communication, and the specific technical scheme is shown in the accompanying drawings Figure 1 .

[0007] 1. Software end (with mobile phone as carrier) The software end is integrated in the mobile phone APP, relying on the existing hardware resources of the mobile phone to realize data collection and analysis function, which specifically includes three function modules: (1) Data collection module: the rear or front camera of the mobile phone is used as the collection unit to capture the joint dynamic data of the user in the fitness process in real time, and the angle and angular velocity parameters of the core motion joints such as elbow joint, shoulder joint and knee joint are collected. For example, when the user performs the back barbell deep squat action, the camera can collect the key data such as the knee joint bending angle and the angle change speed every frame; (2) Computing power and algorithm module: The algorithm utilizes the computing power of the mobile phone's CPU / GPU and combines it with a dedicated joint data algorithm to complete motion analysis. The algorithm has three core capabilities: First, it pre-stores joint motion data models of multiple sets of standard fitness movements (e.g., the elbow flexion angle range of "standard dumbbell curl" is 30°-120°, and the shoulder abduction angle is ≤45°); Second, it pre-stores the joint angle and angular velocity change logic of mainstream barbell and dumbbell movements, which can intelligently identify the starting frame of the movement (e.g., the joint angle when the barbell is pulled out of the barbell squat), the peak frame (e.g., the joint angle when the barbell squat reaches the bottom), and the ending frame (e.g., the joint angle when the barbell squat is fully upright), providing a basis for determining the timing of assistance; Third, it compares the real-time collected joint dynamic data with the standard model and calculates the percentage of deviation (e.g., the actual elbow flexion angle is 20°, and the percentage of deviation from the standard lower limit of 30° is 33%). When the percentage of deviation exceeds the preset threshold, it confirms the user's insufficient strength through the dual judgment logic of "joint angle + joint angular velocity" and generates a "assistance demand signal". (3) Communication and interaction module: Sends control signals such as assist start and stop to the motor end through the mobile phone WiFi module (compatible with 802.11b / g / n protocol); at the same time, it integrates data display function, and displays joint dynamic data, algorithm comparison results, motor end assist status and battery power in real time on the APP interface; it also has parameter setting function, allowing users to adjust the deviation threshold, assist level, WiFi transmission baud rate and standard joint motion data model parameters according to their own situation.

[0008] 2. Bluetooth control unit (active control assist) The Bluetooth control unit features an adjustable touch sensor design, allowing users to activate the assist function with a simple touch and deactivate it by removing the touch point when no further assistance is needed. This convenient operation does not interfere with the continuity of fitness movements.

[0009] 3. Motor end (assist actuator) The motor end adopts an independent modular structure, as shown in the attached diagram of the instruction manual. Figure 2 As shown, it can be fixed to the fitness area (such as a wall or equipment rack) via a bracket or rope, and is detachably connected to the fitness equipment. Specifically, it includes a power component and a control component. (1) Power assembly: contains a direct current servo motor (power range 30W-100W, torque range 1N·m-5N·m), an electromagnetic clutch mechanism (composed of an electromagnetic clutch and a shaft coupling, with a slip torque of 1.5N·m), and a rope winding mechanism (including a winding wheel, a guide pulley, and a travel limiting assembly). The winding wheel has a diameter of 50mm-80mm, and the rope is made of high-strength polyethylene material (diameter 3mm-5mm, breaking strength ≥1000N). One end of the rope is fixed to the winding wheel through a bolted locking + rubber non-slip pad anti-slip fastening structure, and the other end is connected to the exercise equipment moving part (such as the middle of the dumbbell rod, the ends of the barbell rod) through a quick-release buckle; Control assembly: contains an electronic control unit and a battery. The electronic control unit uses an ESP32 module; the battery uses a 18650 rechargeable lithium battery pack (voltage 12V, capacity 6000mAh), with a power detection module (real-time acquisition of remaining power through voltage sampling) and an overcharge and overdischarge protection module (automatic power supply circuit cut-off when the power is <10% or >95%).

[0010] System workflow The system supports two working modes: software automatic control and Bluetooth active control. The specific process is as follows: (1) Software automatic control process: the user starts the mobile phone APP and turns on the motor end power supply, connects with the motor end through WiFi search or selects the saved device; in the APP, select the target action type (such as "dumbbell curl") through the parameter setting module, the APP automatically loads the corresponding standard joint motion data model; after connecting the motor end rope to the exercise equipment through the quick-release buckle, start the exercise action, the mobile phone camera real-time collects joint dynamic data, the algorithm module combines with the exclusive algorithm for real-time analysis; when the algorithm determines that the user's strength is insufficient and needs assistance, the software sends a power assistance control signal to the motor end through WiFi; the motor end electronic control unit receives and analyzes the signal, controls the electromagnetic clutch to connect and attract, drives the motor to run, and drives the rope winding mechanism to tighten the rope to lift the exercise equipment; when the algorithm detects that the joint data deviation is less than the threshold (the user's strength is restored) or the rope reaches the travel limit, the software sends a "stop power assistance" signal, the motor stops running, the electromagnetic clutch is disconnected, and a power assistance cycle is completed.

[0011] (2) Bluetooth active control process: the user fixes the Bluetooth remote control switch assembly through the quick-release buckle beside the hand holding position during exercise, adjusts the touch sensor to a position that can be quickly touched by the fingers; connects the motor end and the Bluetooth remote control switch through the APP; during the exercise, when the user needs power assistance, touching the touch sensor can send a power assistance signal, and releasing the touch can send a stop power assistance signal, and the motor end executes the corresponding action according to the received signal.

[0012] The software can automatically segment and intelligently crop the video captured by the camera based on the frame count data of the movement, and upload the cropped movement video to the cloud platform for storage and recording; during the fitness process, the mobile APP displays joint dynamic data, assistance status and battery level in real time. When the battery level is lower than 10%, the APP will automatically trigger a pop-up reminder.

[0013] Beneficial effects

[0014] Compared with the prior art, the present invention has the following significant advantages: 1. Low cost and high portability: It reuses the camera and computing power resources of the user's existing mobile phone, without the need for additional dedicated acquisition or computing hardware; the motor end adopts a lightweight modular design (weight ≤2kg), which is easy to store and carry, and can be adapted to small home spaces. The overall cost is only 1 / 10-1 / 20 of that of commercial smart fitness equipment. 2. Accuracy of motion analysis: The exclusive joint data algorithm is customized and developed for the characteristics of fitness movements, which can accurately capture the dynamic parameters of core joints and realize the accurate determination of the timing of assistance and signal output; 3. Enhanced safety: The electromagnetic clutch is set with a slippage torque of 1.5 N·m. When the force on the rope end exceeds the threshold, it can automatically slip, avoiding damage to the equipment due to excessive tension and preventing mechanical injury to the user. 4. High compatibility: The quick-release buckle enables rapid connection between the rope and fitness equipment, and it can be used with a variety of common equipment such as dumbbells, barbells, and kettlebells. There is no need to change to special equipment, which greatly improves the utilization rate of the equipment. 5. User experience optimization: The mobile app displays key data in real time, allowing users to intuitively grasp the accuracy of their movements and the assistance status; the parameter setting function can meet the personalized needs of users with different weights and strength levels, improving usability. Attached image description: Figure 1 This is a schematic diagram of the interaction between the software, Bluetooth control, and motor terminals of the fitness movement assistance system of the present invention. Figure 2 This is a schematic diagram of the motor end structure of the fitness movement assist system of the present invention. Detailed Implementation

[0016] To make the technical solution of the present invention clearer and more explicit, the following detailed description is provided with specific embodiments in conjunction with the "10kg dumbbell curl" exercise adaptation scenario: I. System Component Selection 1. Software end: Use mainstream smartphones as carriers, and develop APP based on Android 12 / iOS 16 and above operating systems; action recognition algorithm is deployed on the phone end based on the Mediapipe framework, and the standard joint motion data model is generated by training a large number of standard action and action video samples under insufficient power to ensure accuracy.

[0017] 2. Motor end: Motor: Select 12V DC servo motor, parameters: power 30W, maximum torque 1.8N·m, speed range 0-300rpm; rope winding mechanism: winding wheel diameter 60mm, equipped with 2 nylon material guide pulleys, upper limit of travel limit set to 80cm (adapted to dumbbell curl effective stroke), lower limit set to 5cm; electronic control unit: use ESP32-C3 chip module, support 802.11n protocol, ensure communication stability; battery: 18650 lithium battery pack (specification 12V / 6000mAh), endurance time ≥2 hours, power detection accuracy ±5%; rope and buckle: high-strength polyethylene rope (diameter 4mm, breaking strength 600N), stainless steel quick-release buckle (adapted to dumbbell rod with diameter 25-60mm specification).

[0018] 3. Bluetooth remote control end: use a general Bluetooth remote control module and a custom combination of touch sensors, the core realizes flexible adjustment of touch sensor position, ensuring minimal interference during operation.

[0019] II. Actual working process 1. Software automatic control mode (1) Preparation stage: Users can connect the motor end to the exercise equipment through the following three support forms (one of them can be selected): ① The rope is hung beside the force position of the equipment; ② The floor stand is placed beside the force position of the equipment; ③ Fixed connection with the frame type equipment through the connecting support. Connect one end of the rope to the middle of the dumbbell rod through the quick-release buckle, turn on the power of the motor end, pair the phone APP with the motor end through WiFi (pairing time <3 seconds), select "dumbbell curl" action in the APP, and the system automatically loads the corresponding standard joint data model.

[0020] (2) Motion Acquisition and Analysis: The user holds 10kg dumbbells with both hands, letting them hang naturally (initial position) and begins a curling motion. The mobile phone camera is positioned 1.2m directly in front of the user, and the elbow joint angle and angular velocity data are collected in real time at a default sampling frame rate of 10fps (adjustable range 5-30fps). When the user reaches the halfway point of the movement, the elbow joint flexion angle drops to 20° due to insufficient strength (below the standard lower limit of 30°, deviation percentage 33%). The algorithm determines that the assistance condition is met and generates and sends an "assistance demand signal".

[0021] (3) Assisted execution: The software sends a control signal of "assistance level 2 (corresponding to motor speed of 150rpm and torque of 0.8N·m)" to the motor. After receiving the signal, the ESP32-C3 chip module on the motor receives and analyzes it, and drives the motor to run. The reel tightens the rope synchronously. After the rope changes the direction of force through the guide pulley, it lifts the dumbbell upward (lifting speed of 10cm / s), helping the user to restore the elbow joint angle to 40° (within the standard range).

[0022] (4) Stop Assistance and Safety Protection: When the user's strength recovers and the elbow angle stabilizes at 50° (the percentage deviation is lower than the preset threshold), the software sends a "Stop Assistance" signal, the motor is powered off and stops running, and the electromagnetic clutch disengages. If the user suddenly loses strength and releases their grip, the equipment remains in place when the weight of the dumbbell is less than the slippage torque of the electromagnetic clutch; when the weight exceeds the slippage torque, the equipment slowly descends to the ground via a rope to avoid the risk of falling.

[0023] 2. Bluetooth active control mode (1) Preparation stage: The connection method between the motor and the fitness equipment, the equipment startup and WiFi pairing process are the same as the software automatic control mode. An additional Bluetooth connection step is added: search for the Bluetooth remote control device in the APP and complete the selection and pairing to establish a signal link between the Bluetooth remote control and the motor; the user fixes the Bluetooth remote control to the nearest position next to the hand holding position using the quick-release strap, and adjusts the touch sensor to a position that can be reached with only a slight finger joint movement.

[0024] (2) Assisted execution and stop: When the user subjectively judges that assistance is needed during the dumbbell curl, he touches the touch sensor with a small finger joint movement. The Bluetooth remote control immediately sends an assistance start signal to the motor, and the motor performs the assistance action. When the user recovers his strength or completes the exercise and needs to stop the assistance, he touches the touch sensor again (or removes the touch part). The Bluetooth remote control sends an assistance stop signal, the motor stops running and the electromagnetic clutch is disengaged.

Claims

1. A fitness movement assistance system, characterized in that, Including software, active control, and hardware, these three components work together via signal coordination to achieve closed-loop control of "motion monitoring - intelligent judgment - precise assistance," as detailed below: The software uses a mobile terminal equipped with a camera and a computing module as its carrier. The camera collects dynamic joint data of the user's fitness movements in real time, and the computing module calls a dedicated joint data algorithm to analyze the collected data and outputs the "assistance demand determination result" simultaneously. When the determination result is "assistance is needed", the software sends a control signal containing parameters such as "assistance level and start / stop timing" to the hardware via WiFi. The active control terminal is interconnected with the hardware terminal via Bluetooth. It uses a touch switch as an emergency trigger. The assist signal it sends and the software terminal signal form a dual control system. The software terminal signal takes priority, and the active control terminal signal can independently trigger or interrupt the assist. The hardware is fixedly connected to the fitness equipment and integrates a motor, electromagnetic clutch, electronic control unit, smart battery, and rope lifting mechanism. The smart battery powers the motor and provides real-time feedback of power to the software. The electronic control unit synchronously receives and decodes signals from both ends, driving the motor to rotate the rope lifting mechanism and providing assistance by pulling the fitness equipment with the rope. The electromagnetic clutch, in conjunction with the coil spring, enables the switching between "power connection when assisting and rope taut when not assisting" to ensure continuous movement.

2. The fitness movement assistance system according to claim 1, characterized in that, The "algorithm analysis - signal decoding - motor control" form a precise linkage: the algorithm compares the collected joint dynamic data with the standard model, and outputs the assistance demand when the deviation exceeds the threshold and is determined to be "insufficient force". After decoding the signal, the electronic control unit adjusts the motor speed and torque to match the assistance force. At the same time, the travel limit component monitors the rope position in real time. When the limit is exceeded, the motor is immediately triggered to stop, realizing the joint control of "intelligent assistance + safety protection".

3. The fitness movement assistance system according to claim 1, characterized in that, The hardware features a combined design for safety and compatibility: the rope is made of high-strength, wear-resistant material and is fixed to the reel via an anti-slip structure, with the other end detachably connected to the fitness equipment; the intelligent battery integrates an overcharge and over-discharge protection module, automatically cutting off power when the charge level is below 3% or above 98%; the electromagnetic clutch is controlled by software signals, engaging when powered to transmit motor power and disengaging when powered off to allow the reel to rotate freely, working with the coil spring to keep the rope taut and prevent interruption of movement.

4. The fitness movement assistance system according to claim 1, characterized in that, The software and hardware form an interactive closed loop: the software display module displays joint data, algorithm results, assist status and battery level in real time; the parameter setting module allows users to adjust the judgment threshold, assist level and standard model parameters; the adjustment signal is synchronized to the electronic control unit via WiFi to realize the linkage control of "personalized settings - instant response"; after the touch switch of the active control end is triggered, the Bluetooth signal is directly transmitted to the electronic control unit, which can quickly start the assist to meet emergency needs.

Images