179 results about "Physical education" patented technology

The invention relates to the technical field of physical education, in particular to a physical education method and system based on multimedia and holographic AR fusion, and the method comprises the following steps: collecting the motion data of a trainee based on a multi-source sensor array, constructing a motion parameterized model, and generating a holographic projection generation module through the cooperation with a standard motion database; a holographic AR teaching image is generated in real time, an environment perception module is combined to carry out adaptive projection on illumination and spatial layout, a spatial positioning module is used to establish coordinate mapping between a trainee and a virtual image, and a multi-channel feedback device is started to output visual, tactile and auditory signals. And finally, comparing the action parameters with the standard action data to automatically generate a training report containing wrong action labels and correction suggestions. According to the method, the motion capture precision and the interactive immersion can be effectively improved, and efficient and extensible technical support is provided for modern physical education and training.

The invention relates to the technical field of physical education, and discloses an artificial intelligence-based physical education evaluation system, which comprises a data acquisition module, a multi-target detection module, a key point modeling module, an action trajectory analysis module, a multi-modal fusion module and a dynamic feedback module. The data acquisition module receives a video stream of the camera device and motion physiological data of the wearable sensor; the multi-target detection module locates a target based on a YOLOv8 algorithm; the key point modeling module constructs a three-dimensional action attitude model by using an OpenPose algorithm; the action trajectory analysis module evaluates action standard; the multi-modal fusion module integrates the data to construct a multi-dimensional feature matrix; the dynamic feedback module generates a real-time score and a personalized correction instruction through the optimized LSTM network. In addition, the system further comprises a model self-adaption module, an abnormal action recognition module and a distributed calculation module. According to the system, precise evaluation of physical education is realized, the sports safety is guaranteed, personalized teaching is promoted, and the physical education quality is effectively improved.

The invention relates to the technical field of physical education teaching, in particular to an interactive physical education teaching method based on a large model, and aims to realize real-time monitoring and personalized training guidance of the motion state of a student through an advanced sensor technology and a machine learning model. The method comprises the following steps: firstly, collecting motion data of a student in real time through a camera and a depth sensor, including motion posture, speed, response time, heart rate and other information; then, the collected data are input into a pre-trained large model, the model comprises a convolutional neural network, a support vector machine and a random forest model which are respectively used for evaluating the action standard, the exercise intensity and the fatigue degree of students, the large model automatically generates personalized training items, and the exercise postures of the students are captured in real time; according to the motion deviation correcting system and method, by comprehensively applying various technical means, the scientificity and individuation level of physical education are improved, and the motion deviation correcting system and method have wide application prospects.

The invention relates to the technical field of physical training intellectualization and man-machine interaction, in particular to an interactive physical education method and system based on augmented reality driving, and the method comprises the following steps: S1, collecting three-dimensional motion track data and muscle activation time sequence data; s2, generating kinematic deviation parameters and physiological deviation parameters; s3, generating an AR correction instruction set containing a joint pose correction vector and a muscle activation intensity adjustment suggestion based on the kinematics deviation parameter and the physiology deviation parameter; s4, projecting the AR correction instruction set to the field of view of the trainee in real time through AR display equipment; and S5, generating a physical education efficiency report containing the action precision improvement index and the neuromuscular control optimization scheme. According to the intelligent physical education system, by fusing multi-source motion perception, deviation analysis and AR visual correction, synchronous optimization of motion precision and neuromuscular control ability is realized, and an intelligent physical education system with real-time interaction characteristics is constructed.

The invention relates to the field of education health management, and discloses a physical education teaching management system based on big data, and the system comprises a data collection module which collects physiological, motion, emotion and environment data through a wearable device, a video monitor and a sensor; the data storage module is used for storing and managing structured, semi-structured and unstructured data by adopting a big data architecture; the data processing and analyzing module is used for performing data cleaning, preprocessing, feature extraction, dimension reduction and data fusion; the intelligent decision-making module is used for realizing health risk monitoring, exercise recommendation and teaching optimization by using a deep learning model; and the front-end display and interaction module provides data visualization, real-time feedback and personalized reports. Through real-time acquisition and hierarchical storage of multi-modal data of physiology, movement, emotion and environment, comprehensive monitoring of health and movement states of students is realized, rapid response and trend analysis are supported, and a basis is provided for safety guarantee and scientific decision in teaching.

The invention relates to the technical field of artificial intelligence, in particular to a physical education method and device based on a depth-of-field camera, equipment, a storage medium and a product. The method comprises the steps of obtaining motion image data of a student based on a depth-of-field camera; identifying the motion image data to obtain a motion posture of the student; constructing a 3D motion model based on the motion posture; comparing the 3D motion model with a preset standard motion posture; based on the comparison result, a physical education instruction scheme is generated, and the physical education teaching effect is improved.

The invention belongs to the technical field of teaching management, and discloses a sports classroom multi-dimensional data dynamic weight evaluation method and system. Comprising the steps of collecting physiological index data and classroom image data to obtain student data; performing data cleaning on the student data, and constructing and perfecting a student classroom database; constructing a dynamic weight model and carrying out iterative optimization to obtain an optimized dynamic weight model; performing dynamic weight distribution on the perfect student classroom database by utilizing the optimized dynamic weight model to obtain a weighted student classroom database; constructing a multi-dimensional evaluation framework to perform classroom performance evaluation on the weighted student classroom database, and outputting a classroom performance index of each student; classifying the weighted student classroom database to obtain a marked student classroom database; designing a personalized sports classroom teaching strategy, and sending the personalized sports classroom teaching strategy to a preset teacher terminal; student classroom performance evaluation is carried out by applying the dynamic weight, and personalized teaching of teachers is facilitated.

The present invention discloses an information-based sports education platform management system, which specifically relates to the field of sports training data management, and is used to solve the problem of accurate classification and feedback of multi-device training data. Firstly, based on the fluctuation characteristics of physiological data, the switching period data is captured to ensure the continuity and timeliness of the training data; then, the transition response consistency index and the target device feature matching degree are used to accurately define the effective transition data, clearly distinguish the physiological feature data of each device, and improve the classification accuracy; the defined effective transition data is finely classified and labeled through the target device dynamic feature model to ensure the independence and consistency of the data; finally, an adaptive rejection mechanism is used to screen out abnormal data, retain high-quality data to generate personalized feedback, enhance the pertinence of the feedback, and build a closed-loop system from data collection to feedback generation, providing scientific support for sports training data management.

The invention relates to the technical field of physical education management, and particularly discloses a digital informatization physical education comprehensive management system, which comprises a user interaction module, a data acquisition module, a data processing module, a system optimization module and a distributed database, the data acquisition module acquires teaching content, motion completion degree and user physiological data through a camera, a motion capture sensor and wearable equipment; the data processing module cleans and converts the data, and quantitatively evaluates whether the learning results of the students are qualified or not by using a constructed education evaluation coefficient mathematical model; the system optimization module calculates a scene optimization coefficient by analyzing a user favorable comment trend and multimedia loading performance, and automatically identifies a teaching scene needing to be optimized; the distributed database adopts a master-slave replication and fragmentation technology to improve data access efficiency and fault tolerance, digital management of physical education is achieved, and therefore the quality and efficiency of physical education are improved.

The invention discloses a motion capture method and system for physical education training based on image recognition, and the method comprises the steps: collecting a motion video image of an athlete through a multi-angle image sensor, and recognizing the body contour of the athlete in the motion video image in real time through a Center Net target detection algorithm; if detecting that the athlete extracts the key node, outputting the position of the key node; delimiting an ROI region of each key node according to a prediction result to obtain an ROI region image; and establishing a 3D attitude estimation model based on an HR-Net neural network, extracting an action trajectory of the marked region from the ROI region image, calculating biomechanical parameters based on the action trajectory, and generating personalized scores and improvement suggestions. Multi-dimensional evaluation from an absolute standard to individual differences is realized, and personalized scores and improvement suggestions are generated.

The invention discloses an intelligent turn-back running evaluation method and system based on face recognition. According to the method, firstly, intelligent acquisition devices are installed at the starting point and the ending point of a runway, and a high-resolution camera and an edge calculation module are used for carrying out real-time face snapshot and preprocessing on students; and carrying out face recognition by adopting a deep learning model, comparing a captured face image with a student face library, confirming the identity of the target to be detected, and associating with a running record. In the running process, the motion trail of the student is tracked in real time through the human body recognition technology, the turn-back action is accurately recognized, and the accuracy of turn-back point data is ensured. The problems of large manual timing error, difficulty in turning point monitoring, incomplete data recording and the like in a traditional turn-back running test are solved, and the automation degree and the data accuracy of the test are improved. Meanwhile, the system supports real-time data display and feedback, and provides a scientific basis for physical education and management.

The application relates to the technical field of physical education, and discloses a sports teaching evaluation system based on artificial intelligence, which comprises a data acquisition module, a multi-target detection module, a key point modeling module, an action trajectory analysis module, a multi-modal fusion module and a dynamic feedback module. The data acquisition module receives video streams of a camera device and motion physiological data of a wearable sensor; the multi-target detection module positions targets based on a YOLOv8 algorithm; the key point modeling module constructs a three-dimensional action posture model by using an OpenPose algorithm; the action trajectory analysis module evaluates action standardization; the multi-modal fusion module integrates data to construct a multi-dimensional feature matrix; and the dynamic feedback module generates real-time scores and individualized correction instructions through an optimized LSTM network. In addition, the system also comprises a model self-adaption module, an abnormal action recognition module and a distributed computing module. The system realizes accurate evaluation of sports teaching, guarantees sports safety, promotes individualized teaching, and effectively improves the quality of sports teaching.

The application relates to the technical field of data management, in particular to an intelligent sports teaching data management method and system. The method comprises the following steps: collecting motion video stream data and inertial motion data of students; mapping a camera coordinate system of the video stream data and a body coordinate system of the inertial motion data to a world coordinate system; constructing a state space model based on an extended Kalman filter, and iteratively calculating to obtain a corrected student three-dimensional skeleton posture sequence; calculating an optimal matching path and a similarity distance between the student three-dimensional skeleton posture sequence and a standard action template sequence; and generating a quantitative action score according to the similarity distance. The application eliminates integral drift while ensuring the anti-occlusion capability by using an extended Kalman filter algorithm; meanwhile, an improved dynamic time warping algorithm is introduced, the standardization and objectivity of action quality scoring are realized, and the scientificity, safety and individualization level of sports teaching are effectively improved.

The invention relates to the technical field of motion perception, in particular to an intelligent physical education real-time feedback system and method based on motion perception. The device comprises a joint capturing unit, a deviation coordination unit and an adjustment feedback unit. An angle judgment module calculates an upper arm vector and a front arm vector of an arm part, the dot product of the upper arm vector and the front arm vector is calculated according to the upper arm vector included component of the arm part and the front arm vector included component of the arm part, and the component of the upper arm vector of the arm part is used for calculating the modulus of the upper arm vector; calculating a model of a forearm vector by using a component of the forearm vector of the arm part, calculating an action angle of a joint by combining an upper arm vector and a forearm vector dot product with the model of the upper arm vector and the model of the forearm vector, judging whether the captured whole-body action of the student is a common activity or a movement, and calculating the whole-body action of the student through vector calculation and angle derivation. The action amplitude and speed of the joints can be accurately quantified, the error of subjective judgment is avoided, and the accuracy of analyzing the actions of students is improved.

The invention relates to a three-camera linkage and intelligent evaluation eight-brocade intelligent scoring system based on machine learning and a sum-product method. In an eight-section brocade scoring scene, a traditional subjective scoring mode has many defects. In order to solve the problems, an intelligent scoring system is constructed. The system comprises a video data acquisition module, a posture recognition module, an intelligent scoring module, a score integration module and the like. The video data acquisition module acquires multi-angle videos through three high-definition cameras; the posture recognition module accurately recognizes actions by using an advanced algorithm and a deep neural network technology; the intelligent scoring module designs a multi-dimensional index, and scores based on a sum-product algorithm; and the score integration module gathers scores from different perspectives to obtain a comprehensive score. According to the method, various measures are adopted to overcome the limitation of the existing AI technology in sports detection, including environment optimization and action detail identification, construction of a diversity and balance data set, research of cross-group adaptability, establishment of a man-machine cooperation trust relationship and the like. The system improves objectivity, accuracy and fairness of eight-segment brocade scoring, and is of great significance to eight-segment brocade teaching and competition. Meanwhile, a new technical means is provided for the fields of physical education and detection, the application range of the artificial intelligence technology in the field of physical education is expanded, and fusion innovation of the AI technology and the traditional field is promoted.

The invention discloses a student sports performance assessment method based on weighted GG clustering. The method comprises the following steps: 1, acquiring all body test item data and corresponding evaluation grades of students of all grades, wherein the body test items comprise body weight index, vital capacity, 50-meter running, sitting-position body forward flexion, standing long jump, pull-up (male) / 1-minute sit-up (female) and 1000-meter running (male) / 800-meter running (female); 2, normalizing the acquired body measurement data of each item to [0, 1]; 3, training a weighted GG clustering model by using data obtained by normalization and corresponding scores, and accurately determining a clustering center of each physical test level; 4, classifying actually measured physical measurement parameters by using the weighted GG clustering model trained in the step 3, and obtaining an accurate physical measurement result; and 5, taking an unreasonable or inaccurate result of physical test score evaluation as a new sample to correct the model so as to improve the precision of the model. The physical test score of the student can be rapidly and accurately judged, and the method has good practical application value.

The invention relates to the technical field of campus physical education, and discloses a campus physical digital management system and method based on a cloud platform, and the system comprises a multi-dimensional data acquisition module, a campus education cloud platform, and a feedback module. The multi-dimensional data acquisition module is used for acquiring student basic information data, student extracurricular activity data and student on-campus sports performance data, and the campus education cloud platform is used for storing the received data; according to the basic information data of the students, the extracurricular activity data of the students and the on-campus sports score data of the students, the physical health quality values of the students, the extracurricular participation achievement integral values, the score increase values and the comprehensive actual sports values are calculated, and then the comprehensive actual sports values are sent to the feedback module for feedback display; the method comprehensively considers the current score, physical health quality, extracurricular participation and the change trend of the score of the student, comprehensively evaluates the physical condition of the student, dynamically evaluates the campus education performance, and enables the campus physical education score to be more comprehensive and fair.

The data governance management system comprises a metadata acquisition module, the output end of the metadata acquisition module is bidirectionally and electrically connected with a terminal processor, and the output end of the terminal processor is bidirectionally and electrically connected with a data governance model establishment system for the sports of the Chinese school exam. Student information, physical fitness test scores and physical education course scores are collected through the metadata collection module, data are transmitted to the terminal processor, the terminal processor processes the data and transmits the data to the middle school examination physical education data management model building system, and the middle school examination physical education data management model building system builds a model. According to the system and the method, the operation mode of mastering the physical test state of the student in the middle school examination through a manual screening mode can be replaced, time and labor are saved, a large amount of time can be saved in the statistical process, and measurement and calculation errors are avoided.

The invention discloses a sprint intelligent evaluation method and system based on face recognition. The method comprises the following steps: firstly, installing intelligent acquisition equipment at a starting point and an ending point of a runway, and carrying out real-time face snapshot and preprocessing on students by utilizing a high-resolution camera and an edge calculation module; the system adopts a deep learning model to carry out face recognition, compares a captured face image with a student face database, confirms the identity of a target to be detected, and associates running records. At the final position, the system captures the faces of the students and collects running data including duration and speed allocation. And through real-time data processing and statistical analysis, an intelligent evaluation result is generated, and the intelligent evaluation result comprises a personalized exercise analysis report and a real-time ranking list. The problems of low manual timing efficiency, incomplete data recording, lack of real-time feedback and the like in a traditional sprint test are solved, the automation degree and the data accuracy of the test are improved, and a scientific basis is provided for physical education and management.

The invention discloses an auxiliary leg stretching instrument for physical education, and particularly relates to the technical field of auxiliary leg stretching instruments, the auxiliary leg stretching instrument comprises a base, two symmetrically distributed pressing assemblies are fixedly mounted on the front side of the middle of the upper end of the base, and clamping assemblies are fixedly mounted on the sides, away from each other, of the outer surfaces of the two pressing assemblies. According to the leg auxiliary stretching instrument for physical education, by arranging the pressing assembly, when the legs are stretched, the shanks and the thighs can be stretched, the stretching degree can be autonomously adjusted according to the actual requirements of a user, in the subsequent stretching process, the shanks and the thighs can be relaxed, and the stretching effect is improved. According to the leg stretching device, the situation that the legs are pulled after being kept in the stretching state for a long time is avoided, meanwhile, through the arranged clamping assembly, the thighs can be stretched in the using process, the popliteal space part can be supported in the stretching process, and therefore the legs are in the bent and relaxed state.

The invention discloses an intelligent psychological health education system based on an education robot. The system comprises (1) AI family-school cooperation: an improved YOLOv7 algorithm captures classroom behaviors, dynamic Gaussian blur protects privacy, directionally pushes student participation images containing desensitization IDs, supports AI to generate personalized communication verbal skills, and improves the degree of participation of parents; (2) an interdisciplinary fusion engine: constructing a knowledge graph to associate psychological courses with mathematics, sports and other subjects, predicting an optimal fusion opportunity in combination with educational administration data, and generating a double-teacher collaborative teaching plan; (3) interest and emotion driven daily penetration: analyzing student interest labels based on a latent semantic model (LFM), and triggering personalized family task pushing in combination with multi-modal emotion recognition to realize intelligent penetration of psychological health education in a family scene; and (4) full-link privacy protection: adopting differential privacy, block chain and federated learning technologies to realize'all anonymity 'and'availability and invisibility' of data, and ensuring information security through three-level desensitization and quantum encryption, so that the system meets the requirements of'personal information protection method '.

The utility model discloses a physical training device for physical education, which relates to the technical field of sports equipment and comprises a first mounting column, a second mounting column and a cross rod, positioning seats are fixedly mounted on the outer sides of the first mounting column and the second mounting column, and connecting mechanisms are fixedly mounted at two ends of the cross rod. Clamping grooves are symmetrically formed in the inner sides of the positioning seats, and the transverse rod is connected with the clamping grooves of the positioning seats between the opposite faces of the uppermost ends of the first mounting column and the second mounting column in a clamped mode through a connecting mechanism. By the adoption of the structure, when the cross rod is taken out of the positioning seat, only the two shifting blocks on the connecting shell need to be extruded at the same time, so that the shifting blocks move close to each other at the limiting holes of the baffles, and after limiting, the movable plates connected with the shifting blocks drive the clamping blocks to be separated from the clamping grooves of the positioning seat; therefore, the cross rod can be conveniently and rapidly disassembled, assembled and adjusted, and meanwhile, the situation that students move when using the cross rod can be avoided through the installation mode.

The invention discloses a multi-target sports classroom exercise density evaluation method and device under a single view angle and a medium. The method comprises the following steps: S110, a video acquisition and preprocessing stage; s120, constructing a data management framework; s210, constructing an electronic fence; s220, performing human body detection and ROI extraction by using a deep learning model; s230, designing a gridding index mechanism; s240, adopting a video stabilization method based on feature points; s310, designing a multi-target state updating mechanism; s320, establishing an identity mapping table and a target identity feature set; s330, constructing a target tracking model; according to the single-view-angle analysis technology, multi-target motion density evaluation based on a single-view-angle video is achieved, multi-angle camera arrangement is not needed, hardware cost and site requirements are remarkably reduced, the system can conduct off-line or real-time analysis through videos collected by portable devices such as a mobile phone and a common camera, and the analysis efficiency is improved. And the practicability and the universality of the system are greatly improved.

The invention discloses a multifunctional comprehensive strength training device, and belongs to a teaching device in the technical field of physical education training devices. According to the technical scheme, a mounting support and a multi-posture supporting assembly are fixedly connected with the upper ends of two supporting frames; one end of the tension driving device is connected with the winder, and the other end of the tension driving device is provided with the pulling assembly. A horizontal reciprocating assembly is arranged in the tension driving device, a hinged connecting rod set is arranged in the multi-posture supporting assembly, and the output end of the horizontal reciprocating assembly is connected with the hinged connecting rod set and used for posture change of the multi-posture supporting assembly. According to the multifunctional comprehensive strength training device, when an operator carries out strength pulling training on the multi-posture supporting assembly, the sitting posture of the body can be supported and assisted by strength, and therefore safe and convenient strength training can be carried out on trainees with small strength and poor strength control ability.

The invention discloses basketball throwing equipment for physical education, and particularly relates to the field of physical education, the basketball throwing equipment for physical education comprises a supporting seat, a backboard is arranged above the supporting seat, two supporting plates are fixedly connected to the front surface of the backboard, a collecting net bag is fixedly connected to the bottom surfaces of the two supporting plates, a basket is fixedly connected to the front surface of the backboard, and the basket is fixedly connected to the basket. A net barrel bag is fixedly connected to the bottom face of the basket, a base is fixedly connected to the bottom face of the supporting seat, a mechanism box is arranged on the front face of the supporting seat, a supporting cover is fixedly embedded in the right side face of the mechanism box, an electric push rod is fixedly installed on the inner side wall of the supporting cover, and a push plate is fixedly connected to the telescopic end of the electric push rod. According to the basketball throwing training device, basketballs can be discharged through the front end of the mechanism box, sports personnel can conveniently take the basketballs, then the device is provided with a structure for collecting the thrown basketballs, the basketballs can be continuously thrown for training, and the basketball throwing training effect during physical education is improved.

The invention relates to the technical field of data management, in particular to an intelligent physical education data management method and system. The method comprises the following steps: collecting motion video stream data and inertial motion data of a student; uniformly mapping a camera coordinate system of the video stream data and a body coordinate system of the inertial motion data to a world coordinate system; a state space model based on extended Kalman filtering is constructed, and a corrected student three-dimensional skeleton attitude sequence is obtained through iterative calculation; calculating an optimal matching path and a similarity distance between the three-dimensional skeleton posture sequence of the student and the standard action template sequence; and according to the similarity distance, generating a quantized action score. According to the invention, by using the extended Kalman filtering algorithm, integral drift is eliminated while the anti-shielding capability is ensured; meanwhile, an improved dynamic time warping algorithm is introduced, standardized and objective scoring of the motion quality is achieved, and the scientificity, safety and individuation level of physical education are effectively improved.

The invention belongs to the field of sports equipment, and relates to campus sports detection equipment, in particular to an early warning accurate timing system for a long-distance running scene in smart sports, which comprises a data intelligent body box, and a wearable heart rate device and a Bluetooth working module are arranged in the data intelligent body box; a six-axis acceleration sensor and a heart rate sensor are arranged in the wearable heart rate equipment; the system further comprises an intelligent Bluetooth base station. The sports safety is detected in real time, the sports risk is greatly reduced, and sports casualties are avoided. And safe dead-angle-free real-time detection is realized. The whole playground is covered through two intelligent Bluetooth base stations with 90-degree teaching signal coverage, heart rate movement data of each student is detected in real time, the heart rate and heart rate abnormity early warning can be displayed in real time in a test, an individual is prompted through bracelet vibration when an abnormity occurs, a physical education teacher is prompted through a tablet computer, and a person on the playground is prompted through Bluetooth sound equipment. And human intervention physical education is carried out.

The invention discloses a college student sports course selection recommendation method based on big data analysis, and relates to the technical field of data analysis, and the method comprises the steps: obtaining the physical measurement feature data of a jth student; acquiring course feature data of the ith physical education course; inputting the course feature data and the body measurement feature data into a pre-configured convolutional neural network to predict an estimated average heart rate interval when the jth student participates in the ith physical education course; according to the estimated heart rate interval and the standard heart rate interval, whether the ith physical education course is matched with the jth student or not is judged, and if yes, the ith physical education course serves as a first recommended physical education course selection course of the jth student; if not, i = i + 1; repeating the steps 1-4 until j is equal to N, and stopping circulation; according to the invention, full utilization of educational resources is guaranteed, and the risk of each student participating in the corresponding physical education course is minimized.

The invention relates to the field of physical training, in particular to a multifunctional track and field training device for physical education, which comprises a track assembly, a driving assembly mounted on the track assembly, a traction assembly mounted on the driving assembly and a pedal assembly mounted on the traction assembly, the driving assembly is used for driving the traction assembly and the corresponding pedal assembly to move on the track assembly, and the traction assembly is used for fixing the pedal assembly to any position on the track assembly; the track assembly is provided with a first transmission assembly, and the output end of the first transmission assembly is provided with a measuring assembly. By means of the measuring assembly integrated on the starting block, the distance between the starting block and the starting line can be measured when the starting block is installed, under long-term use, it can be guaranteed that the starting block is installed at the same position when training is conducted every time compared with the mode that the starting block is installed with a foot as a measuring reference, and the training efficiency is improved. The long-term muscle memory of athletes is favorably cultivated, and the training level is improved.

The invention discloses ball serving equipment for physical education training, belongs to the technical field of physical training, and aims to solve the problems that continuous feeding is interrupted and impurity particles on the surface of a tennis ball splash to a training area at a relatively high speed due to friction clamping stagnation of the tennis ball in a funnel due to a stacking effect. The device comprises a moving vehicle and an adjusting seat arranged on one side of the moving vehicle, a launching cylinder is fixedly installed on one side of the adjusting seat, and a launching piece is arranged at one end of the launching cylinder. Through the double-pushing design that the launching piece is in linkage with an elastic jacking piece, the phenomena of accumulation clamping stagnation and bridging of tennis balls in a discharging piece are effectively broken through; in the ball pushing process of the pushing piece, impurities on the surfaces of the tennis balls are separated under the action of extrusion and friction and are collected into the collecting box in a centralized mode through the discharging holes, and the safety accidents that the impurities splash along with high-speed tennis ball launching, skin scratching, foreign matter entering eyes and the like are avoided.