Limb movement recognition and comparison method based on artificial intelligence

Abstract

The invention provides a limb movement recognition and comparison method based on artificial intelligence, and the method comprises the following steps: S1, collecting standard limb movement data, andtraining a standard movement model; s2, collecting body movement data to be compared; s3, inputting to-be-compared body movement data into the standard movement model, and respectively extracting body key point data in the standard body movement data and the to-be-compared body movement data; and S4, according to the standard limb movement data and the limb key point data in the to-be-compared limb movement data, calculating the deviation between the to-be-compared limb movement and the standard limb movement, and scoring the to-be-compared limb movement. The body movement recognition and comparison cost is low, the convenience is high, the efficiency is high, and the body movement recognition and comparison method can be applied to AI fitness and other scenes.

Description

technical field [0001] The invention relates to the technical field of artificial intelligence, in particular to an artificial intelligence-based body movement recognition and comparison method. Background technique [0002] The technologies currently used in sports and fitness body movement gesture recognition and teaching guidance mainly fall into the following categories: [0003] 1. Use accelerometers, gyroscopes, and magnetometers to recognize body posture, movement, and speed. Although this method has the ability to quickly and accurately measure the speed of human body parts, but whether it is a coach, a teacher, or a student, it is necessary to wear the corresponding equipment. The disadvantage is that it is inconvenient, expensive, and difficult to popularize; [0004] 2. Use special close-fitting clothing and the sensors on it to perceive and recognize body postures and movements. The use of special clothing can detect the wearer's movement, breathing, pressure, ...

AI Technical Summary

Problems solved by technology

This method also needs to be equipped with corresponding equipment, whether it is on the coach and teacher side or the student side. The course production process needs it, and the student practice process also needs it; the first disadvantage is that the cost is high; the second disadvantage is that all teaching content needs to be re-shot Production; the third disadvantage is that the convenience is not good, and the scene is limited; the fourth disadvantage is that the current mainstream technology intelligently recognizes 15-20 bone points, and it is very affected by the light and background at the time. It is not very practical in sports and fitness, and the experience is not good. , poor experience in somatosensory games, not widely used;

This method is more traditional, and the disadvantage is that it needs to invest a lot of manpower and material resources for shooting, production and identification, the cost is very high, and the efficiency is very low

[0007] To sum up, at present, it is still difficult to fully understand the movement standardization of students in the whole process of fitness movements at low cost and high efficiency, and it is also impossible to develop application models based on this technology to improve the fun and sense of participation in online exercise and fitness, resulting in online It is difficult to achieve good results in sports and fitness, and it is also difficult to persist

Images