Fitness movement error correction method and system based on deep learning and image recognition

Abstract

The invention discloses a fitness movement error correction method based on deep learning and image recognition. The method comprises the steps of: receiving each fitness image intercepted from a fitness video, recording time corresponding to each fitness image, and preprocessing each fitness image; inputting the pre-processed fitness images into a human body joint recognition model, and outputting each human body joint included in the fitness images and position coordinates thereof through calculation of the human body joint recognition model; carrying out orderly line connection on each joint according to a coordinate position to obtain human body skeleton graphs; and carrying out comparison analysis on the human body skeleton graphs according to standard fitness movement, and outputtingnon-standard fitness images to realize error correction on fitness movement. The invention also discloses a fitness movement error correction system based on deep learning and image recognition. Thesystem and the method can truly and effectively solve the problem of difficult comparison between fitness movement with or without a musical rhythm.

Description

technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a fitness action error correction method and system based on deep learning image recognition. Background technique [0002] With the development of economy and the improvement of people's living standards, people's fitness awareness is gradually enhanced, and people pay more and more attention to the improvement of physical fitness. However, due to factors such as the limitation of fitness venues and the high cost of fitness coaches, many bodybuilders do not have enough opportunities to learn professional fitness movements, so the fitness exercise of bodybuilders has not achieved the expected results. [0003] For the current fitness exercise, it can be divided according to the rhythm of music and the rhythm without music. For example, most of the single-function and multi-function fitness equipment in the gym are fitness exercises without music rhythm, and t...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to provide a method and system for error correction of fitness movements based on deep learning image recognition, so as to solve the problem that bodybuilders lack instructive suggestions on fitness movements

Images