Rope skipping counting method and device and computer storage medium

Abstract

The invention discloses a rope skipping counting method and device and a computer storage medium, and relates to the field of artificial intelligence. The rope skipping counting method comprises the steps of obtaining a rope skipping video; detecting a plurality of key limb parts in the first frame of image in the rope skipping video through a human body key point detection model, and outputting athermodynamic diagram corresponding to the key limb parts; setting the coordinate of the position with the maximum thermodynamic value in each thermodynamic diagram as a first key point coordinate; obtaining a plurality of second key point coordinates of each key limb part in a plurality of second frame images in the rope skipping video, wherein the second key point coordinates are obtained by tracking the previous frame image through a tracking algorithm or through a human body key point detection model; and associating all the first key point coordinates and all the second key point coordinates with the height change caused by each rope skipping action to obtain rope skipping counting data. Statistical efficiency is not reduced while rope skipping counting accuracy can be improved through the human body key point detection model in combination with a tracking algorithm.

Description

technical field [0001] The invention relates to the field of artificial intelligence, in particular to a method, device and computer storage medium for counting by skipping rope. Background technique [0002] Rope skipping is one of the sports that people often engage in. Rope skippers need to count during the rope skipping process, that is, they need to know the amount of exercise as an exercise index. The counting in the rope skipping process usually adopts two counting methods, one is to count by electronic counting rope skipping; The recognition accuracy rate is low, resulting in counting errors; while the manual statistical method requires one-to-one, the statistical efficiency is low, and it is easy to make statistical errors when the statistician is distracted. Therefore, although the two statistical methods can obtain counting data in real time, they are prone to counting errors, resulting in inaccurate counting. Contents of the invention [0003] The present inv...

AI Technical Summary

Problems solved by technology

The counting in the rope skipping process usually adopts two counting methods, one is to count by electronic counting rope skipping; The recognition accuracy rate is low, resulting in counting errors; while the manual statistical method requires one-to-one, the statistical efficiency is low, and it is easy to make statistical errors when the statistician is distracted

Therefore, although the two statistical methods can obtain counting data in real time, they are prone to counting errors, resulting in inaccurate counting

Images