The invention provides a parallel bar support forward
movement detection method, device and equipment and a medium, and the method comprises the steps: synchronously collecting multi-
source data when a trainer executes a parallel bar support forward movement, preprocessing the collected multi-
source data, then achieving data alignment and
coordinate mapping based on a
timestamp, forming a standardized
data set, and storing the standardized
data set in a
database; the multi-
source data at least comprises joint motion visual data, two-
hand pressure data and
muscle group electrical activity data; based on the standardized
data set, a plurality of quantitative indexes representing the motion quality are obtained through calculation, and the quantitative indexes at least comprise the shoulder
elbow joint angle, the body displacement speed, the core stable duration, the two-
hand pressure balance degree and the
muscle group cooperation efficiency; and inputting the plurality of quantitative indexes into a pre-trained
deep learning model, comparing the quantitative indexes with a standard threshold value, and triggering risk grading early warning when at least one quantitative index exceeds the range of the standard threshold value, so as to solve the problems of subjective evaluation mode, fragmentation of
monitoring data, missing of risk early warning and no individuality of a training scheme.