Rope skipping counting method based on rope target extraction

Abstract

The invention discloses a rope skipping counting method based on rope target extraction. A rope skipping video is shot from an overlook angle, so that the difficulty of subsequently extracting people and ropes from a background is reduced, the establishment of a coordinate system is facilitated, and the rope skipping counting analysis is performed through a rope direction vector in the coordinate system in a later period. Besides, a video image is pre-processed, the resolution of the video image is controlled, then a color threshold value is set, when the color of a rope reaches the threshold value range, the rope is extracted and analyzed,so that the resolution is controlled, and the calculation speed is higher. Finally, the analysis process of the rope direction vector is very strict, a body front trigger and a body rear trigger are arranged, and by combining the operations of normal rope skipping, reverse rope skipping, rope skipping breaking, reverse shaking after breaking, crossing rope skipping after breaking, during the rope skipping process, the values of the body front trigger and the body rear trigger can change according to different logic, therefore, the rope skipping frequency is influenced, and the accuracy rate is higher.

Description

technical field [0001] The invention relates to the technical field of sports testing, in particular to a rope skipping counting method based on rope target extraction. Background technique [0002] For rope skipping, counting the number of times of rope skipping is usually carried out manually. When a large amount of counting is required, the manual method is not only inefficient, but also requires a lot of manpower. For example, in the physical examination, in order to improve the efficiency of counting, many people need to skip rope counting for different examinees at the same time, which wastes manpower. There are two ways for automatic counting. One is to realize the counting through the intelligent skipping rope detection device and the structural design of the skipping rope itself and its supporting products. The application of this method is relatively limited, and the specified rope must be used to complete the automatic counting. Counting cannot be done on any jum...

AI Technical Summary

Problems solved by technology

The biggest problem with the detection of rope skipping times based on rope skippers is the lack of rope processing, so the behavior of empty jumps cannot be reasonably distinguished. At the same time, the audio signal is greatly affected by the environment.

Patent CN112044046A uses deep learning to divide images into three categories for counting analysis. It has high requirements on the angle of training set and test set, and is more dependent on the model. The process complexity is high and the calculation speed is high.

Images