Method, system and embedded device for limb rehabilitation training

Abstract

The invention discloses a limb rehabilitation training method which is applied to embedded equipment integrated with a limb rehabilitation training system. The method comprises the steps that a deep neural network is trained in advance according to human body standard exercise data; when a patient carries out limb rehabilitation training, a limb movement video, collected by a video acquisition module in real time, of limb rehabilitation training of the patient is received; spatial-temporal features are extracted according to the limb movement video, the spatial-temporal features are input into the pre-trained deep neural network, and a limb movement recognition result of limb rehabilitation training of the patient is obtained; the limb movement of the patient is evaluated and corrected according to the recognition result and a rehabilitation training scheme. The limb rehabilitation training method is high in recognition precision when the limb movement of the patient is recognized. The limb rehabilitation training system is integrated with the embedded equipment, so that the equipment cost and the operation difficulty are reduced, and the portability is improved. The invention further discloses the limb rehabilitation training system and the embedded equipment which have the advantages the same as those of the method.

Description

technical field [0001] The invention relates to the field of limb training, in particular to a limb rehabilitation training method. The invention also relates to a limb rehabilitation training system and an embedded device. Background technique [0002] The current rehabilitation treatment of limb function training is mainly to promote the recovery of limb movement ability in the hospital through the assistance of some training equipment or the manipulation of the therapist. However, in my country, due to the wide distribution of patients and the lack of rehabilitation institutions, many patients are unable to undergo modular and institutional rehabilitation training. In order to solve this problem, people began to use human-computer interaction to carry out limb rehabilitation training, guide patients to complete rehabilitation training through sensors and computers integrated with limb rehabilitation training systems, and correct unreasonable training actions. [0003] S...

AI Technical Summary

Problems solved by technology

Because there may be differences in height, body shape and movement speed between the person and the patient who entered the standard action template, these differences will have a greater impact on the recognition accuracy, resulting in low recognition accuracy and poor user experience

In addition, in the existing technical solutions, since the rehabilitation training system is installed on a computer, the cost of the equipment is relatively high, and the computer operation level of the patient is required to be relatively high.

Images