Acupuncture manipulation recognition method, device, system and medium based on tactile sensor

Abstract

The present invention relates to an acupuncture manipulation recognition method, device, system and medium based on a tactile sensor. The method includes filtering collected data, window segmentation, window combination, and window screening, and extracting time-domain feature structures of data. A feature data set reflecting the acupuncture technique is obtained, and finally a neural network is used for identification. The invention aims at performing window segmentation on piezoelectric signal sequences collected by piezoelectric sensors designed based on PVDF piezoelectric films during acupuncture treatment by doctors. Then, combine the data of each window, analyze the time-domain features, and finally form a feature data set, and then extract features, solve the influence of light and the occlusion of gestures by objects, etc. Recognition and classification of complex gestures such as stabbing, which are composed of several sub-actions in a certain time sequence.

Description

technical field [0001] The present invention relates to the technical field of neural networks, in particular to a method, device, system and medium for recognizing acupuncture manipulations based on tactile sensors. Background technique [0002] Acupuncture manipulation is an important part of acupuncture and moxibustion medicine. Under the guidance of TCM theory, the needles are inserted into specific acupoints of the patient at a certain angle and depth, and the acupoints of the human body are stimulated by lifting, inserting or twisting to activate the acupuncture points. The internal reaction of the human body, so as to achieve the purpose of treating diseases. [0003] At present, gesture recognition based on computer vision is a research hotspot, such as video games and remote surgery technology. Many current gesture methods are based on extracting features, which typically classify gestures by capturing attributes such as shape, motion cues, or human skeleton. Howe...

AI Technical Summary

Problems solved by technology

However, in the process of gesture recognition, the predecessors often cared about gesture segmentation based on optical flow characteristics, gesture recognition with time series features, and the dimensionality of gesture recognition technology, lacking a quantitative description of subtle movements.

In addition, due to the influence of light and the occlusion of gestures by objects, gesture recognition will be limited, and there is a lack of research on the identification and classification of subtle techniques.

Images