Human body back acupoint recognition method and device and computer storage medium

Abstract

The invention relates to a human body back acupoint recognition method and device and a computer storage medium. The recognition method comprises the following steps: performing single-branch convolution calculation on a back feature map of a target image by using an improved OpenPose algorithm to obtain a back acupuncture point recognition model; acquiring a back image data set by using the human body back image; training a back acupoint recognition model by using the back image data set to obtain an improved back acupoint recognition model; the to-be-recognized back image is input into the improved back acupoint recognition model, the back acupoints of the to-be-recognized back image are obtained, intelligent recognition of the back acupoints of the human body can be achieved without manual participation, and intelligentization of recognition of the back acupoints of the human body is achieved.

Description

technical field [0001] The invention relates to the technical field of identification of acupuncture points on the back of the human body, in particular to a method for identifying acupuncture points on the back of the human body, an identification device and a computer storage medium. Background technique [0002] In traditional Chinese medicine, meridian points play an important role in traditional Chinese medicine physical therapy such as acupuncture and moxibustion. With the development of intelligent medical care, the automation and intelligence of acupoint identification has become an inevitable development trend. At present, when the intelligent medical system performs acupoint identification, it requires TCM doctors to identify the basic acupoints, input the coordinates of the basic acupoints into the model, and then identify other acupoints according to the relative positional relationship between the acupoints, which largely depends on the For experienced doctors,...

AI Technical Summary

Problems solved by technology

At present, when the intelligent medical system recognizes acupoints, it requires doctors of traditional Chinese medicine to identify the basic acupoints, input the coordinates of the basic acupoints into the model, and then identify other acupoints according to the relative positional relationship between the acupoints, which largely depends on For experienced doctors, labor costs are high

Moreover, the number of acupoints that can be identified by existing acupoint identification methods is limited, and most of the identifiable acupoints are acupoints that can be recognized by ordinary people.

Images