Acupuncture Points Localization Algorithm about Bell′s Palsy on Facial Infrared Images

ZHAO Yang,ZHANG Tao,LIAN Qiu-sheng
DOI: https://doi.org/10.3969/j.issn.1000-1220.2012.07.043
2012-01-01
Abstract:This paper uses infrared image processing into traditional Chinese medicine diagnosis,put forward a kind of positioning algorithm to automatic positioning facial acupuncture points,and provide an objective and standardization basis for Bell ′s palsy to choose acupuncture.This method firstly distinguishes the foreground(namely head images) and background use head threshold,divide the facial features in accordance with the rule of "three court five eyes".Then use MinimumEigenvalue operator detect facial features′ corner,use Log operator detect facial features′ edges,integrated apply the corner and edge information,positioning the position of each facial feature;Finally,use the facial features position as a reference coordinate,complete facial acupuncture points positioning.After the validation of many facial infrared images,this method′s average accuracy is over 94%,the accuracy of ren zhong is close to 100%.
What problem does this paper attempt to address?