Facial wrinkle detection and scoring mechanism based on adaptive scale hybrid Hessian filter
A facial wrinkle detection method and scoring mechanism based on adaptive scale hybrid Hessian filter(ASHHF)are proposed to solve the problem of lacking quantitative evaluation method for the whole face wrinkles.According to the age of the subject,the size(σ)range and step parameters of the Hessian filter were adaptively adjusted,and the high-resolution facial images were filtered.Based on 81 facial feature points,the face and background features were removed from the filtering results,and only facial wrinkles were kept.The wrinkle detection results of different depths were marked in the original image with different colors.Finally,the score was calculated to quantify the degree of skin aging.Taking wrinkles marked by professional doctors as a reference,the detection results of 50 subjects(aged between 22 and 65)show that compared with the traditional fixed scale hybrid Hessian filter(FSHHF),the detection accuracy of the ASHHF algorithm was increased by 68.57%on average,and the running time was shortened by 26.26%on average.The result of scoring mechanism is consistent with the changing trend of detection accuracy.In summary,the detection method in this paper can accurately and quickly detect the distribution,depth and width information of facial wrinkles,and the proposed scoring mechanism can scientifically reflect the degree of skin aging of subjects.It is expected to provide a quantitative evaluation means for anti-aging methods such as cosmetics,medical cosmetology,etc.
国家科技期刊平台
NSTL
万方数据
维普
