Objective To propose the ideas and methods for processing and analysis of the images of Guasha marks,which can provide references for digital research of Guasha.Methods Based on computer vision technology,aiming at the processing and a-nalysis of Guasha images in patients with low back pain,a series of methods such as image preprocessing,feature extraction and screening,and analysis of Guasha image were proposed and verified by experiments.Results The image preprocessing method can achieve color correction of the original image and accurate segmentation of the back area.The feature extraction method extracted 94 color and texture features of low back pain patients,and screened 20 key features.The correlation analysis identified key fea-tures associated with different syndromes.Conclusion The use of computer vision methods can better complete the preprocessing,feature recognition and extraction,analysis and understanding of Guasha images,which can provide support for the digitization and objectification study of Guasha.
关键词
计算机视觉/腰背痛/刮痧/痧象特征
Key words
computer vision/low back pain/Guasha/features of Guasha marks
维普
万方数据