We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
探索图像分析中适用于评估皮肤毛孔的指标.
- Authors
黄冰婷; 马 琳; 孙丽丽; 简苑晴; 张晓洁; 李 静
- Abstract
Facial pores were assessed by subjective assessment and image analysis, and the results were comparatively analyzed to explore the applicability as an assessment indicator for evaluating pores. 184 subjects participated and completed the test. The subjects' facial images were captured by VISIA®, and the images were analyzed by VISIA® and Image-Pro® Plus (hereinafter referred to as IPP) and subjective assessed by the investigator, the results were statistically analyzed. The results show that the pore size and density parameters of VISIA® and Image-Pro® Plus are correlated by two-correlation coefficient analysis and linear regression analysis. The pore counts (VISIA) is positively correlated with the number of pores (IPP) (P<0.001). The pore score (VISIA) is positively correlated with pore area percentage (IPP), mean pore area (IPP), and mean pore diameter (IPP) (P<0.001), with a lower correlation between the pore score (VISIA) and mean pore diameter (IPP) (the R2 of linear regression analysis is 0.270 2). There is a correlation between the subjective assessment results and the pore size and density parameters analyzed by VISIA® and Image-Pro® Plus. The subjective assessment result about pore size score is positively correlated with the pore score (VISIA), also with the pore area percentage (IPP) and the mean pore area (IPP) (P<0.001). The subjective assessment result about pore density score is positively correlated with the pore counts (VISIA), also with the number of pores (IPP) (P<0.001). In conclusion, the pore count (from VISIA®) and pore score (from VISIA®) and the number of pores (from Image-Pro® Plus), pore area percentage (from Image-Pro® Plus) and mean pore area (from Image-Pro® Plus) are suitable as the assessment indicators for evaluating pore features.
- Publication
Detergent & Cosmetics, 2024, Vol 47, Issue 5, p30
- ISSN
1006-7264
- Publication type
Article
- DOI
10.3969/j.issn.1006-7264.2024.05.007