C-I Moon1, O Lee1. 1. Department of Medical IT Engineering, College of Medical Sciences, Soonchunhyang University, Asan, Korea.
Abstract
BACKGROUND/ PURPOSE: The aim of this study was to evaluate skin condition, expensive equipment is required, continuous skin care is difficult. Therefore, we obtain the skin image using mobile camera, and propose a new algorithm that easily and simply extracts skin features. METHODS: We analyze skin features, extracting the wrinkle length, cell area, and the number of cells. To get accurate skin features, we obtain a new skin binary image, and apply Watershed segmentation to it. So, we improve the accuracy of skin analysis. Therefore, we compare and analyze the degree of matching distribution of wrinkles, the shape of the cell, etc., using similarity between the ground truth and the proposed algorithm result image. RESULTS: We extract skin surface features using a mobile camera image, and verify the change in skin features with age. Also, we demonstrate the superiority of the proposed algorithm through the similarity between the ground truth and proposed result image. CONCLUSION: The proposed method in this study shows that the skin surface can be quantitatively evaluated by the similarity with ground truth. We also propose a method to diagnose and manage individual skin condition using a mobile camera in real life.
BACKGROUND/ PURPOSE: The aim of this study was to evaluate skin condition, expensive equipment is required, continuous skin care is difficult. Therefore, we obtain the skin image using mobile camera, and propose a new algorithm that easily and simply extracts skin features. METHODS: We analyze skin features, extracting the wrinkle length, cell area, and the number of cells. To get accurate skin features, we obtain a new skin binary image, and apply Watershed segmentation to it. So, we improve the accuracy of skin analysis. Therefore, we compare and analyze the degree of matching distribution of wrinkles, the shape of the cell, etc., using similarity between the ground truth and the proposed algorithm result image. RESULTS: We extract skin surface features using a mobile camera image, and verify the change in skin features with age. Also, we demonstrate the superiority of the proposed algorithm through the similarity between the ground truth and proposed result image. CONCLUSION: The proposed method in this study shows that the skin surface can be quantitatively evaluated by the similarity with ground truth. We also propose a method to diagnose and manage individual skin condition using a mobile camera in real life.