Automatic wrinkles detection on face image

International Conference on Information Technology and Digital Applications 2018

计算机科学;无线电电子学

Muhimmah, Izzati^1 ; Abriyani, Fadhillah^1 ; Kurniawardhani, Arrie^1

Department of Informatics, Islamic University of Indonesia, Yogyakarta, Indonesia^1

关键词: Canny filter;    Colour image segmentation;    Face images;    Image processing technique;    Laplacian of gaussian filters;    Phase image;    Validation data;   

Others  :  https://iopscience.iop.org/article/10.1088/1757-899X/482/1/012026/pdf
DOI  :  10.1088/1757-899X/482/1/012026

学科分类：计算机科学（综合）

来源: IOP

【 摘 要 】

Image processing techniques are beneficial for reconstructing and processing an image in resulting different images. One of its benefits is it can be used for detecting wrinkles on a face. Line characteristic of wrinkles can be detected also with the support of the features that had been provided. The purpose of this research is to design a system that could detect the exact location of the line of wrinkles which occur from one face image. Generally, the step that required for this research is the HSV colour image segmentation step and K-Means Clustering for dividing skin object and non-skin object. The second step is phase image improvement step with Laplacian of Gaussian filter to sharpening line of wrinkles on image face, and the last step is feature extraction which used for taking wrinkles. After all the steps had been done, each of the wrinkle's areas are detected with Canny filter and eliminated based on Area characteristic. From research experiment, the result validation data of system based on an expert are 40.6% for sensitivity, 96.9% for specificity, and 94.2% for accuracy, with the performance of the system 5,247 seconds.