Uncovering Vein Pattern using Generative Adversarial Network

被引：4

作者：

Ma, Gehua ^{[1
]}

Wang, Biao ^{[1
]}

Tang, Chaoying ^{[1
]}

机构：

[1] Nanjing Univ Aeronaut & Astronaut, Coll Automat Engn, Nanjing, Peoples R China

来源：

ELEVENTH INTERNATIONAL CONFERENCE ON DIGITAL IMAGE PROCESSING (ICDIP 2019) | 2019年 / 11179卷

关键词：

vein pattern; deep learning; GAN; IMAGE QUALITY ASSESSMENT;

D O I：

10.1117/12.2539601

中图分类号：

O43 [光学];

学科分类号：

070207 ; 0803 ;

摘要：

Vein distribution is important in medical treatments. It could also be used for identity authentication(1). As a basic part of our body, the blood vessel has the merits of universality and distinctiveness. However, vein patterns are usually not visible in color images, which carries significant limitation. To address this limitation, we proposed a deep-learning-based method. Our method can uncover vein distributions from color images, help relieving pains to patients and widening the application scenarios of vein patterns. Experimental results showed that the proposed method has reliable performance and robustness in varying environments.

引用

页数：7