Related Experiment Video
Updated: Nov 15, 2025
06:08
Quantitative Visualization and Detection of Skin Cancer Using Dynamic Thermal Imaging
Published on: May 5, 2011
17.0K
Skin tactile surface restoration using deep learning from a mobile image: An application for virtual skincare
Donghyun Kim1, Myeongseob Ko1, Kwangtaek Kim1,2
1Haptic Engineering Research Laboratory, Department of Information and Telecommunication Engineering, Incheon National University, Incheon, Republic of Korea.
Summary
This study introduces a deep learning method to reconstruct 3D skin tactile surfaces from mobile images, overcoming illumination challenges for virtual skincare applications.
Area of Science:
- Biomedical Engineering
- Computer Vision
- Dermatology
Background:
- Reconstructing 3D skin tactile surfaces from mobile images is crucial for virtual skincare and remote palpation.
- Variable illumination conditions present a significant challenge for accurate tactile surface reconstruction.
Purpose of the Study:
- To develop a deep learning-based scheme for tactile reconstruction from mobile skin images.
- To address challenges posed by light distortion and improve 3D skin tactile surface generation.
Main Methods:
- A novel deep learning approach utilizing a conditional generative adversarial network (cGAN) for light distortion removal.
- 3D tactile surface generation based on image gradients after distortion correction.
Main Results:
- The proposed method effectively removes light distortion, restoring tactile properties.
- Demonstrated superior performance in illumination-free image restoration and 3D surface reconstruction compared to existing methods.
Conclusions:
- Deep learning enables complete restoration of illumination-distorted tactile properties with smaller training datasets.
- Achieved precise 3D skin tactile surface reconstruction, enabling remotely touchable interfaces for virtual skincare.
