Accurate Photometric Stereo Using Four Surface Normal Approximations

2006 ◽  
Author(s):  
O. Ikeda
Keyword(s):  
Photometric Stereo ◽  
Normal Approximations ◽  
Surface Normal

scholarly journals Pay Attention to Devils: A Photometric Stereo Network for Better Details

2020 ◽  
Author(s):  
Yakun Ju ◽  
Kin-Man Lam ◽  
Yang Chen ◽  
Lin Qi ◽  
Junyu Dong
Keyword(s):  
Neural Network ◽  
State Of The Art ◽  
Photometric Stereo ◽  
Normal Map ◽  
Surface Normal ◽  
3D Surface ◽  
Surface Reconstructions ◽  
Higher Weights

We present an attention-weighted loss in a photometric stereo neural network to improve 3D surface recovery accuracy in complex-structured areas, such as edges and crinkles, where existing learning-based methods often failed. Instead of using a uniform penalty for all pixels, our method employs the attention-weighted loss learned in a self-supervise manner for each pixel, avoiding blurry reconstruction result in such difficult regions. The network first estimates a surface normal map and an adaptive attention map, and then the latter is used to calculate a pixel-wise attention-weighted loss that focuses on complex regions. In these regions, the attention-weighted loss applies higher weights of the detail-preserving gradient loss to produce clear surface reconstructions. Experiments on real datasets show that our approach significantly outperforms traditional photometric stereo algorithms and state-of-the-art learning-based methods.

Design and Calibration of a 3D High-Resolution Surface Profiling System Using Photometric Stereo

2015 ◽  
Author(s):  
Boren Li ◽  
Tomonari Furukawa
Keyword(s):  
High Resolution ◽  
Light Intensity ◽  
Target Surface ◽  
Photometric Stereo ◽  
Radiometric Calibration ◽  
Geometric Calibration ◽  
Surface Profiling ◽  
Surface Normal ◽  
3D Methods ◽  
Intensity Calibration

This paper presents the design and calibration of a 3D high-resolution surface profiling system using photometric stereo (PS). This system is mainly composed of a high resolution DSLR camera with a macro lens facing perpendicularly to the target surface, and several LEDs tilting towards the surface constrained by a light fixture. With each LED turned on at a time to create one lighting direction, the camera fixed at the same position captures an image. PS with surface normal integration (SNI) are then performed to reconstruct the surface in 3D. Methods of four calibrations for the developed system are proposed to achieve better accuracy, which are the camera radiometric calibration, the camera geometric calibration, the light direction calibration and the light intensity calibration. Experiments have demonstrated that the developed system with the calibration processes could achieve the accuracy in the order of 10 microns.

scholarly journals Deep Photometric Stereo Networks for Determining Surface Normal and Reflectances

2020 ◽  
pp. 1-1
Author(s):  
Hiroaki Santo ◽  
Masaki Samejima ◽  
Yusuke Sugano ◽  
Boxin Shi ◽  
Yasuyuki Matsushita
Keyword(s):  
Photometric Stereo ◽  
Surface Normal

MPS-Net: Learning to recover surface normal for multispectral photometric stereo

2020 ◽  
Vol 375 ◽  
pp. 62-70 ◽  
Author(s):  
Yakun Ju ◽  
Lin Qi ◽  
Jichao He ◽  
Xinghui Dong ◽  
Feng Gao ◽  
...  
Keyword(s):  
Photometric Stereo ◽  
Surface Normal

scholarly journals Estimation of Surface Normal Vector Based on Majority Voting using Multi-Light Photometric Stereo

2008 ◽  
Vol 128 (10) ◽  
pp. 1567-1573
Author(s):  
Hisao Fukumoto ◽  
Hiroshi Terashima ◽  
Masashi Ohchi ◽  
Tatsuya Furukawa
Keyword(s):  
Photometric Stereo ◽  
Majority Voting ◽  
Normal Vector ◽  
Surface Normal

Examining the uncertainty of the recovered surface normal in three light photometric stereo

2007 ◽  
Vol 25 (7) ◽  
pp. 1073-1079 ◽  
Author(s):  
Jiuai Sun ◽  
Melvyn Smith ◽  
Lyndon Smith ◽  
Abdul Farooq
Keyword(s):  
Photometric Stereo ◽  
Surface Normal
Sign in / Sign up

Export Citation Format

Share Document