关注公众号,发现CV技术之美
期刊:Photonics 影响因子 2.536
专刊名:Advanced Photometric 3D Reconstruction and Beyond
学科领域:计算机视觉,光学,三维重建,深度学习,光度立体
专刊官网网址:https://www.mdpi.com/journal/photonics/special_issues/604639UE1N
截止日期:2023年9月30日
▍Guest Editors ▍
Yakun Ju, Department of Electronic and Information Engineering, The Hong Kong Polytechnic University, Hong Kong
Wuyuan Xie, Research Institute for Future Media Computing, Shenzhen University
Junyu Dong, The College of Computer Science and Technology, Ocean University of China
▍Topics ▍
Acquiring three-dimensional (3D) geometry from two-dimensional (2D) scenes is a fundamental problem in computer vision, which provides an opportunity to obtain a better understanding of the surrounding environment for machines. Over the past decade, 3D reconstruction has received much attention from the research community, due to its many potential applications, which are changing the way we live. Photometric reconstruction is a multidisciplinary field that combines optics and computer vision. It aims at modeling the relationship between surface geometry and irradiance, and leverages it to compute the 3D shape of a surface from single-view images.
Typically, a photometric algorithm works under specific circumstances due to the varying reflectance and global illumination effects. The recent advances in deep learning have rapidly found a place in this field and have been used to solve the above problems. However, many challenges still exist. Photometric stereo techniques still require specialized knowledge from researchers to properly address the problems of interest, for example, how to balance the performance of Lambertian and non-Lambertian surfaces; how to increase the robustness of photometric stereo methods that can be used in outdoor environments; how to combine the photometric method with the geometry method; how to accurately reconstruct the surfaces from normal maps, etc.
This Special Issue aims to gather recent advances in photometric 3D reconstruction and its applications. Topics of interest include, but are not limited to, the following:
photometric stereo
shapes from shading
shapes from polarization
multiview photometric stereos
applications with photometric 3D reconstruction
optical devices and systems for photometric reconstruction
surface reconstruction
normal map enhancement
lighting direction estimation
cloud points processing
other learning-based 3D reconstruction methods
END
欢迎加入「征稿」交流群👇备注:papers