论文阅读 [TPAMI-2022] Deep Photometric Stereo Networks for Determining Surface Normal and Reflectances

论文阅读 [TPAMI-2022] Deep Photometric Stereo Networks for Determining Surface Normal and Reflectances

论文搜索(studyai.com)

搜索论文: Deep Photometric Stereo Networks for Determining Surface Normal and Reflectances

搜索论文: http://www.studyai.com/search/whole-site/?q=Deep+Photometric+Stereo+Networks+for+Determining+Surface+Normal+and+Reflectances

关键字(Keywords)

Estimation; Shape; Computational modeling; Neural networks; Pattern analysis; Deep learning; Photometric stereo; surface normal; bidirectional reflectance distribution functions (BRDFs); deep learning

机器视觉

光度立体

摘要(Abstract)

This article presents a photometric stereo method based on deep learning.

本文提出了一种基于深度学习的光度立体方法。.

One of the major difficulties in photometric stereo is designing an appropriate reflectance model that is both capable of representing real-world reflectances and computationally tractable for deriving surface normal.

光度立体中的一个主要困难是设计一个合适的反射模型，该模型既能表示真实世界的反射，又能在计算上易于处理，用于推导表面法线。.

Unlike previous photometric stereo methods that rely on a simplified parametric image formation model, such as the Lambert’s model, the proposed method aims at establishing a flexible mapping between complex reflectance observations and surface normal using a deep neural network.

与以前依赖简化参数成像模型（如Lambert模型）的光度立体方法不同，该方法旨在使用深度神经网络在复反射观测值和表面法线之间建立灵活的映射。.

In addition, the proposed method predicts the reflectance, which allows us to understand surface materials and to render the scene under arbitrary lighting conditions.

此外，所提出的方法可以预测反射率，这使我们能够理解表面材质，并在任意光照条件下渲染场景。.

As a result, we propose a deep photometric stereo network (DPSN) that takes reflectance observations under varying light directions and infers the surface normal and reflectance in a per-pixel manner.

因此，我们提出了一种深度光度立体网络（DPSN），它在不同的光方向下进行反射观测，并以每像素的方式推断表面法线和反射。.

To make the DPSN applicable to real-world scenes, a dataset of measured BRDFs (MERL BRDF dataset) has been used for training the network.

为了使DPSN适用于真实场景，已使用测量的BRDF数据集（MERL BRDF数据集）来训练网络。.

Evaluation using simulation and real-world scenes shows the effectiveness of the proposed approach in estimating both surface normal and reflectances…

通过仿真和真实场景的评估，表明了该方法在估计表面法线和反射方面的有效性。。.

作者(Authors)

[‘Hiroaki Santo’, ‘Masaki Samejima’, ‘Yusuke Sugano’, ‘Boxin Shi’, ‘Yasuyuki Matsushita’]