PISR: Polarimetric Neural Implicit Surface Reconstruction for Textureless and Specular Objects

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Chen, Guangcheng, He, Yicheng, He, Li, Zhang, Hong
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917781906128896
author Chen, Guangcheng
He, Yicheng
He, Li
Zhang, Hong
author_facet Chen, Guangcheng
He, Yicheng
He, Li
Zhang, Hong
contents Neural implicit surface reconstruction has achieved remarkable progress recently. Despite resorting to complex radiance modeling, state-of-the-art methods still struggle with textureless and specular surfaces. Different from RGB images, polarization images can provide direct constraints on the azimuth angles of the surface normals. In this paper, we present PISR, a novel method that utilizes a geometrically accurate polarimetric loss to refine shape independently of appearance. In addition, PISR smooths surface normals in image space to eliminate severe shape distortions and leverages the hash-grid-based neural signed distance function to accelerate the reconstruction. Experimental results demonstrate that PISR achieves higher accuracy and robustness, with an L1 Chamfer distance of 0.5 mm and an F-score of 99.5% at 1 mm, while converging 4~30 times faster than previous polarimetric surface reconstruction methods.
format Preprint
id arxiv_https___arxiv_org_abs_2409_14331
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle PISR: Polarimetric Neural Implicit Surface Reconstruction for Textureless and Specular Objects
Chen, Guangcheng
He, Yicheng
He, Li
Zhang, Hong
Computer Vision and Pattern Recognition
Neural implicit surface reconstruction has achieved remarkable progress recently. Despite resorting to complex radiance modeling, state-of-the-art methods still struggle with textureless and specular surfaces. Different from RGB images, polarization images can provide direct constraints on the azimuth angles of the surface normals. In this paper, we present PISR, a novel method that utilizes a geometrically accurate polarimetric loss to refine shape independently of appearance. In addition, PISR smooths surface normals in image space to eliminate severe shape distortions and leverages the hash-grid-based neural signed distance function to accelerate the reconstruction. Experimental results demonstrate that PISR achieves higher accuracy and robustness, with an L1 Chamfer distance of 0.5 mm and an F-score of 99.5% at 1 mm, while converging 4~30 times faster than previous polarimetric surface reconstruction methods.
title PISR: Polarimetric Neural Implicit Surface Reconstruction for Textureless and Specular Objects
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2409.14331