arXiv:2505.21890v3 Announce Type: replace 
Abstract: Hyperspectral imaging (HSI) has been widely used in agricultural applications for non-destructive estimation of plant nutrient composition and precise quantification of sample nutritional elements. Recently, 3D reconstruction methods, such as Neural Radiance Field (NeRF), have been used to create implicit neural representations of HSI scenes. This capability enables the rendering of hyperspectral channel compositions at every spatial location, thereby helping localize the target object's nutrient composition both spatially and spectrally. However, it faces limitations in training time and rendering speed. In this paper, we propose Diffusion-Denoised Hyperspectral Gaussian Splatting (DD-HGS), which enhances the state-of-the-art 3D Gaussian Splatting (3DGS) method with wavelength-aware spherical harmonics, a Kullback-Leibler divergence-based spectral loss, and a diffusion-based denoiser to enable 3D explicit reconstruction of the hyperspectral scenes for the entire spectral range. We present extensive evaluations on diverse real-world hyperspectral scenes from the Hyper-NeRF dataset to show the effectiveness of our DD-HGS. The results demonstrate that DD-HGS achieves the new state-of-the-art performance compared to all the previously published methods. Project page: https://dragonpg2000.github.io/DDHGS-website/

يقدم المقال Diffusion-Denoised Hyperspectral Gaussian Splatting (DD-HGS)، وهي طريقة جديدة تعزز 3D Gaussian Splatting (3DGS) من خلال دمج التوافقيات الكروية الحساسة لطول الموجة ومزيل الضوضاء القائم على الانتشار. تهدف هذه التطورات إلى تحسين وقت التدريب وسرعة العرض في التصوير الطيفي (HSI) للتطبيقات الزراعية، مما يسمح بتحليل دقيق لتكوين العناصر الغذائية للنباتات.

El artículo presenta el Diffusion-Denoised Hyperspectral Gaussian Splatting (DD-HGS), un método novedoso que mejora el 3D Gaussian Splatting (3DGS) al integrar armónicos esféricos sensibles a la longitud de onda y un desruido basado en difusión. Este avance tiene como objetivo mejorar el tiempo de entrenamiento y la velocidad de renderizado de la imagen hiperespectral (HSI) para aplicaciones agrícolas, permitiendo un análisis preciso de la composición nutricional de las plantas.

L'article présente le Diffusion-Denoised Hyperspectral Gaussian Splatting (DD-HGS), une méthode novatrice qui améliore le 3D Gaussian Splatting (3DGS) en intégrant des harmoniques sphériques sensibles à la longueur d'onde et un débruiteur basé sur la diffusion. Cette avancée vise à améliorer le temps d'entraînement et la vitesse de rendu de l'imagerie hyperspectrale (HSI) pour les applications agricoles, permettant une analyse précise de la composition nutritionnelle des plantes.

The paper introduces Diffusion-Denoised Hyperspectral Gaussian Splatting (DD-HGS), a novel method that enhances 3D Gaussian Splatting (3DGS) by integrating wavelength-aware spherical harmonics and a diffusion-based denoiser. This advancement aims to improve the training time and rendering speed of hyperspectral imaging (HSI) for agricultural applications, enabling precise nutrient composition analysis of plants.

Diffusion-Denoised Hyperspectral Gaussian Splatting

Was this article worth reading? Share it

Deptho.ai

Zemith-3bda3b

Attentive AI