Unstructured lumigraph rendering

By Chris Buehler, M. Bosse, Leonard McMillan, Steven J. Gortler, Michael F. Cohen. International Conference on Computer Graphics and Interactive Techniques, 2001

1,074 citations recorded by the source corpus.

We describe an image based rendering approach that generalizes many current image based rendering algorithms, including light field rendering and view-dependent texture mapping. In particular, it allows for lumigraph-style rendering from a set of input cameras in arbitrary configurations (i.e., not restricted to a plane or to any specific manifold). In the case of regular and planar input camera positions, our algorithm reduces to a typical lumigraph approach. When presented with fewer cameras and good approximate geometry, our algorithm behaves like view-dependent texture mapping. The algorithm achieves this flexibility because it is designed to meet a set of specific goals that we describe. We demonstrate this flexibility with a variety of examples.

Subfields: Computer Graphics and Computer-Aided Design, Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/0112657f62ccbd036bae4c84cb41557ee2c8c80d

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Graphics and Computer-Aided Design

In Computer Vision and Pattern Recognition

Citations in the corpus

Soft 3D reconstruction for view synthesis (2017, background): “Unstructured lumigraphs [Buehler et al. 2001] specifies the ‘eight desirable properties’ for IBR algorithms, such as continuity, proxy geometry, and epipole consistency.”
MobileNeRF: Exploiting the Polygon Rasterization Pipeline for Efficient Neural Field Rendering on Mobile Architectures (2022, methodology): “While early methods applied view-dependent texturing to a 3D mesh [7, 12, 13], later methods incorporated convolutional neural networks as a post-processing step to improve quality [20, 31, 44].”
A Survey on 3D Gaussian Splatting (2024, neutral): “Most properties of a 3D Gaussian can be optimized directly through back-propagation.”
Free-viewpoint video of human actors (2003, methodology): “This texture is created by blending the projection of all available camera images onto the geometry [Buehler et al. 2001; Raskar and Low 2002].”
Polyhedral Visual Hulls for Real-Time Rendering (2001, methodology): “The weighting that we use is the k-nearest neighbor weighting used in [3] and summarized here.”
View Synthesis by Appearance Flow (2016, background): “[26] presented a unifying framework for these image-based rendering techniques.”
Common Objects in 3D: Large-Scale Learning and Evaluation of Real-life 3D Category Reconstruction (2021, methodology): “Finally, a very recent IBRNet [68] learns to copy existing colors from known views in an IBR fashion [24, 5], whereas our method can hallucinate new colors, which is crucial since we leverage far fewer source views (at most 9).”
D-NeRF: Neural Radiance Fields for Dynamic Scenes (2020, background): “…ﬁeld of neural rendering, which learn scene representations encoding both geometry and appearance [26, 23, 19, 50, 29, 35], have achieved re-sults that largely surpass those of traditional Structure-from-Motion [14, 41, 38], light-ﬁeld photography [18] and image-based rendering approaches [5].”
Enhancing Photorealism Enhancement (2021, background): “real imagery of a scene for rendering novel views [36], [37], [38], [39], [40], [41], [42], [43], [44].”
NeRF in the Dark: High Dynamic Range View Synthesis from Noisy Raw Images (2021, background): “A more feasible capture scenario is to capture more widely spaced inputs and use a “proxy” geometry ( e.g ., a reconstructed tri-angle mesh) to reproject and combine colors from the input images, using either a heuristic [6] or learned [21, 39, 40] blending function.”
Block-NeRF: Scalable Large Scene Neural View Synthesis (2022, methodology): “This geometric “proxy” is then used to reproject pixels from the input images into new camera views, where they are blended by heuristic [6] or learning-based methods [24,52,53].”
Free View Synthesis (2020, background): “[3] describe the Unstructured Lumigraph, which utilizes dense and accurate 3D geometry to map and blend the input images in a novel target view.”
Let There Be Color! Large-Scale Texturing of 3D Reconstructions (2014, result): “Similar to this are the Lumigraph’s [4] view blending weights, which account for the very same effects.”
Local light field fusion (2019, neutral): “…a global mesh from all inputs using the screened Poisson surface reconstruction algorithm [Kazhdan and Hoppe 2013], and use the heuristic Unstructured Lumigraph blending weights [Buehler et al. 2001] to blend input images after reprojecting them into the novel viewpoint using the global mesh.”
Real-time consensus-based scene reconstruction using commodity graphics hardware (2002, methodology): “• The model can be used with traditional compute graphics objects, or as input to other image-based rendering methods, such as [5, 2, 8], to provide more compelling results.”
blue-c: a spatially immersive display and 3D video portal for telepresence (2003, methodology): “This is accomplished by weighted averaging, similar in spirit to Unstructured Lumigraph Rendering [Buehler et al. 2001].”
Depth Anything 3: Recovering the Visual Space from Any Views (2025, neutral): “Feed-Forward Novel View Synthesis Novel view synthesis (NVS) has long been a core problem in computer vision and graphics [8, 34, 49], and interest has increased with the rise of neural rendering [28, 44, 57, 77, 78].”