Photon Mapping Research Articles

Rasterisation-based progressive photon mapping

Ray tracing on the GPU has been synergistically operating alongside rasterisation in interactive rendering engines for some time now, in order to accurately capture certain illumination effects. In the same spirit, in this paper, we propose an implementation of progressive photon mapping entirely on the rasterisation pipeline, which is agnostic to the specific GPU architecture, in order to synthesise images at interactive rates. While any GPU ray tracing architecture can be used for photon mapping, performing ray traversal in image space minimises acceleration data structure construction time and supports arbitrarily complex and fully dynamic geometry. Furthermore, this strategy maximises data structure reuse by encompassing rasterisation, ray tracing and photon gathering tasks in a single data structure. Both eye and light paths of arbitrary depth are traced on multi-view deep G-buffers, and photon flux is gathered by a properly adapted multi-view photon splatting. In contrast to previous methods exploiting rasterisation to some extent, due to our novel indirect photon splatting approach, any event combination present in photon mapping is captured. We evaluate our method using typical test scenes and scenarios for photon mapping methods and show how our approach outperforms typical GPU-based progressive photon mapping.

Deep Kernel Density Estimation for Photon Mapping

AbstractRecently, deep learning‐based denoising approaches have led to dramatic improvements in low sample‐count Monte Carlo rendering. These approaches are aimed at path tracing, which is not ideal for simulating challenging light transport effects like caustics, where photon mapping is the method of choice. However, photon mapping requires very large numbers of traced photons to achieve high‐quality reconstructions. In this paper, we develop the first deep learning‐based method for particle‐based rendering, and specifically focus on photon density estimation, the core of all particle‐based methods. We train a novel deep neural network to predict a kernel function to aggregate photon contributions at shading points. Our network encodes individual photons into per‐photon features, aggregates them in the neighborhood of a shading point to construct a photon local context vector, and infers a kernel function from the per‐photon and photon local context features. This network is easy to incorporate in many previous photon mapping methods (by simply swapping the kernel density estimator) and can produce high‐quality reconstructions of complex global illumination effects like caustics with an order of magnitude fewer photons compared to previous photon mapping methods. Our approach largely reduces the required number of photons, significantly advancing the computational efficiency in photon mapping.

A fiber scanning tunneling microscope for optical analysis at the nanoscale.

A hybrid scanning tunneling/optical near-field microscope is presented, in which an optical fiber tip coated with 100 nm thick Ag/Cr films scans the surface. The tip metallization enables operating the instrument via a current-based distance control and guarantees sub-nanometer spatial resolution in the topographic channel. The fiber tip simultaneously serves as nanoscale light source, given the optical transparency of the metal coating. The emission response of the tip-sample junction is collected with two parabolic mirrors and probed with a far-field detector. To test the capabilities of the new setup, the evolution of the optical signal is monitored when the tip approaches a gold surface. The intensity rise and frequency shift of the emission provide evidence for the development of coupled plasmon modes in the tip-sample cavity. Photon mapping is employed to probe the optical inhomogeneity of Ru(0001) and TiO2(110) surfaces covered with silver deposits. While the 2D Ag flakes on Ru give rise to a near-field enhancement, the 3D particles on titania locally damp the gap plasmons and lower the emitted intensity. The lateral resolution in the optical channel has been estimated to be ∼1 nm, and optical and topographic signals are well correlated. Our fiber microscope thus appears to be suitable for probing optical surface properties at the nanoscale.

Photon mapping of geometrically complex glass structures: Methods and experimental evaluation

Advances in digital fabrication and additive manufacturing have enabled the creation of geometrically complex glass structures and building components, opening up new design opportunities across scales. Quantifying and evaluating their optical performance, however, remains a technical challenge. In order to accurately predict light behavior, common approaches of daylight modeling utilizing light-backward raytracers are insufficient. This paper evaluates the use of the photon mapping approach within the Radiance render engine to simulate artificial and natural lighting conditions. 3D printed optically transparent glass components are used to benchmark the simulations. We present a workflow to gather geometric data of the glass objects and a series of validation experiments. Pairs of physical studies and digital simulations are compared to assess optical performance in the context of both indoor and outdoor lighting conditions. These experiments demonstrate that the photon mapping approach can reliably measure and predict caustic light patterns and indoor light levels with some limitations, specifically of glare and scattered light from the glass objects themselves.

Denoising Stochastic Progressive Photon Mapping Renderings Using a Multi-Residual Network

Stochastic progressive photon mapping (SPPM) is one of the important global illumination methods in computer graphics. It can simulate caustics and specular-diffuse-specular lighting effects efficiently. However, as a biased method, it always suffers from both bias and variance with limited iterations, and the bias and the variance bring multi-scale noises into SPPM renderings. Recent learning-based methods have shown great advantages on denoising unbiased Monte Carlo (MC) methods, but have not been leveraged for biased ones. In this paper, we present the first learning-based method specially designed for denoising-biased SPPM renderings. Firstly, to avoid conflicting denoising constraints, the radiance of final images is decomposed into two components: caustic and global. These two components are then denoised separately via a two-network framework. In each network, we employ a novel multi-residual block with two sizes of filters, which significantly improves the model’s capabilities, and makes it more suitable for multi-scale noises on both low-frequency and high-frequency areas. We also present a series of photon-related auxiliary features, to better handle noises while preserving illumination details, especially caustics. Compared with other state-of-the-art learning-based denoising methods that we apply to this problem, our method shows a higher denoising quality, which could efficiently denoise multi-scale noises while keeping sharp illuminations.

Multiple Photon Sampling Technique Based on Stochastic Progressive Photon Mapping

This paper introduces a multiple photon sampling technique based on stochastic progressive photon mapping. We use the image space concept to divide the scene into continuous sub-blocks and then we calculate our proposed distance function and photon number function in each of the sub-blocks. The distance function is used to calculate the distance error of the hit point and to determine whether each sub-block is located at a boundary between different objects. The photon number function is used to calculate the photon number error and to determine whether the photon distribution in each sub-block is uniform. Based on the values of the distance error and the photon number error, the multiple photon sampling technique is used to acquire multiple samples of the hit point in each sub-block. Instead of using a single radius for the radiance estimate, we use three different radii and compute the final radiance estimate as a weighted average of the three values. When compared with the existing stochastic progressive photon mapping method, our method provides a better solution to the photon distribution problem and can also reduce bias and noise, especially in the scene with drastic changes in light and dark.

Field localization of hexagonal and short-range ordered plasmonic nanoholes investigated by cathodoluminescence

Plasmonic nanoholes have attracted significant attention among nanoplasmonic devices, especially as biosensing platforms, where nanohole arrays can efficiently enhance and confine the electromagnetic field through surface plasmon polaritons, providing a sensitive detection. In nanohole arrays, the optical resonances are typically determined by the inter-hole distance or periodicity with respect to the surface plasmon wavelength. However, for short-range ordered (SRO) arrays, the inter-hole distance varies locally, so the plasmon resonance changes. In this study, we investigate the local resonance of SRO nanoholes using a cathodoluminescence technique and compare it with hexagonally ordered nanoholes. The cathodoluminescence photon maps and resonance peak analysis reveal that the electric fields are confined at the edges of holes and that their resonances are determined by inter-hole distances as well as by their distributions. This demonstrates the Anderson localization of the electromagnetic waves showing locally enhanced electromagnetic local density of states in SRO nanoholes.

Эффективное вычисление оптимальных весов множественной выборки по значимости в двунаправленной трассировке лучей с фотонными картами

Эффективное вычисление оптимальных весов множественной выборки по значимости в двунаправленной трассировке лучей с фотонными картами

Progressive Photon Elimination With a Status Tree

We describe a modified progressive photon mapping (PPM) method applied with sample elimination, referred to as progressive photon elimination, to pursue accurate results and accelerated iterations. Meanwhile, an elimination status tree is proposed for progressive photon elimination, which retains the information used in the elimination and can be updated along with the multi-pass process to solve the challenge of the accuracy of local parameters. By using the status tree, our method can obtain a uniform photon distribution at each iteration by eliminating a certain number of photons. The tree is also improved to adapt to the complicated photon distributions. This strategy also facilitates the parallel elimination with a priori elimination ratio in the status tree, making the algorithm further accelerated. In parallel block processing, the equilibrium and edge problems are resolved. The experimental results show that our method requires about half of the number of iterations to achieve the same visual effect compared with progressive photon mapping.

Реалистичный рендеринг на основе прямых и обратных фотонных карт

Реалистичный рендеринг на основе прямых и обратных фотонных карт

Charge Carrier Injection Electroluminescence with CO-Functionalized Tips on Single Molecular Emitters.

We investigate electroluminescence of single molecular emitters on NaCl on Ag(111) and Au(111) with submolecular resolution in a low-temperature scanning probe microscope with tunneling current, atomic force, and light detection capabilities. The role of the tip state is studied in the photon maps of a prototypical emitter, zinc phthalocyanine (ZnPc), using metal and CO-metal tips. CO-functionalization is found to have an impact on the resolution and contrast of the photon maps due to the localized overlap of the p-orbitals on the tip with the molecular orbitals of the emitter. The possibility of using the same CO-functionalized tip for tip-enhanced photon detection and high resolution atomic force is demonstrated. We study the electroluminescence of ZnPc, induced by charge carrier injection at sufficiently high bias voltages. We propose that the distinct level alignment of the ZnPc frontier orbitals with the Au(111) and Ag(111) Fermi levels governs the primary excitation mechanisms as the injection of electrons and holes from the tip into the molecule, respectively. These findings put forward the importance of the tip status in the photon maps and contribute to a better understanding of the photophysics of organic molecules on surfaces.

Photon mapping to accelerate daylight simulation with high-resolution, data-driven fenestration models

Data-driven modelling provides a general means to represent optically complex fenestration in daylight simulation by its Bidirectional Scattering Distribution Function (BSDF). Radiance employs the tensor tree as a compact data structure to store the BSDF at high directional resolution. The application of such models under sunny sky conditions is, however, computationally demanding, since the density of stochastic backward samples must match the BSDF resolution. The bidirectional Photon Map is proposed to rapidly forward-sample the BSDF, starting from the known sun direction. Its exemplary application shows a potential speed-up of ⩾ 98% when compared to backward ray-tracing.

Photon-mapping in Climate-Based Daylight Modelling with High-resolution BSDFs

Visual comfort assessments employing luminance-based metrics rely on efficient CBDM techniques for image synthesis. Data-driven BSDF models allow to isolate internal light paths in optically CFS from CBDM. Bidirectional photon mapping is proposed for the efficient sampling of such models in the calculation of the direct solar component in CBDM. The method allows accurate image synthesis for visual comfort assessments with only two calculation steps, achieving comparable accuracy as the established but complex 5PM. The validity of the approach is confirmed by comparison with backward ray-tracing. Its exemplary application to compare two CFS in terms of glare control demonstrates the importance to achieve reconcilability of conflicting targets such as view and glare control in daylighting.

Illumination estimation for augmented reality based on a global illumination model

With the rapid development of 3D technology, the illumination consistency plays an important role in realistic rendering of virtual objects which are superimposed into the real scene. In this paper we proposed a new approach to estimate illumination, given the images of a scene. Our algorithm is designed from principles in optics, which is based on a global illumination model. First, we get the images of a scene and perform scene reconstruction using images. Then, a photon emission hemispherical model is built for emitting photons into the scene. The photons are traced, meanwhile the result is stored in multiple photon maps. Finally, the estimated illumination is obtained by estimating the photon radiance value in the photon emission hemispherical model. Experiments on our newly collected real scene databases and virtual scene databases show the accuracy of our method subjectively and objectively. By comparing with related work, it is found that more accurate results can be obtained using the global illumination inversely.

Photon mapping in image-based visual comfort assessments with BSDF models of high resolution

Data-driven models replicate the irregular Bidirectional Scattering Distribution Functions (BSDFs) of optically Complex Fenestration Systems in daylight simulation. RADIANCE employs the tensor tree to store the BSDF at high directional resolution. Its application in backward ray-tracing is however challenging, since the density of stochastic samples must match the model resolution. BSDF proxy and peak extraction address this problem, but are limited to cases when either the fenestration geometry, or the shape and direction of the transmission peak are known. Photon Mapping is proposed to efficiently sample arbitrary BSDFs from the known sun direction. The existing implementation in RADIANCE is extended to account for light sources and their reflections in the field of view, that are of particular importance for visual comfort assessments. The method achieves a high degree of accordance with ray-tracing, and reduces simulation times by ≈95% with data-driven models of high resolution.Abbreviations: BRT: Backward Ray-Tracing; BSDF: Bidirectional Scattering Distribution Function; CBDM: Climate-Based Daylight Modelling; CFS: Complex Fenestration System; DGI: Daylight Glare Index; DGP: Daylight Glare Probability; FPM: Five Phase Method; GPGPU: General Purpose Graphics Processing Unit; LCP: Laser Cut Panel; OoC: Out-of-Core; PM: Photon Mapping; RMSE: Root Mean Squared Error

Analysis of nuclear response in CFETR toroidal field coils with density reduction VR technique

The China Fusion Engineering Test Reactor (CFETR) has been designed as steady-state operated tokamak device. The CFETR new engineering design has been initialized corresponding to the latest core parameters (major/minor radius is R = 7.2 m/a = 2.2 m). The nuclear analysis of CFETR was carried out to evaluate the neutron shielding capabilities and nuclear responses to toroidal field coils (TFC). The 3D neutronics model of CFETR was built with the support of McCad tool to convert the CAD model automatically. An extended variance reduction (VR) technique of combined MAGIC GVR with density reduction method has been involved to generate suitable weight windows to enhance the efficiency of particle transport. Highly detailed neutron and photon map has been obtained which meet the requirements for Monte Carlo nuclear analyses. The global neutron flux distribution of 3D CFETR with Pf = 200 MW to 1.5 GW was obtained to evaluate its shielding performance. The nuclear responses to TFC such as the nuclear heating and fast neutron fluence have been evaluated. Initial results show the shielding of current divertor design is not sufficient and the optimizing work has been done. The final results satisfy the referred ITER design limit and it can provide the support for the future more detailed design of divertor and other shielding systems.

A task and data balanced distributed photon mapping method

Photon mapping is an important global illumination method of simulating some visual effects, such as caustic. However, photon mapping is difficult to be parallelized efficiently, due to unbalanced task distribution and memory management. In this paper, we present a combined data-parallel and task-parallel approach to improve performance across multiple rendering clients in a distributed system. The shading points (receivers) from pixels are divided into groups. We associate the photons with each receiver group by computing a tree cut of the photon hierarchy, thus both the data and the task are tightly coupled. Furthermore, we organize both the photon hierarchy and the receivers with Morton code for efficient tree construction. In the end, our method balances both the illumination task of receivers and related photons, resulting in better load balancing and higher performance without any quality degradation.

Precomputed Multiple Scattering for Rapid Light Simulation in Participating Media.

Rendering translucent materials is costly: light transport algorithms need to simulate a large number of scattering events inside the material before reaching convergence. The cost is especially high for materials with a large albedo or a small mean-free-path, where higher-order scattering effects dominate. We present a new method for fast computation of global illumination with participating media. Our method uses precomputed multiple scattering effects, stored in two compact tables. These precomputed multiple scattering tables are easy to integrate with any illumination simulation algorithm. We give examples for virtual ray lights (VRL), photon mapping with beams and paths (UPBP), Metropolis Light Transport with Manifold Exploration (MEMLT). The original algorithms are in charge of low-order scattering, combined with multiple scattering computed using our table. Our results show significant improvements in convergence speed and memory costs, with negligible impact on accuracy.

Photonic bandgap transmission map of graphene in a defective optical structure

Interaction of graphene with a defect layer of a 1D defective photonic structure correspondingly could open a Dirac gap in its nonlinear energy dispersion. Also, an excitonic gap could be created upon applied strong magnetic fields at low temperatures. Now, the exact solution for the transmission of the allowed defective states under the influence of a homogeneous magnetic field requires providing a new scheme for the transfer matrix method to cover the circularly polarized propagation of the light in the terahertz frequency regime. In this paper, through an easy-to-follow framework for obtaining the distinct expression of the related transfer matrix method, it is observed that the hybrid defective states in the suggested optical structure undergo an unconventional transmission map as a function of the Dirac gap. Moreover, it is shown that mapping the transmission spectrum in the frequency space reveals peculiar results for narrow-band transport of the defective states. Finally, it is demonstrated that the transmission spectrum shows plateaus as a function of the applied magnetic field analogous to the situations observed for the conductance of a 2D electron gas system in the quantum Hall effect regime. This almost simple suggested design which can be fully controlled marks the first optical counterpart of the quantum Hall effect in optics.

BRDF Analysis with Directional Statistics and its Applications.

Data-driven BRDF models using real material measurements have become increasingly prevalent due to the development of novel gonioreflectometers, but efficient use of these models in many graphical applications remains challenging due to the few functionalities the raw data could provide. To ameliorate this issue, we propose to analyze BRDFs using directional statistics for better handling and exploring measured materials, especially isotropic materials, with efficient computation and compact storage. We conduct a thorough statistical analysis on both analytical BRDF models and measured materials from the MERL database. We show that different aspects of visual appearance can be characterized by different spherical moments, from which several descriptive measures can be derived to further facilitate their usage. We demonstrate how these measures are best leveraged in some graphical applications including gamut mapping using a new BRDF similarity measure, BRDF or SVBRDF reconstruction based on material clustering, and importance sampling for measured materials based on fast extracted GGX distributions. We finally show the potential of our approach in the categorization of surface reflectance types which is common for traditional photon mapping.