Indoor Heat Estimation from a Single Visible-Light Panorama

TL;DR

The paper tackles the challenge of estimating indoor thermal distribution from a single visible-light panorama, addressing the limitations of infrared-only methods and narrow FOV. It proposes a joint pipeline that uses paired indoor–outdoor HDR panoramas to reconstruct the indoor 3D floor layout, derive a 360° outdoor environment map for spatially varying illumination and material estimation, and run a transient heat model to produce per-pixel indoor temperature maps. Its key contributions are the Visible-Light Estimation that links light transport to heat transfer for panoramas and the Transient Heat Estimation that simulates heat diffusion with a per-pixel heat source derived from illuminance, followed by validation against real infrared images. The approach enables photorealistic rendering and integrated light–heat visualization suitable for virtual home staging and indoor design, offering a practical means to assess solar heat gains and lighting without relying solely on infrared data.

Abstract

This paper introduces a novel image-based rendering technique for jointly estimating indoor lighting and thermal conditions from paired indoor-outdoor high dynamic range (HDR) panoramas. Our method uses the indoor panorama to estimate the 3D floor layout, while the corresponding outdoor panorama serves as an environment map to infer spatially-varying illumination and material properties. Assuming indoor surfaces are Lambertian and that all heat originates from outdoor visible light, we model the relationship between light transport and heat transfer, and perform transient heat simulations to generate indoor temperature distributions. The simulated heat maps are validated against real-world thermal images captured with an infrared camera. This approach supports photorealistic and physically informed visualization, enabling integrated light and heat estimation to advance traditional virtual home staging.

Figures (9)

• Figure 1: Overview of this work: An indoor panorama (a) is captured under natural illumination, with a paired outdoor panorama providing real-time, spatially-varying light to the scene. The scene is virtually edited with new indoor layout objects (b). The absolute light level ($cd/m^2$) (c) is estimated for the virtual scene, and we introduced a heat transport equation to compute an indoor heat map (d) displaying per-pixel temperature values ($^{\circ}\mathrm{C}$).
• Figure 2: Our pipeline for indoor light and heat estimation: First, a single camera (Ricoh Theta Z1) captures paired indoor and outdoor panoramas. Second, the indoor panorama is used to estimate the 3D floor layout, while the outdoor panorama provides a 360$^{\circ}$ environment map for 3D scene model. Third, virtual objects are inserted into the scene, and the objects are relit within the scene. Finally, an indoor light map with absolute light level ($cd/m^2$) and a heat map ($^{\circ}\mathrm{C}$) are computed for this scene.
• Figure 3: Heat simulation process: (a) The input is a single panorama; (b) a 3D layout is reconstructed from the input panorama; (c) planar mesh surfaces are uniformly sampled to generate a grid of vertices; (d) the heat equation is applied to compute the per-pixel heat value at each vertex; and (e) the output heat map is projected from 3D coordinates onto an equirectangular representation.
• Figure 4: Transient heat simulation over time: The heat map is initialized and recorded at successive time steps, shown from a top-down view to illustrate the 3D floor layout as heat propagates from 0$s$ to 200$s$.
• Figure 5: Indoor light and heat estimation under new outdoor sun illuminations: The virtual scene is rendered under the captured outdoor panorama (a). The same scene is then rendered with a virtual sun at 16:45 (b), 17:45 (c), and 18:45 (d), respectively.