Sky Based Light Metering for High Dynamic Range Images

c. Tonemapped after scaling with our estimate d. Tonemapped after scaling with ground truth e. CIE ΔΕ94 between c and d Figure 1: Our algorithm employs sky models to estimate a scaling that can convert images to absolute luminance values. (a) We detect and segment sky pixels, which are (b) analysed to determine properties of the sky dome and camera parameters (the boxes show the visible area of the sky dome in the image). (c) By scaling the image with our estimation, we obtain absolute radiometric values, allowing images to be processed by appearance models such as Reinhard et al. [RPK∗12]. (d) The same processing was applied to the image after scaling it with the ground truth scale factor. (e) The CIE ∆E94 differences between (c) and (d) indicate good correspondence between our estimate and the ground truth. Image calibration requires both linearization of pixel values and scaling so that values in the image correspond to real-world luminances. In this paper we focus on the latter and rather than rely on camera characterization, we calibrate images by analysing their content and metadata, obviating the need for expensive measuring devices or modeling of lens and camera combinations. Our analysis correlates sky pixel values to luminances that would be expected based on geographical metadata. Combined with high dynamic range (HDR) imaging, which gives us linear pixel data, our algorithm allows us to find absolute luminance values for each pixel—effectively turning digital cameras into absolute light meters. To validate our algorithm we have collected and annotated a calibrated set of HDR images and compared our estimation with several other approaches, showing that our approach is able to more accurately recover absolute luminance. We discuss various applications and demonstrate the utility of our method in the context of calibrated color appearance reproduction and lighting design