Improving the quality of high dynamic range images

TLDR

HDR photography is a valuable luminance measurement tool, yet its accuracy depends on camera, lens, and photometric calibration, for which technical guidance is lacking. The authors captured 128 HDR images of 54 grey and 18 colour targets using a Canon 350D with a Sigma 10–20 mm lens across multiple apertures, focal lengths, ambient light levels, and exposure ranges, then applied vignetting correction and physical calibration before comparing the images for luminance acquisition quality. The results demonstrate that the calibrated HDR imaging protocol improves image quality, offering a useful framework for future HDR photography.

Abstract

High dynamic range (HDR) photography is a useful luminance measurement tool but relies on the camera and lens used and photometric calibration for its accuracy, yet these aspects lack technical guidance. This study used a Canon 350D camera fitted with a Sigma 10—20 mm lens to record the luminance data of 54 grey and 18 colour targets plus a light source in a windowless classroom lit by fluorescent lighting. Six apertures (f/4—f/22), three focal lengths (10, 14, 20 mm), two ambient light levels and four ranges of exposure values were used for generating 128 HDR images in Photosphere. After vignetting correction and physical calibration, these HDR images were compared with each other for the quality of luminance acquisition. This study could benefit future HDR imaging by ensuring higher image quality.