# HDR

Luminance is a photometric measure of the luminous intensity per unit area of light traveling in a given direction. measured in candela per square meter (cd/m^2)

  • Human static contrast ration: :1 -> about 65 f-stops
  • Human dynamic contrast ration: :1 -> about 20 f-stops

Challenge with expressing brightness effectively in images

  • Need about 5-10 million values to store all brightness around
  • 8-bit images provide only 256 values
  • so you end up capturing values only in one side of the spectrum (focusing on either the dark side of the spectrum or bright side of the spectrum)

# Image Acquisition Pipeline

Image Acquisition Pipeline

# Camera Calibration

  • Geometric: How pixel coordinates relate to directions in the world

  • Radiometric/Photometric: How pixel values relate to radiance amounts in the world

  • Panorama: How pixel coordinates relate to directions in other images

  • HDR: How pixel values relate to radiance amounts in other image

# Radiometric Calibration

  • A color chart with known reflectances
  • Multiple camera exposures to fill up the curve
  • Method assumes constant lighting on all patches and works best when source is far away (example sunlight)
  • Unique inverse exists because g is monotonic and smooth for all cameras

# Response Curves

hdr-response-curve

# Computing Response Curve

computing-response-curve

# Radiance map formats

Require an additional dimension to record expontent

# Tone Mapping

  • Map one set of colors to another
  • Displaying on a medium that has limited dynamic range
    • printers, monitors and projectors all have a limited dynamic range
    • Inadequate to reproduce the full range of light intensities present in natural scenes
  • Addresses the problem of
    • Strong contrast reduction from the scene radiance to the displayable range
    • preserves the image details and color appearance
  • many well-known algorithms exist for this
  • See Banterle, et al. (2011), Reinhard et al. (2002) and Durang and Dursey (2002)