Imagery with great dynamic range

Data-processing principle

The traditional Digital image that everyone knows is coded on 256 values (between 0 and 255) on each green and blue red level, i.e. of 24 Bit S by pixel (3 X 8 bits). The variation of luminous intensity between the most luminous pixel and the weakest pixel, not black, is thus only of 255. However in reality, it is current that dynamics between the most luminous zones and darkest of a scene is larger. Images HDR use more bits/pixel than the traditional images and make it possible to store a largely higher dynamics. The most current technique is to store the images with a floating by color (96 bits per pixel) but it exists also images HDR in 32 bits/pixel, like format RGBE of Radiance or the LogLuv format of SGI.

Applications

One finds the use of the imagery with great dynamic range in the following fields:

Numeric photography

Principle

In 2006, the apparatuses Fuji FinePix S3 Pro (Reflex camera) and Fuji FinePix F700 (compact) are the only ones with being able to produce photographs with great dynamic range thanks to the use of particular Photodiode S.

However, obtaining a photograph HDR can also be done with a traditional apparatus by taking several photographs LDR ( Low Dynamic Range ) then by amalgamating them with a Logiciel. With this intention, it is necessary to carry out several stereotypes of the same scene with different durations (Bracketing) (what makes it possible to obtain details for the dark zones and at the same time for the clear zones). During the various catches of sight, the stability of the apparatus must be maintained to obtain same framing. Then, the photographs with weak dynamic range can be combined thanks to a software (like Photomatix or Qtpfsgui) to form only one photograph with great dynamic range (HDR). There exists also a application Web which makes it possible to combine images (WebHDR) which is developed by Low Energy Architecture Research Unit (LEARN) off the London Metropolitan University.

An image HDR has only little interest if it is not exploited and is not reconverted into a general format (.jpeg for example). This is why it is then necessary to use an artifice, the “Mapping Ton”, to reveal the image correctly. This algorithm causes to create a photograph where the totality of the elements are correctly exposed, without any over-exposure or under-exposure. Returned is often very realistic, in particular for the reflections, the cloudy landscapes and the night environments.

Examples

Random links: