Orasis is a brain-inspired, image processing software, which attempts to simulate the sensation of appearance, formed by our Visual System. The name "Orasis" comes from the Greek word Όρασις, which means Vision.
or not, what we “see” is not just the light which reaches our eyes, but rather,
the result of a very complicated spatial processing that the retina and visual
cortex apply to the image captured by our eyes. Part of this processing is
dedicated to form the sensation of appearance. The sensation of appearance
determines how much dark or light are the objects that we perceive, what is
their exact shade and their level of details.
Cameras capture just light. As a result, photographs may differ significantly from the images that we perceive. Humans can perceive many more details in the shadows or highlights, while, most of the time, captured photographs suffer from considerable loss of visual information in these areas. This discrepancy is much more obvious in the following three cases:
Captured photograph What we perceive
Make the following experiment! Use your camera or iPhone to take a picture of your window across the room, during the day. Turn off any lights. Point your camera in a way, as to capture partially the window and partially the room (see sample photos below) and then take the photo. Observe it carefully, looking at the real scene and back to your captured photograph. The dimmer the ambient room light is, the greater the discrepancy you will see. You will experience something similar to this:
Captured photograph What you will perceive
Orasis attempts to bridge the gap between "what you see" and " the camera’s output". Orasis enhances the shadow or the highlight regions of an image, while keeping intact all the correct ones. The final result is a lot closer to the human perception of the scene, than the original captured image, revealing visual information that otherwise wouldn't be available to the human observer.
Orasis’ algorithm is the result of years of academic research in the field of biologically-inspired image processing, developed by the Electronics Lab of the Democritus University of Thrace, department of Electrical & Computer Engineering in Greece, and has been published in peer reviewed scientific journals. It combines the shunting characteristics of the ganglion cells of the retina, as well as, the filling-in mechanisms of the visual cortex, into an efficient spatially-processing algorithm. This means that Orasis does not globally transform the whole image, but rather locally, according to its contents. When pressing the “Auto” correction button, a neural network, which has been trained using psychophysical data from many observers, selects the most appropriate parameters in order for the corrected photograph to approximate, as close as possible, the image as seen by your eyes. Since however, photography is also a matter of personal taste, controls are provided for any additional manual adjustment.
Orasis will make your photographs look closer to what you see with your eyes, compared to what they originally looked like, when they were captured by the camera. This means that, you should not always expect “WOW” results, since there are cases where captured photographs are a quite accurate representation of what we see. These kinds of photographs are usually taken indoors, in a uniformly well lit room, without any shadows or strong light sources in the picture. In these cases, Orasis does not change the captured images significantly.
However, in cases where shadows, strong light
sources (e.g. the sun), camera flash, or generally, high contrast imaging
conditions are present, Orasis can significantly improve your captured images.
More specifically, it will lighten all the dark areas, such as shadows or dark
backgrounds caused by camera flash, and at the same time it will darken all
the unnaturally bright areas, equalizing the appearance of the image and making
it look “more natural”. Most important though, it will do so without
compromising local contrast (e.g. object details) and without affecting all the
good image regions. These features are clearly depicted in the following image.
Orasis is not an HDR software. It does not take multiple photographs of the same scene and combine them into one final image. Although its results can be similar to some HDR applications, Orasis uses only a single image. As a consequence, its results are bounded by the quality and the range of visual information contained in the captured image. Particularly, image regions which are totally white, due to their very strong brightness during capturing, contain absolutely no visual information, and thus, cannot be corrected by Orasis or any other image processing software.