Product Code Database
The red-eye effect in photography is the common appearance of red pupils in color photographs of eyes. It occurs when using a photographic flash at low lighting or at night. When a flash passes through the eyes and rebounds at the back of the eye, it causes a red reflex in an image, turning the subject's eyes red. The hue is mostly caused by a high concentration of blood in the choroid. The effect can also be influenced by the near proximity of the flash and camera lens. In children, a different hue red reflex, such as white or yellow, may indicate an illness. In animals, a similar effect could cause their eyes to change colors in photographs.
The effect can be avoided physically by instructing the subject to look away from the lens, increasing the brightness of the photographic location, or moving the flash further away from the lens, or digitally by using the red-eye correction option on or by removing the effect in editing software. Scholars have developed a number of red-eye detection techniques to improve digital red-eye removal.
The effect could also infrequently occur in only one eye of persons with or tumors, or it may be exacerbated by photographing subjects that are drunk. In children, if the red reflex is white, it may indicate retinoblastoma malignancy, whereas yellow color may indicate Coats' disease. If the effect is asymmetrical, it may indicate strabismus. A similar reaction to the red-eye occurs in animals, turning their eyes different colors.
Scholars have proposed a number of red-eye detection methods for digital red-eye correction. The red-eye detection algorithm developed by electronic engineers Seunghwan Yoo and Rae-Hong Park is composed of face detection, region-growing, and red-eye detection. Following the detection of the subject's face and eyes, the algorithm searches for red-eye regions, which are subsequently enlarged using the region-growing technique. They claim that pupil size calculation, pupil painting, and iris detection make up the red-eye correction that may then take place. Computer scientists Xiao-Ping Miao and Terence Sim have proposed their technique for detecting red-eye by fusing subjects' flash and non-flash photos. Scholars Tauseef Ali, Asif Khan, and Intaek Kim put forth an algorithm that, following facial recognition, the image before applying red-eye correction. On the contrary, researchers Richard Youmaran and Andy Adler presented a method for improving the quality of videos with the red-eye phenomenon. They have stated that their method consists of two frames, the first shot under standard illumination with no infrared light and the second one using infrared light.
|
|
