Performance of the face verification system depend on many conditions. One of the most problematic is varying illumination condition. In this paper 14 normalization algorithms based on histogram normalization, illumination properties and the human perception theory were compared using 3 verification methods. The results obtained from the experiments showed that the illumination preprocessing methods significantly improves the verification rate and it's a very important step in face verification system.
2
Dostęp do pełnego tekstu na zewnętrznej witrynie WWW
Retinex, a model of human color vision suitable for unsupervised chromatic equalization, due to Land and McCann, is receiving a reneved interest after several years. Different versions have been developed so far, and it has been used for various applications. Most of the implementations skip the classical random paths approach because of the high frequency chromatic noise it introduces. To solve the noise problem, avoiding the increase of path number and witout substituting the random path approach, we present in this paper two new retinex versions: one based on lookppup table transformation, and another based on multilevel image decomposition. These versions strongly decrease the dependency of the computed pixel value on the path randomness, eliminating in this way a great part of the chromatic noise.
JavaScript jest wyłączony w Twojej przeglądarce internetowej. Włącz go, a następnie odśwież stronę, aby móc w pełni z niej korzystać.