Abstract
Getting clear high detailed images with high contrast is an important task in many spheres of science and engineering. However, it's not always possible because of imperfection of devices or environment conditions. This all leaded to development of different methods of image enhancement. In this article a developed two-phase full-color image enhancement algorithm is described. During the first phase the picture is denoised. Wavelet transformation has been chosen to perform it, because it allows easily remove high-frequency parts. Also, noise components, especially big random surges of signal, could be presented like set of local features of signals. Noise can be reduced by thresholding this features. During the second phase brightness and contrast are automatically tuned up using evolutionary algorithm. Evolutionary algorithms, which are effective methods of multidimensional optimization, allow quick selection of optimal values of transformation parameters, using objective optimization criterion.
Original language | English |
---|---|
Title of host publication | Proceedings of SPIE - The International Society for Optical Engineering |
Volume | 6522 |
DOIs | |
Publication status | Published - 2006 |
Event | Thirteenth Joint International Symposium on Atmospheric and Ocean Optics/ Atmospheric Physics - Tomsk, Russian Federation Duration: 2 Jul 2006 → 7 Jul 2006 |
Other
Other | Thirteenth Joint International Symposium on Atmospheric and Ocean Optics/ Atmospheric Physics |
---|---|
Country | Russian Federation |
City | Tomsk |
Period | 2.7.06 → 7.7.06 |
Keywords
- Evolutionary algorithm
- Image enhancement
- Wavelet transformation
ASJC Scopus subject areas
- Electrical and Electronic Engineering
- Condensed Matter Physics