A neuromorphic model for achromatic and chromatic surface representation of natural images
Neural Networks - 2004 Special issue Vision and brain
Color Characterization and Balancing by a Nonlinear Line Attractor Network for Image Enhancement
Neural Processing Letters
Detecting and removing specularities in facial images
Computer Vision and Image Understanding
Detecting and removing specularities in facial images
Computer Vision and Image Understanding
| Hi-index | 0.00 |
The last version of Edwin Land''s retinex model for human vision''s lightness and color constancy has been implemented. Previous research has established the mathematical foundations of Land''s retinex but has not examined specific design issues and their effects on the properties of the retinex operation. Here we have sought to define a practical implementation of the retinex without particular concern for its validity as a model for human lightness and color perception. Here we describe issues involved in design the surround function. We find that there is a trade-off between rendition and dynamic range compression that is governed by the surround space constant. Various functional forms for the retinex surround are evaluated and a Gaussian form found to perform better than the inverse square suggested by Land. Preliminary testing led to the design of a Gaussian surround with a space constant of 80 pixels as a reasonable compromise between dynamic range compression and rendition.