ABSTRACT
In this paper, the direct relationship between the gradient and the contrast are considered, which means a sharp contrast is always associated with a large gradient value and a low contrast corresponds to a small gradient, in order to find a way to deal with images of all degrees of clarity. Thus we introduce the gradient domain method to enhance vein image. Firstly, we calculate the gradient field of the original image and design a gradient distribution model from which we modify the gradient field of the original image to create a new gradient field where the previously small gradients are stretched and the zero gradients and large ones remain unchanged. Finally, a new image is obtained by reconstruction from the new gradient field. Our method can not only enhance the typical picture but can also be effective in improving low contrast images. Furthermore, a high process speed is delivered by our strategy that satisfies the real-time requirement of the vein recognition system or vein imaging system.
