An intuitive contrast enhancement of an image data employing the self-organizing relationship (SOR)
K. Horio, Takuma Haraguchi, T. Yamakawa · 2003
The user intuition of an image is very important factor to evaluate the image. In this paper, the new image enhancement method, which provides the enhanced image satisfying the intuitive evaluation of the user, is proposed. The self-organizing relationship (SOR) network proposed by the authors is employed in order to obtain the relationship, which reflects the user intuition, between intensity histogram of an original image and intensity mapping curve. The SOR network can construct the relationship, which corresponds to the arbitrary evaluation, between input vector and output vector by the learning. Employing the user intuition as the evaluation, the relationship corresponding to the user intuition is obtained. The intensity histogram of original images and the intensity mapping curve are employed as the input vector and the output vector of the SOR network, the relationship between them is constructed by SOR network. The intensity histogram of an image is applied to the SOR network after the learning, the intensity mapping curve which reflects the user intuition is generated.