Optimizing Color Preservation in Artistic Style Transfer with VGG-19 and Luminance Transfer Approach

Artistic Style Transfer(AST) a fascinating aspect of Neural Style Transfer(NST) has redefined the creative manipulation of digital images by imbuing photographs with the aesthetic characteristics of famous paintings. However a persistent challenge in NST-color distortion—often compromises the color fidelity of the content image during the style transfer process. This paper introduces an innovative approach employing the VGG-19 neural network to extract style features while preserving the color harmony of the content image through luminance transfer. This method ensures that the original color palette remains unaltered achieving an authentic blend of style and content characteristics. In our experimentation, we applied the innovative artistic style transfer technique to a variety of content images, incorporating the aesthetic elements from renowned paintings. Notably, our method successfully preserved the color fidelity of the content images throughout the style transfer process. This approach contributes significantly to the field by offering a improved technique for NST that harmonizes artistic style with natural coloration serving a valuable asset for artists and technologists.