A New Automatic Watercolour Painting Algorithm Based on Dual Stream Image Segmentation Model with Colour Space Estimation

Image processing plays a crucial role in automatic watercolor painting by manipulating the digital image to achieve desired effect. segmentation algorithms is essential for region-based processing, color mixing and blending, capturing brushwork texture, providing artistic control over final result. It allows more realistic expressive watercolor-like paintings different regions individually applying appropriate effects each segment. Hence, this paper proposed an effective Dual Stream Exception Maximization (DSEM) segmentation. DSEM combines both texture information segment into meaningful regions. This approach begins converting from RGB space perceptually-based space, such as CIELAB, account variations lighting conditions human perception of color. With conversion, extracts relevant features image. Color are computed based on values channels chosen nuances distribution within Simultaneously, derived computing statistical measures local variance or co-occurrence matrices, textural characteristics Finally, model applied deep learning classification painting. Simulation analysis performed compared with conventional techniques CNN RNN. The comparative states that exhibits superior performance terms estimation, region merging. ~12% higher than techniques.