This quiet elaborate method of image creation is based on a combination of color- and information maps plus a suitable prompt. This method offers artistic control over the elements in the picture by their shape, color and overall apperance. The method is based on a hand drawn segment map, that is used as a figurative guideline – on top of that offset comes a handdrawn color map, that guides the color interpretation of the motif and the overall lighting and appearance.

using segmentation and color map in A1111

As a fundament, we use the img2img render pipeline with our handdrawn color image. Render pixel dimensions should align with the size or ratio of your maps.

hyper realistic contemporary acrylic painting of (many old female persons in white futuristic gardener shaman suits and sitting on the floor observing :1.1), black water like oil glossy flowing , white linen cloth sheets with traditional textile patterns on the grass, concrete distressed dirty buildings in background, (triadic unsaturated color palette, sharp, rough strokes,:1.1) curvy black an white cables tentacles from the sky , fog and dust fluffy
deformed face, naked, beard, sexy, ugly, boat, sun, building, city, cute, cute girl ,(low quality, worst quality:1.2), (monochrome:1.2), photo, (( disfigured )), distorted limbs, disfigured, bad anatomy, bad hand, extra digits, extra fingers, twisted fingers, long body, jpeg artifacts, signature, watermark, username, censor, text, asian, animals, frame, symmetrical, butterfly, asian

The pipline uses the controlNET extension for A1111. This needs to be installed plus the necessary controlNET models. They can be downloaded here:
https://huggingface.co/lllyasviel/ControlNet-v1-1/tree/main

The ADetailer is another extension for A1111 to fix face, hand and limb render flaws in the render process. With a simple prompt, we can determine what the recognized faces should be rendered like.