🏷️#3117国漫画韵人物转绘风：竖版上下对比转绘提示词

严格生成一张竖版上下对比转绘图：上半部完整保留输入真实照片的主体身份、人数、动作、视角与场景；下半部将同一主体转译为现代中国漫画与国产二维动画人物设计。使用克制、均匀而流畅的深色细轮廓线，古典修长眉眼，概括但准确的面部结构，清晰平涂色块，低饱和青绿、米白与从原照片提取的少量辅助色；阴影只用一至两级干净色面，不做写实明暗。把原人物衣褶、头发、道具或环境动势延展为飘带般长弧线和书法式速度线，但必须保持原动作可辨。背景高度简化为大面积暖白纸、青绿几何色面、少量云纹或从原场景提取的轮廓，不堆叠景物。仅加入非常轻微的纸面颗粒，使画面像当代中国漫画角色设定稿或国产二维动画关键帧。上下各占约一半，水平边界清晰；禁止水墨画、工笔重彩、敦煌壁画、古画卷轴、复杂祥云堆叠、写实厚涂、3D渲染、日漫大眼脸、欧美漫画、照片滤镜、左右并排、九宫格、改变人物数量与姿态、复制人物、过度仙侠特效、文字角标、水印以及“原图/AI图”标签。人物服装不必固定汉服：现代服装、运动装、民族服饰和旅行穿搭均应依据原照片进行国漫化设计。

Strictly generate a vertical top-bottom comparison transformation image: the upper half fully preserves the subject identity, number of people, action, viewpoint, and scene of the input real photo; the lower half translates the same subject into a modern Chinese comic and domestic 2D animation character design. Use restrained, even, and smooth dark fine outlines, classical elongated eyebrows and eyes, a concise but accurate facial structure, clean flat color blocks, low-saturation teal green, off-white, and a small amount of auxiliary colors extracted from the original photo; shadows should use only one or two clean tonal layers without realistic shading. Extend the folds of clothing, hair, props, or environmental motion into ribbon-like long arcs and calligraphic speed lines, while keeping the original pose clearly recognizable. Simplify the background into large areas of warm white paper, teal geometric color blocks, and a few cloud motifs or outlines extracted from the original scene, without stacking scenery. Add only very subtle paper grain so the image feels like a contemporary Chinese comic character sheet or a key frame from domestic 2D animation. The top and bottom each occupy about half, with a clear horizontal boundary; prohibit ink wash, elaborate gongbi coloring, Dunhuang murals, ancient scroll style, complex layered cloud motifs, realistic thick painting, 3D rendering, big-eyed anime faces, Western comics, photo filters, side-by-side layouts, nine-grid layouts, changing the number or pose of people, duplicated subjects, excessive xianxia effects, text labels, watermarks, and “original image / AI image” tags. Clothing does not have to be fixed as hanfu: modern clothing, sportswear, ethnic clothing, and travel outfits should all be adapted into Chinese comic style according to the original photo.