Create a vertical split-style image. The entire canvas is strictly vertically arranged, with the top and bottom halves precisely 1:1 in height, forming a clear and complete vertical visual contrast relationship.
Top half:
Directly present the original image uploaded by the user, treating the original image as the sole primary reference source. Focus on preserving the core subjects such as people, pets, or animals from the original image, prioritizing the extraction of the subject's facial features, hairstyle or fur characteristics, facial features, expressions, postures, clothing, accessories, body contours, and overall temperament.
If buildings, indoor environments, street scenes, furniture, landscapes, etc., appear in the original image, these are not the primary objects of expression and only exist as weakened auxiliary backgrounds.
The top half as a whole maintains a sense of real photography, natural lighting, and original shooting atmosphere, making it obvious at a glance that this is the original reference image itself; do not make the top half into an illustration or a stylized redraw.
Bottom half:
Reconstruct the same person/same pet/same animal subject from the top half into a black and white ink line drawing.
Highlight the most recognizable features of the subject, not mechanically copying photo details, but preserving the most identifiable face shape, silhouette, posture, expression, action relationship, and identity temperament.
The overall style adopts the 2D animation aesthetics of the 1970s, emphasizing bold and rhythmic high-contrast outlines, sharp and clear line art, minimalist but hierarchically varied line width control, and the graphic language of old animation clean-up drawings.
The image uses a high-key composition and adds a pure white backing around the subject or in local areas to make the subject emerge more clearly from the background.
Only black and white are allowed; do not add shadows, colors, grayscale, or midtones; do not do sketch-like light and shadow modeling, and do not add color accents.
The bottom half does not pursue complex environmental depiction but instead forms a black and white visual effect with a sense of era, graphics, and animation through the weight, density, and generalization of outlines.
If the original subject is a person, highlight the expression, hairstyle, movement, and clothing silhouette; if it is a pet or animal, highlight the ears, fur edges, eyes, posture, and rhythm of movement.
The overall effect should make people clearly feel: the top is the real original image, and the bottom is the same subject translated into a 1970s black and white animation ink line image.
Overall requirements:
There must be a clear correspondence between the top and bottom parts, ensuring the bottom part is instantly recognizable as a stylized translation of the subject in the top part.
The image should be clean, restrained, and designed, without generating redundant text, logos, or watermarks; no irrelevant subjects should appear, and buildings or background elements should not seize the visual center.
创作一张竖版上下双拼风格图像,整张画布严格纵向排版,上半部分与下半部分高度精准 1:1,形成清晰而完整的上下视觉对照关系。
上半部分:
直接呈现用户上传的原图,将原图视为唯一主体参考来源。重点保留原图中的人物、宠物、动物等核心主体,优先提取主体的脸部特征、发型或毛发特征、五官、神态、动作、服装、配饰、体型轮廓与整体气质。
如果原图中同时出现建筑、室内环境、街景、家具、风景等内容,这些都不作为主要表现对象,只作为弱化的辅助背景存在。
上半部分整体保持真实照片感、自然光感和原始拍摄氛围,让人一眼看出这是原始参考图本身,不要将上半部分也做成插画或风格化重绘。
下半部分:
将上半部分中的同一人物 / 同一宠物 / 同一动物主体重构为一张黑白墨线画。
突出主体最具辨识度的特征,不机械照搬照片细节,而是保留最有识别性的脸型、轮廓、姿态、表情、动作关系和身份气质。
整体采用二十世纪七十年代二维动画美学,强调粗犷而富有节奏感的高对比轮廓、锐利清晰的线稿、极简但有层级变化的线宽控制,以及老动画清稿式的图形语言。
画面采用高调构图,并在主体周围或局部区域加入纯白衬底,使主体从背景中更清晰地浮现出来。
只允许使用黑白两色,不得添加阴影、颜色、灰度或中间调,不要做素描式明暗塑造,也不要加入彩色点缀。
下半部分不追求复杂环境刻画,而是通过线条的轻重、疏密和轮廓的概括,形成具有年代感、图形感和动画感的黑白视觉效果。
如果原图主体是人物,则突出神态、发型、动作和服装轮廓;如果是宠物或动物,则突出耳朵、毛发边缘、眼神、体态和动作节奏。
整体效果要让人明显感受到:上方是真实原图,下方是同一主体被转译成七十年代黑白动画墨线形象。
整体要求:
上下两部分之间必须存在明确呼应关系,保证下半部分一眼就能识别出是上半部分主体的风格化转译。
画面干净、克制、有设计感,不要生成多余文字、Logo、水印,不要出现无关主体,不要让建筑或背景元素抢夺视觉中心。
Model: ChatGPTSourceUpdated: 9/18/2026, 6:05:23 AM