原生艺术上下双拼构图插画

来源: @zhidawang219555模型: ChatGPT

📸 示例图片

示例 1

提示词

English
Create a vertical illustration with a "top-bottom split composition," where the vertical layout of the canvas is divided into top and bottom areas with a strict 1:1 ratio. It should be a single complete work, not a collage of two independent images.

Top Half:
Use the user-uploaded original image as the primary source, automatically identifying the main characters, cats, dogs, or other common pets in the original image and presenting them as the sole core subject. Focus on extracting identifiable information of the subject itself, including the person's face shape, facial features, hairstyle, hair color, skin tone, age, clothing, accessories, posture, and temperament, or the pet's fur color, patterns, body shape, ear shape, eye characteristics, and expression. The top half should center on "original subject display," maintaining a visual state that is as realistic, natural, and clear as possible. Appropriate cropping and composition optimization can be done according to the needs of the split screen, but do not change the identity of the subject. If the original image contains buildings, indoor spaces, street scenes, or other complex environmental elements, these contents should only be treated as auxiliary background weakening and not as key extraction objects; there is no need to specifically preserve the architectural style. The overall requirement for the top half is that the subject is clear, complete, and not cut off, the image is clean, the visual center is stable, and the main task is to clearly show "who the original subject uploaded by the user is."

Bottom Half:
Perform a stylized translation of the same subject from the top half in the bottom half. It must still be the same person or animal, maintaining core recognizability and not becoming an unrelated new character. Create a complete illustration in the visual language of "Raw Art / Art Brut." Core stylistic features: unrefined academic lines, rough forms, direct symbols, and irregular color blocks. General principle: allow for exaggeration, distortion, and irregular brushstrokes, creating visual impact with strong lines, ink marks, and emotional tension. The composition uses a single clear subject or a clear scene, the visual center is stable, the subject is complete and not cut off, and the background is moderately simplified while retaining a sense of breathability; lines, color blocks, light and shadow, materials, and paper texture must serve the same style and not mix in unrelated painting styles. The image should be natural, have a handmade feel and recognizable stylistic features, avoiding common AI issues like excessive smoothness, meaningless piles of detail, and random decoration. Unless the theme specifically requires it, do not generate text, logos, watermarks, borders, or UI, and do not make multi-image collages.

Overall Supplementary Requirements:
The top and bottom parts must clearly correspond to the same subject, forming a top-bottom contrast relationship of "original subject / stylized subject." The bottom half should reflect a raw, direct, and unrefined sense of crude art, but must not lose the recognizability of the subject. Maintain a handmade, simple, primitive, and impactful visual language overall, and do not make it into a polished commercial illustration. Do not generate additional explanatory text, titles, logos, watermarks, UI, or borders.
中文
创作一张“上下双拼构图”的竖版插画作品,画布纵向排版,上下区域高度比例严格为 1:1,整体为同一张完整作品,不是两张独立图片拼贴。

上半部分:
使用用户上传的原图作为主体来源,自动识别原图中的主要人物、猫、狗或其他常见宠物,并将其作为唯一核心主体进行呈现。重点提取主体本身的可识别信息,包括人物的脸型、五官、发型、发色、肤色、年龄感、服饰、配饰、体态、气质,或者宠物的毛色、花纹、体型、耳朵形状、眼睛特征、神态等。
上半部分应以“原图主体展示”为核心,尽可能保持真实、自然、清晰的视觉状态,可根据双拼画面的需要做适度裁切与构图优化,但不要改变主体身份。
如果原图中带有建筑、室内空间、街景或其他复杂环境元素,这些内容只作为辅助背景弱化处理,不作为重点提取对象;建筑风格不用专门保留。
上半部分整体要求主体明确、完整、不被裁断,画面干净,视觉中心稳定,主要任务是清楚展示“用户上传的原始主体是谁”。

下半部分:
以下半部分对上半部分同一主体做风格化转译,必须仍然是同一个人物或同一只动物,保持核心辨识度,不得变成无关的新角色。
创作一张「Raw Art / Art Brut 原生艺术」视觉语言的完整插画。核心风格特征:未经学院化修饰的原生线条、粗糙形体、直接符号和不规则色块。整体原则:允许夸张、扭曲和不规则笔触,以强烈线条、墨迹和情绪张力制造视觉冲击。构图采用单一明确主体或一个清晰场景,视觉中心稳定,主体完整不被裁断,背景适度简化并保留呼吸感;线条、色块、明暗、材质和纸面纹理必须服务于同一风格,不混入无关画风。画面自然、有手作感和可识别的风格特征,避免AI常见的过度光滑、无意义细节堆砌和随机装饰。除非主题明确要求,否则不生成文字、Logo、水印、边框、UI,不做多图拼接。

整体补充要求:
上下两部分必须明显对应同一主体,形成“原图主体 / 风格转译主体”的上下对照关系。下半部分要体现原生、直接、未经修饰的粗粝艺术感,但不能失去主体辨识度。整体保持手工、质朴、原始、有冲击力的视觉语言,不要做成精致商业插画。不要额外生成说明文字、标题、Logo、水印、UI、边框。

更多推荐

12 个精选提示词