OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image
摄影照片转建筑草图设计海报
Architectural Sketch Design Poster from Photo
来自 OpenNana 的图像提示词，已按本地图像分类和站点 gpt-image-2.5 归类整理。
提示词 / Prompt
中文提示词
请将我上传的每一张照片分别制作成一张独立的高级设计海报，不多图拼接，每张照片单独输出。整体采用 3:4 竖版构图，上下两个区域高度严格 1:1，各占画面 50%。 上半部分保留原始照片，保持主体身份、结构、姿态、真实质感、自然光影和原有色彩氛围，仅进行轻微高级调色，使其具有艺术杂志与展览图像质感。为适配画幅，可自然扩展环境背景，但不得拉伸、扭曲或改变主体。 下半部分只提取照片中最具识别性的主体、轮廓、结构、姿态与叙事关系，重构为建筑概念透视草图 / 建筑方案表现手绘。不要完整保留场景，不要逐物复制照片；主动删去大部分背景、配景和无关细节，只留下最能代表原始主题的核心形体、空间关系与视觉记忆点。 下半部分重新做美术判断。哪怕原图构图差、背景乱、主体小，也要通过删减、重组、裁切、尺度变化和留白，把它救成一张成立的作品，真正赋予模型重新导演画面的权力。 画风采用自由徒手透视线稿 + 少量选择性色块。线条松弛、克制、略带重复与停顿感，可保留少量辅助线、延长线、构造线与建筑手绘的几何参考线，形成明确的设计推敲感。色彩只在少数关键部位轻微铺陈，用于强调主体、结构层次、局部阴影或空间重点，其余区域尽量不画。 画面必须以大量有意识的留白为核心。主体应更小、更精炼，其余空间大面积留空。留白不是空背景，而是构图的一部分，要与主体、线稿、参考线、阴影和少量色块共同形成呼吸感、空间感与节奏感。主体可偏心、贴边、局部裁切或缩小尺度；宁可少画，也不要填满。 可加入少量几何阴影色块强化体量、进深与空间层次，但避免厚重渲染、复杂环境和写实效果图感。背景保持极简，只保留极少量必要的场地暗示。 配色从上方照片中提取 2–4 种最具识别度和生命力的颜色，重新整理为更干净、克制、耐看的草图色系。整体可适度提亮、提纯、去灰，以米白、浅灰、暖灰、柔和绿色、淡木色、浅蓝灰等为主，局部保留少量更鲜明但不刺眼的点睛色。避免脏灰、暗沉、过度复古和廉价高饱和感。 文字只作极少量编辑性介入，不限制语种。可根据主体、地点、动作或情绪生成少量字词、短句或注记，安静地落在留白区域，与画面形成克制的图文关系。 整体呈现建筑概念手绘、自由线稿、几何参考线、少量色块、几何阴影、超大量留白与编辑式排版共同构成的高级视觉效果。避免完整场景保留、逐物复制、复杂背景、满幅上色、写实效果图、粗重描边、卡通感、3D感和模板化表现。
English prompt
Please make each photo I upload into a separate high-end design poster, no multi-image stitching, each photo output individually. The overall composition adopts a 3:4 vertical format, with the top and bottom two areas strictly at a 1:1 height ratio, each occupying 50% of the screen. The upper part retains the original photo, maintaining the subject identity, structure, posture, real texture, natural light and shadow, and original color atmosphere, with only slight high-end color grading to give it the quality of an art magazine or exhibition image. To adapt to the frame, the environmental background can be naturally extended, but the subject must not be stretched, distorted, or changed. The lower part only extracts the most recognizable subject, outline, structure, posture, and narrative relationship from the photo, reconstructing it as an architectural concept perspective sketch / architectural design performance hand-drawing. Do not fully retain the scene, do not copy the photo object by object; actively delete most of the background, scenery, and irrelevant details, leaving only the core form, spatial relationship, and visual memory points that best represent the original theme. Re-make artistic judgments for the lower part. Even if the original composition is poor, the background is messy, or the subject is small, it must be saved into a viable work through deletion, reorganization, cropping, scale changes, and white space, truly giving the model the power to re-direct the image. The painting style adopts freehand perspective line drawings + a small amount of selective color blocks. The lines are relaxed, restrained, with a slight sense of repetition and pause, and a small amount of auxiliary lines, extension lines, construction lines, and geometric reference lines for architectural hand-drawing can be retained, forming a clear sense of design deliberation. Colors are only slightly laid out in a few key parts to emphasize the subject, structural layers, local shadows, or spatial focuses, and other areas are kept unpainted as much as possible. The image must be centered on a large amount of conscious white space. The subject should be smaller and more refined, with other spaces left large and empty. White space is not an empty background but part of the composition, forming a sense of breath, space, and rhythm together with the subject, line drawings, reference lines, shadows, and small amounts of color blocks. The subject can be eccentric, edge-aligned, partially cropped, or reduced in scale; it is better to draw less than to fill it up. A small amount of geometric shadow color blocks can be added to strengthen volume, depth, and spatial layers, but avoid heavy rendering, complex environments, and realistic rendering effects. Keep the background minimalist, retaining only a minimal amount of necessary site hints. Extract 2–4 most recognizable and vibrant colors from the photo above, and reorganize them into a cleaner, more restrained, and durable sketch color palette. The overall tone can be moderately brightened, purified, and de-greyed, dominated by off-white, light gray, warm gray, soft green, light wood color, light blue-gray, etc., with a small amount of more distinct but not glaring accent colors retained locally. Avoid dirty gray, dullness, excessive vintage, and cheap high-saturation feelings. Text is used only as a minimal amount of editorial intervention, with no restrictions on language. A small amount of words, short phrases, or notes can be generated based on the subject, location, action, or emotion, quietly falling on the white space areas, forming a restrained graphic-text relationship with the image. Overall, it presents a high-end visual effect composed of architectural concept hand-drawings, freehand line drawings, geometric reference lines, a small amount of color blocks, geometric shadows, an ultra-large amount of white space, and editorial layout. Avoid full scene retention, object-by-object copying, complex backgrounds, full-page coloring, realistic renderings, thick strokes, cartoonishness, 3D feeling, and templated expression.
探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作
相关作品
- GPT-Image-2.5 像素解构主义
以上传的实拍照片作为唯一素材源，制作完整竖向构图作品，禁止单独生成抽象画、禁止直接给原图加滤镜 画面上下分区，上方为原图摄影区域，占画面50%，完整保留原始照片内容:下方将上传照片转换为“像素消隐景观“风格 插画。保留原照片中最具识别性的主体、基本构图、透视关系和主要色彩氛围,将真实场景提炼为由少量方块组成成的极 简像素化景观。不要直接复制照片细节,而是将建筑、街道、树木、人物、动物、交通工具或其他主体概括为简洁的几何 块面,用规则小方块重新组织画面,呈现一种照片被数字化、提纯后逐渐显现又逐渐消散的视觉效果。严格控制像素方块总 量,整体不超过500个方块。其中主体核心区域使用较密集的方块构成,
- 河边
清澈河水中的写实水下第一人称视角：左下近景有一条橘红鱼，镜头贴近河床仰望岸边石台上并坐的三位成年古风女子。三人分别穿月白、藕荷、青绿色交领襦裙，挽起裙摆避免浸湿衣料，深V领口，丰满身材，双腿与脚浸入浅水；水草、卵石、气泡、折射和午后光柱营造自然戏水氛围。
- H3爆裂冲击帧之【复仇之刃】
H3最强特效生成 粉丝福利邀请码：rh-v1631 10 15 33 三合一 SD2.5之下皆蝼蚁 全网最干净方便的空间折叠工作流，去除繁复连线、虚空桥接+空间折叠； B站主页：https://space.bilibili.com/111985620/lists?sid=7895659&spm_id_from=333.788.0.0 youtube主页：
- 极简双色品牌吉祥物头像与字标
9:16 极简扁平品牌视觉：纯色背景中央放置小型白色反白吉祥物头像，内部以底色镂空形成五官、刘海和发丝分界；底部居中为微型白色品牌字标，右侧边缘另有极小 voxcat 字样和一颗青金石蓝圆点。模板保留主体与品牌名变量，示例默认使用抽象吉祥物头像和来源中出现的 voxcat 字样。
- 镜压成字实验展览海报
9:16 实验性展览海报模板：四个不等大的中文标题字从四角向中心压入，中央镜面金属通过反射切分并重组部分笔画，让材质真实参与字体与版面变化。用深蓝黑背景、暖象牙白文字、镜面银色和少量朱红强调色，保持清楚的三层阅读顺序、精细编辑排版与可读留白。
- 服装店内活力双V手势街拍人像
35mm 街拍风格全身肖像：一位 20 岁出头的成年东亚女性在明亮现代服装店内戴黑框眼镜微笑，双手在头侧比出 V 字，右腿弯曲向后踢起、左腿支撑身体。她穿黑色白条纹五角星针织毛衣、深灰瑜伽裤、白袜和米白厚底运动鞋，斜挎小型棕色包；橡木地板、折叠毛衣货架与柔和商业顶灯构成真实生活方式场景。