OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image

圆形底座上的写实等距城市场景

Realistic Isometric City Scene on a Circular Base

将与固定参考身份一致的成年人物置入写实等距城市街景剖面，以圆形底座、真实风化材质和比例协调的道具呈现完整场景。

圆形底座上的写实等距城市场景 — AI image creation
圆形底座上的写实等距城市场景

提示词 / Prompt

中文提示词

使用上传的参考图作为画面中主要人物的面部身份参考。除与身份参考冲突的部分外，保留来源场景、造型、服装、姿势、构图、光线和视觉风格。创作一幅高度写实的3D等距广角画面，呈现紧凑、略显风化的城市街角场景，并完整置于清晰可见的圆形底座上。画面包含一位明确年满25岁、身份绑定至ref-10a-v1的成年人，穿着简洁日常服装，以全身入镜的写实真人形象呈现，保留自然皮肤纹理、毛孔和细微真实瑕疵；不得像玩具、娃娃或微缩人偶。人物自然放松地站在道具旁。人物后方是一段真实质感的风化砖石与水泥店面墙，砖块、水泥或木材应呈自然老化质感，而非涂绘模型质感。人物旁停放一辆无品牌的复古风踏板车；底座上放置一张小咖啡桌，桌上有披萨和一杯葡萄酒。所有道具都要有明确落地关系，人物、建筑和物件之间保持准确比例。圆形底座、人物全身和所有道具必须完整留在画面内，不得裁切。使用等距广角视角、深景深，并确保全景清晰对焦；背景为纯净中性灰色。整体呈现精致、高细节、照片般写实的3D渲染效果，具有可信的全局光照和真实材质反应；不要声称使用了特定引擎、光线追踪流程或输出分辨率。不要玩具、塑料、娃娃、动作人偶、黏土或树脂质感；不要浅景深、特写、元素裁切、比例失真、巨大头部、卡通风格、肢体缺失或未完成表面。不要文字或水印。

English prompt

Use the uploaded reference image as the facial identity reference for the main visible person. Preserve the source scene, styling, clothing, pose, composition, lighting and visual style unless they conflict with the identity reference. Create a highly realistic 3D isometric wide view of a compact weathered urban street-corner vignette on one clearly visible circular base. Include one clearly adult person age 25 or older, with identity bound to ref-10a-v1, wearing simple everyday clothing, shown full-body as a realistic living human with natural skin texture, pores and subtle imperfections—not a toy, doll or miniature. Have the person stand naturally beside the props in a relaxed posture. Behind them, build a realistic section of weathered brick-and-cement storefront wall, with naturally aged brick, cement or wood rather than painted model textures. Place an unbranded vintage-style scooter beside the subject and a small café table with a pizza and a glass of wine on the base. Ground every prop visibly and keep accurate relative scale among the human, architecture and objects. Keep the complete circular base, entire person and every prop inside the frame without cropping. Use an isometric wide-shot perspective, deep depth of field and sharp focus throughout, against a solid neutral-gray background. Aim for a polished, high-detail photorealistic 3D-render aesthetic with convincing global illumination and realistic material response; do not claim a specific engine, ray-tracing pipeline or output resolution. No toy, plastic, doll, action-figure, clay or resin appearance; no shallow depth of field, close-up, cropped elements, distorted proportions, giant head, cartoon style, missing limbs or unfinished surfaces. No text; no watermark.

查看 API 模型

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