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

参考主体超写实真人电影海报转换

Hyperrealistic Movie Poster Conversion of Reference Subject

将上传参考主体转换为原创真人电影关键艺术海报，保留身份、轮廓、色彩和标志性道具线索。

参考主体超写实真人电影海报转换 — AI image creation
参考主体超写实真人电影海报转换

提示词 / Prompt

中文提示词

将上传参考图转换为一张超写实4:5竖版真人电影院线海报，保留参考主体的可识别身份、轮廓、面部神态、发型或头部形状、服装结构与色彩、标志性道具、符号、个性和存在感。将卡通或风格化特征重新诠释为可信的高端材质、真实皮肤或表面纹理和协调解剖，同时不能失去辨识度。根据参考图可见线索构建原创电影世界：顶部安排主角巨大特写，中部为处于故事环境中的全身英雄姿态，仅在参考有依据时于两侧和下方加入克制的盟友、对手、生物、载具或世界细节。使用戏剧性轮廓光、背光、体积氛围、反射高光、电影调色、风化、烟雾、余烬、雾气、火花和可信景深。下方三分之一加入清晰的原创电影标题The Last Horizon，下方小字加入Every world leaves a mark；其他演职员字样保持极小、虚构且拼写正确，不得出现真实制片厂或系列名称。采用高端好莱坞关键艺术构图、丰富世界构建、真实材质和8K细节。排除身份丢失、复制真实系列、除提供参考主体外的名人相似度、扭曲手部、多余肢体、变形面孔、不可读标题、标志、水印和廉价角色扮演质感。

English prompt

Transform the uploaded reference image into an ultra-realistic 4:5 vertical live-action theatrical poster, preserving the reference subject's recognizable identity, silhouette, facial attitude, hairstyle or head shape, costume structure and palette, signature props, symbols, personality, and iconic presence. Reinterpret stylized or cartoon traits as believable premium materials, realistic skin or surface texture, and coherent anatomy without losing recognition. Build an original cinematic world from the reference's visible cues, with a giant close-up of the lead at the top, a full-body heroic pose in a story-related environment at center, and restrained supporting allies, rivals, creatures, vehicles, or world details at the sides and lower areas only when supported by the reference. Use dramatic rim light, backlight, volumetric atmosphere, reflective highlights, filmic color grading, weathering, smoke, embers, fog, sparks, and believable depth. Add a clean lower-third title reading The Last Horizon and a smaller line reading Every world leaves a mark; all other credits must be tiny, fictional, and correctly spelled, with no real studio or franchise names. Use premium Hollywood key-art composition, rich world-building, realistic materials, and 8K detail. No identity loss, copied real franchise, celebrity likeness beyond the supplied reference, distorted hands, extra limbs, warped faces, unreadable title text, logos, watermark, or cheap cosplay look.

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