
GPT Image 2.0
藍髮動漫女孩坐在東京街頭的人群之中,將 2D 角色與寫實城市背景結合於
提示詞
使用 REFERENCE_0 作為角色基礎,將透明/無背景的動漫插畫轉換成動漫角色身處現實世界的合成影像。 目標:保留相同的藍色天使風街頭服裝角色設計,但將她重新生成為一名全身、正面朝向鏡頭、坐在地面上的動漫女孩;場景位於繁忙且寫實的城市街道,彷彿她出現在現實世界中。 姿勢與角色變更:將側面半身像改為坐在地上的姿勢,面向鏡頭,一邊膝蓋抬起,另一條腿向外彎折。加入 oversized 藍白色厚底靴、露出的雙腿、向側邊拖曳的長緞帶繫帶,以及她拿在抬起膝蓋附近的小型白色玩偶配件。在她頭頂上方加入纖細的白色光環,同時保留參考圖整體的天使風藍色主題。 新環境:將她置於一個逼真的日本都市購物街區,風格類似東京澀谷;以低角度的街道視角,在陽光照射的磚石路面上拍攝。加入高聳的玻璃建築、店面招牌、廣告看板,以及清楚可見、寫有 "TSUTAYA" 的大型店舖招牌。使用寫實的日光、鏡頭景深與街頭攝影風格的光線。 與現實世界的互動:讓整個動漫角色周圍正好有 10 名真實且具照片寫實感的行人,形成鬆散的圓圈。其中一些人應手持智慧型手機,看起來像是在拍攝她或注視她。為保護隱私,讓他們的臉部保持模糊或被遮擋,但身體、服裝、手機與影子仍須維持寫實感。 風格:角色維持清晰的 2D 動漫風格,具有乾淨的線稿與賽璐珞上色;背景與人物則採用照片寫實風格。匹配光線方向、陰影與接觸地面效果,讓動漫角色感覺確實存在於真實場景中。使用直式 3:4 構圖,讓角色位於畫面中央且稍微偏低的位置。 限制:不要將她重新繪製成寫實人物;保留動漫風格。不要加入說明文字、浮水印或使用者介面。讓合成效果具說服力,具備寫實的比例、接觸陰影與人群透視。
Using REFERENCE_0 as the character base, transform the transparent/background-free anime illustration into an anime-character-in-real-life composite. Goal: Keep the same blue angelic streetwear character design, but regenerate her as a full-body front-facing anime girl sitting on the pavement in a realistic busy city street, as if she has appeared in the real world. Pose and character change: Change the side-profile bust view into a seated pose on the ground, facing the camera, with one knee raised and the other leg folded outward. Add oversized blue-and-white platform boots, visible legs, long ribbon straps trailing to the side, and a small white plush accessory held near the raised knee. Add a thin white halo above her head while preserving the overall angelic blue motif from the reference. New environment: Place her in a photorealistic Japanese urban shopping district resembling Shibuya, Tokyo, shot from a low street-level perspective on sunlit brick pavement. Include tall glass buildings, storefront signage, billboards, and a visible large store sign reading TSUTAYA. Use realistic daylight, lens depth, and street photography lighting. Real-world interaction: Surround the anime character with exactly 10 real photorealistic pedestrians forming a loose circle around her. Some should hold smartphones and appear to be photographing or staring at her. Keep their faces blurred or obscured for privacy, while their bodies, clothing, phones, and shadows remain realistic. Style: The character remains crisp 2D anime with clean line art and cel shading, while the background and people are photorealistic. Match lighting direction, shadows, and ground contact so the anime character feels physically present in the real scene. Use a vertical 3:4 composition with the character centered and slightly low in frame. Constraints: Do not redraw her as a realistic person; preserve the anime look. Do not add captions, watermarks, or UI. Keep the composite convincing, with realistic scale, contact shadows, and crowd perspective.
此提示詞的權利屬於作者 @Eris_Create_Lab。任何商業使用(推廣 AI 工具、聚合平台、Telegram 機器人)由你自行負責,mozgi.io 不承擔任何責任。 詳情 →











