
GPT Image 2.0
動作電影感畫面:一名身穿紅色背心的健美女子在便利商店櫃檯上騰空飛踢對抗劫匪,GPT Image 2.0中生成
提示詞
創作一張電影感、寫實的動作場景定格畫面,場景設定在夜晚燈火通明的便利商店內。畫面中是一名膚色黝黑、留著深色長髮並綁成高馬尾的健美年輕女子,身穿紅色短版背心和黑色長褲,正處於打鬥之中,身體凌空定格:她側身躍過結帳櫃檯,一手撐在玻璃展示櫃上,一條腿伸出,踢出有力的飛踢。畫面中必須剛好出現四個人:1) 位於畫面中央偏左、神情專注而堅定的女主角;2) 位於最左側、戴著黑色滑雪面罩、穿著深灰色連帽上衣的蒙面搶匪,雙手舉起,身體向後閃避;3) 位於前景中央偏右、從背後拍攝、戴著黑色滑雪面罩並穿著深色連帽上衣的第二名蒙面搶匪,面向女子;4) 位於右側、穿著橄欖綠 T 恤和米色工裝褲、外表邋遢的男性旁觀者,站在零食貨架旁,手拿一瓶飲料,焦慮地注視著眼前情況。商店內部應包含擺滿各色小包裝商品的玻璃結帳櫃檯、左側的收銀機、櫃檯後方的滑動玻璃入口門和深色窗戶、天花板上的螢光燈反光、磚紅色磁磚地板,以及右側和後牆上多座擺滿色彩鮮豔的洋芋片袋與零食的貨架。採用寬幅 16:9 電影畫面、平視視角、動態定格效果、寫實的人體比例、粗獷的現代犯罪驚悚片氛圍、自然的商店螢光照明與夜間倒影交織、電影感淺景深、細緻的紋理,以及高預算電影劇照般的質感。避免文字疊加、標誌、字幕、浮水印、額外人物、血腥畫面或卡通風格。
Create a cinematic, realistic action still set inside a brightly lit convenience store at night. The scene shows an athletic young woman with tan skin and long dark hair in a high ponytail wearing a red cropped tank top and black pants, frozen mid-fight as she vaults sideways over the checkout counter with one hand planted on the glass display case and one leg extended in a powerful flying kick. Count exactly four visible people: 1) the woman hero at center-left, focused and intense; 2) a masked robber on the far left in a black ski mask and dark gray hoodie, recoiling with both hands raised; 3) a second masked robber in the foreground center-right, seen from behind in a black ski mask and dark hoodie, facing the woman; 4) a scruffy male bystander on the right wearing an olive green T-shirt and beige cargo pants, standing near snack shelves and holding a bottled drink while watching anxiously. The store interior should include a glass checkout counter filled with colorful small packaged goods, a cash register on the left, sliding glass entrance doors and dark windows behind the counter, fluorescent ceiling reflections, brick-red tile flooring, and multiple racks of bright chip bags and snacks on the right and back wall. Use a wide 16:9 movie frame, eye-level perspective, dynamic frozen motion, realistic anatomy, gritty modern crime-thriller mood, natural fluorescent store lighting mixed with night reflections, shallow cinematic depth of field, detailed textures, and a high-budget film still look. Avoid text overlays, logos, captions, watermarks, extra characters, gore, or cartoon styling.
此提示詞的權利屬於作者 @Spectromachina。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 →











