穿黃色洋裝的女孩手拿粽子，站在重慶老城陡峭的街道上，周圍有傳統招牌與綠意

提示詞

建立一幅電影感、寫實的直式街頭人像，場景設定在端午節期間的重慶老山城街區。一位年輕女子站在陡峭向下延伸、由石階和柏油路面構成的街道中央，面向鏡頭，雙手在胸前高度捧著一顆綠色三角形竹葉粽。她穿著一件黃色碎花細肩帶短洋裝和白色球鞋，短短的深色頭髮上別著一個小紅蝴蝶結髮夾；為了保護隱私，她的臉應柔和地模糊或遮擋，但身體、衣著和粽子要保持清晰細緻。街道狹窄而富有戲劇感，兩旁排列著老舊的多層建築、木門、陽台、老店面、垂掛的植物、盆栽綠意、裸露的電線、燈籠造型的直式招牌，以及傳統的中文招牌。恰好加入五個醒目且清晰可讀的招牌：左側一個溫暖發光的直式招牌，寫著「十八梯」；一塊藍色小門牌，寫著「十八梯 18」；中景一個直式霓虹招牌，寫著「洪崖洞」；右側一個吊掛招牌，寫著「茶」；以及一塊木製店招，寫著「茶葉店」。道路陡峭地向下延伸，通往遠方背景中密集的現代天際線與河港，高樓塔樓在霧氣中柔化。光線為午後偏晚時分從左側灑落的金色陽光，在街道上投下長長的陰影，營造溫暖懷舊的賽博寫實城市氛圍。採用高視角，帶有 35mm 紀實鏡頭的感覺、深遠的透視、寫實的材質、高動態範圍、細微的景深霧感和自然色彩；主角附近不要有其他人，不要浮水印，不要字幕，不要卡通風格。

Create a cinematic photorealistic vertical street portrait in an old hillside neighborhood of Chongqing during the Dragon Boat Festival. A young woman stands centered on a steep descending stone-and-asphalt stair street, facing the camera and holding one green triangular bamboo-leaf rice dumpling, a zongzi, with both hands at chest height. She wears a short yellow floral spaghetti-strap sundress, white sneakers, and a small red bow hair clip in short dark hair; her face should be softly blurred or obscured for privacy while keeping the body, clothes, and zongzi sharply detailed. The street is narrow and dramatic, lined with weathered multi-story buildings, wooden doors, balconies, old shopfronts, hanging plants, potted greenery, exposed wires, lantern-like vertical signs, and traditional Chinese signage. Include exactly five prominent readable signs: a warm glowing vertical sign on the left reading 「十八梯」, a small blue address plaque reading 「十八梯 18」, a vertical neon sign in the mid-distance reading 「洪崖洞」, a right-side hanging sign reading 「茶」, and a wooden shop sign reading 「茶葉店」. The road drops steeply into a dense modern skyline and river harbor in the far background, with high-rise towers softened by haze. Lighting is late-afternoon golden sunlight from the left, casting long shadows down the street and creating a warm nostalgic cyber-real city atmosphere. Use a high vantage point with a 35mm documentary lens feel, deep perspective, realistic textures, high dynamic range, subtle depth haze, natural colors, no extra people near the subject, no watermark, no captions, no cartoon styling.