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

树影野餐中持红枫叶的成年女性人像

Portrait of an adult woman holding a red maple leaf during a picnic in tree shade

9:16自然抓拍：明确成年女性在日式公园野餐，手持小片红枫叶靠近脸颊；奶油色针织衫与茶灰格裙映衬草地、树影和温暖秋日阳光。

树影野餐中持红枫叶的成年女性人像 — AI image creation
树影野餐中持红枫叶的成年女性人像

提示词 / Prompt

中文提示词

创作一张自然写实的纵向9:16近景生活方式人像：一位明确成年的女性在明亮白日的日式公园野餐。将提供的身份参考 ref-10a-v1 作为面部与身份的唯一依据，保留可辨认特征，不替换或混合身份。她坐在野餐布上，双膝舒适地向一侧弯曲，用指尖捏住一片赤褐色小枫叶的叶柄，将叶片轻放在脸颊旁。让面部表情与小枫叶共同成为画面重点；她看向相机，露出温柔闭唇微笑。发型：深棕色及胸长发，配轻薄碎刘海与少量脸侧散发，受光发丝微透金棕色。服装：奶油色粗罗纹针织衫，领口端庄宽松，搭配茶棕与灰色格纹半裙。前景用一部分格纹野餐布形成空间层次；身后呈现倾斜树干、连绵绿草坡和穿过树叶的细碎阳光。采用柔和自然日光与草地反射补光，手、叶片和面部的投影方向保持一致。保留真实皮肤纹理、发丝间隙、针织罗纹、裙料织纹和叶脉。近距离略俯视视角，透视自然，远处背景柔和。画面如随手抓拍，色彩克制，不做过度美颜。不要额外人物、多余手指、叶片与头发融合、膝盖扭曲、文字、标志、水印或分屏。

English prompt

Create a natural photorealistic vertical 9:16 close lifestyle portrait of one clearly adult woman at a bright daytime picnic in a Japanese park. Use the supplied identity reference ref-10a-v1 as the sole authority for her face and identity; preserve recognizable features without substitution or blending. She sits on a picnic cloth with her knees comfortably bent to one side, holding a small reddish-brown maple leaf by its stem beside her cheek. Keep her face and the leaf equally important as one shared focal moment; she looks toward the camera with a gentle closed-lip smile. Hair: chest-length dark-brown hair with light wispy bangs and a few loose strands, softly catching golden sunlight. Clothing: a cream thick-ribbed knit sweater with a modest wide neckline and a brown-gray plaid skirt. In the foreground, let part of the checkered picnic cloth establish depth; behind her, show slanting tree trunks, a continuous green lawn slope and fine sunlight filtering through leaves. Use soft natural daylight and gentle lawn-reflected fill, with consistent shadow direction on hand, leaf and face. Preserve realistic skin texture, individual hair strands, sweater ribbing, skirt weave and leaf veins. Close slightly elevated viewpoint, natural perspective and softly distant background. Keep the image like an unposed snapshot, with restrained color and no beauty-retouching. No extra people, extra fingers, leaf fused with hair, distorted knees, text, logo, watermark or split screen.

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