OpenZhenZhen AI 提示词画廊 · gpt-image-2.5 · image
霓虹故障动漫旅行双联编辑海报
Neon Glitch Anime Travel Double-Panel Poster
来自 OpenNana 的图像提示词，已按本地图像分类和站点 gpt-image-2.5 归类整理。 使用时需上传人物参考图；本流程使用统一参考人物生成示例。
提示词 / Prompt
中文提示词
以用户上传的参考图作为主要可见人物的面部身份依据。在不与身份参考冲突的情况下，保留来源的场景、造型、服装、姿势、构图、光线和视觉风格。 将人物渲染为风格化插画，只保留参考图的面部身份特征；不要生成写实摄影人像。 请将我上传的每一张人物照片分别制作成一张独立的高级旅行肖像编辑海报，不多图拼接，每张照片单独输出。整体采用3:4竖版双联构图，上方保留清楚可见的真实原图，下方进行“霓虹故障动漫肖像”转绘；上下边界水平、明确、克制，不雾化、不互相渗透。 上半部分约占画面45%–50%，忠实保留原始照片。准确保持人物身份、数量、脸型、五官关系、发型、眼镜、目光、表情、姿态、手势、衣着、随身物件以及人物与环境的空间关系。只进行轻微的艺术杂志式调色与极薄胶片颗粒处理，仍必须明显是真实摄影。为适配画幅，可自然扩展天空、墙面或周边环境，但不得重画人物、改变面孔、拉伸、镜像、复制或移动主体。 下半部分从同一张照片中提取人物最具识别度的面部、发型、服装、手势和关键物件，重构为具有强烈画面冲击的当代动漫编辑肖像。不要照搬原照片的完整取景和背景密度；将人物半身明显放大，占下区宽度约65%–95%，可适度裁切出画，让视线、动作方向和轮廓成为主导。背景只保留一至三条来自原图的建筑线、灯光、栏杆、家具或空间结构，并压缩成低信息光色场与结构碎片。 转绘使用角面化色块、清楚的动漫线面关系、深靛蓝阴影、钢蓝与青色高光，并从原图中选择一处真实存在的暖色，转译为克制的珊瑚粉或品红跳色。加入有限网点、局部RGB套色偏移、边缘信号断裂和少量像素切块。故障效果只能发生在头发边缘、肩部轮廓、衣物外缘及背景结构，不得覆盖或破坏脸、眼睛、手和关键物件；不得改变人物身份，也不得生成无关的赛博城市。 下区保留约18%–28%连续的深蓝或低信息色面作为呼吸区。文字集中成一组，主标题使用在正常聊天预览中清楚可读的粗窄游戏标题字，带轻微像素切口或RGB错位；辅助字段较小，但不能缩成灰线。 根据照片中真实可见的地点、人物状态、动作或光线提炼1–3个英文单词标题，并加入字段“SIGNAL NOTE 01”与一行5–12个英文单词的观察句。若地点无法确认，不得编造地名，可使用标题“AFTERLIGHT”，观察句使用“A quiet face held inside moving light.”。只渲染已确定的文字，不添加作者名、品牌、账号、Logo、二维码或乱码。 整体色彩从原照片提炼并压缩为有限色盘，保持人物衣物和环境原有的冷暖性格。画面应兼具真实旅途气息、动漫人物张力、电子游戏封面感和编辑出版物的克制感；故障是节奏，不是遮挡。 避免通用赛博朋克、满屏霓虹彩虹、密集HUD、小到不可读的界面字、蒸汽波模板、面部故障、额外肢体、廉价滤镜和全画面插画化。
English prompt
Use the uploaded reference image as the facial identity reference for the main visible person. Keep the source scene, styling, clothing, pose, composition, lighting and visual style unless they conflict with the identity reference. Render the person as a stylized illustration that preserves the reference facial identity cues; do not turn the result into a photorealistic camera portrait. Please turn each character photo I upload into an individual high-end travel portrait editorial poster, no multi-image stitching, output each photo separately. Overall use a 3:4 vertical diptych composition, with the top retaining the clearly visible real original image and the bottom performing a "neon glitch anime portrait" redraw; the upper and lower borders are horizontal, clear, and restrained, with no atomization or mutual penetration. The upper part accounts for about 45%–50% of the frame, faithfully retaining the original photo. Accurately maintain character identity, quantity, face shape, facial feature relationships, hairstyle, glasses, gaze, expression, posture, gestures, clothing, personal items, and the spatial relationship between characters and the environment. Only perform slight art magazine-style color grading and extremely thin film grain processing, it must still clearly be real photography. To adapt to the frame, the sky, walls, or surrounding environment can be naturally extended, but the character must not be redrawn, faces changed, stretched, mirrored, copied, or the subject moved. The lower part extracts the most recognizable face, hairstyle, clothing, gestures, and key items of the character from the same photo, reconstructing them into a contemporary anime editorial portrait with strong visual impact. Do not copy the full framing and background density of the original photo; significantly enlarge the character's half-body to occupy about 65%–95% of the lower area's width, which can be moderately cropped out of the frame, letting the line of sight, action direction, and silhouette become dominant. The background retains only one to three architectural lines, lights, railings, furniture, or spatial structures from the original image, compressed into low-information light-color fields and structural fragments. The redraw uses angular color blocks, clear anime line-plane relationships, deep indigo shadows, steel blue and cyan highlights, and selects one existing warm color from the original image to translate into a restrained coral pink or magenta accent color. Add limited halftones, partial RGB color offset, edge signal breaks, and a small amount of pixel chunks. Glitch effects can only occur at hair edges, shoulder silhouettes, clothing outer edges, and background structures, and must not cover or destroy faces, eyes, hands, and key items; must not change character identity, nor generate irrelevant cyber cities. The lower area retains about 18%–28% of continuous deep blue or low-information color surface as a breathing area. The text is concentrated into a group, the main title uses bold narrow game title fonts that are clearly readable in normal chat previews, with slight pixel cuts or RGB misalignments; auxiliary fields are smaller, but cannot be shrunk into gray lines. Extract a 1–3 English word title based on the actually visible location, character state, action, or light in the photo, and add the field "SIGNAL NOTE 01" and a 5–12 English word observation sentence. If the location cannot be confirmed, do not fabricate place names, the title "AFTERLIGHT" can be used, and the observation sentence "A quiet face held inside moving light." Only render confirmed text, do not add author names, brands, accounts, logos, QR codes, or garbled characters. The overall color is extracted from the original photo and compressed into a limited palette, maintaining the original warm and cold character of the character's clothing and environment. The image should have a sense of real travel atmosphere, anime character tension, video game cover feel, and editorial publication restraint; glitch is rhythm, not obstruction. Avoid generic cyberpunk, full-screen neon rainbows, dense HUD, UI text too small to read, vaporwave templates, facial glitches, extra limbs, cheap filters, and full-frame illustrationization.
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
相关作品
- 90年代中国乡野家庭童年抓拍
可复用的竖版135胶片家庭生活照模板：表现90年代中国乡野普通家庭中孩子自然玩耍的一瞬，使用可替换的童年行为变量；保留日常抓拍的真实、随意与胶片偶然性。
- 手绘岩彩插画
竖幅的手绘岩彩插画，主题是一片竹影与一轮将隐的月。画面不追求完整构图，像从一段夜色里裁下来的一角：竹从画面右侧斜出，向左侧留出大片空白，月光落在留白的最上方。第一眼读到的是墨绿，第二眼才是金。 底子是暖白的宣纸，纸面有细密的纤维与一层极浅的磨砂颗粒，颗粒是纸本身的质地，不是浮在画上的噪点。整张画在纸的暖白上压了一层岩彩的底色：先用水墨淡彩铺出一层稀薄的青绿，让颜色顺着纸纤维渗开，边缘有不规则的渗边；再在上层叠矿物颜料的颗粒，颗粒在中景最清楚，越往边缘越散。 竹是这幅画的主角，但不需要画全。竹竿用湿墨一笔带过，起笔重、收笔飞白，墨色在纸里洇开成一团边缘模糊的深墨绿；竹叶不做单片交代，用侧锋扫
- desert dessert
16:9横屏，盛夏沙漠绿洲实景中的真人COS写真。一位29岁的成年女性扮演原创越野机械师，深栗色长发编成略松的辫子，小麦色肌肤，眉骨清晰，鼻尖与脸颊带轻微晒红，颈侧有细小汗珠。胸部非常丰满，肩背舒展，穿砖红色深V宽肩带比基尼，罩杯贴合、乳沟清晰自然，搭配沙米色泳裤与窄款工具腰带。泳装面料不透明，表面带细密织纹，古铜色扣件存在轻微磨痕；一双短指工作手套随意别在腰侧。 她站在绿洲修理棚的遮阳帆下，身体斜靠旧越野车的翼子板，一手拿着冰水瓶，另一手搭在车身边缘，望向镜头，笑容自信而松弛。炽烈的日光照亮棚外沙地，人物处于柔和阴影里，明亮沙面的反射光从下方轻轻托亮面部；遮阳帆边缘漏下的一道阳光落在肩头和
- rain
16:9横屏，真实港口外景的真人COS摄影。一位28岁的成年女性扮演原创深海勘探员，深棕色长发松松束起，耳边有几缕被潮气打湿的碎发，灰蓝色眼瞳，眉毛与睫毛自然清晰，脸颊保留细小毛孔。身形丰腴，胸部丰满，穿深海青色宽肩带低领比基尼，胸前以哑光银色搭扣连接，领口自然显出明显乳沟；搭配同色高腰泳裤，面料厚实不透，肩带边缘与金属扣上留有细小水珠。潜水面镜挂在腰侧，手腕戴一只略有使用痕迹的机械深度表。 她坐在码头雨棚下的木箱上，身体微侧，一手扶着箱沿，另一手拿着搪瓷杯，抬眼望向镜头，神情安静而有精神。雨刚停，天空呈冷灰蓝色；雨棚右上方的暖色检修灯照亮面庞与肩头，形成明确的明暗分区，港湾天光轻轻补亮另一
- 柔粉短发天使与白羽翼维多利亚礼服圣像
竖版对称神圣幻想肖像：柔粉短发成年天使侧身低眸，身着维多利亚蕾丝礼服、持金雕长弓，巨大白羽翼与花卉曼陀罗挂毯光环构成极繁复但清晰的装饰画面。
- vacation time
16:9横屏，真人COS写真，一位27岁的成年女性扮演原创海岛飞行员。铜棕色短发，浅绿色眼瞳，头顶戴奶油白飞行护目镜。身形丰腴，上围格外丰满，肩颈舒展。穿珊瑚橙色宽肩带低领比基尼，胸前以小巧银色搭扣连接，自然形成清晰乳沟，下身搭配奶油白拼色泳裤，布料不透明。她倚在海边木栈桥的栏杆旁，一手拿冰橙汽水，另一手扶着护目镜，笑着望向镜头。人物位于右侧，平视膝上构图，左侧是碧蓝海湾与停泊的白色小飞艇。夏日阳光经过遮阳棚柔化，橙色泳装与蓝色海面鲜明呼应，真实杂志摄影质感，无文字、无水印。