黑白巨字城市构成风格对比海报

完整 Prompt

Convert the input city travel photo into a vertical 3:4 top-bottom contrast transformation poster; the top approximately 50% retains the city architecture, streets, vehicles, pedestrians, and landmark perspectives from the real photo; the bottom approximately 50% is translated into "MONOCHROME TYPE CITY | Black and White Giant Letter City Composition Style". The bottom half uses high-contrast black and white architectural silhouettes, carbon black and warm white, coarse halftone photocopy grain, uneven toner, and old paper wear, extracting the most accurate city landmarks from the photo, and using the giant cropped English title "CITY / STREET" as the composition skeleton, allowing letters and buildings to intersperse, bleed out of the frame, and obscure parts of the strokes. Only retain a maximum of one or two spot colors from the photo, no gradients or full-image tinting; the bottom half must change the scale of the subjects and information density and retain continuous low-information empty spaces. The midline is horizontally clear; no ordinary duotone filters, photo collages, irrelevant landmarks, logos, watermarks, fake Chinese, gibberish, or original/AI image icons/badges.

将输入的城市旅行照片制作成一张竖向3:4上下对比转绘海报，上半约50%保留真实照片中的城市建筑、街道、车辆、行人和地标透视；下半约50%转译为“MONOCHROME TYPE CITY｜黑白巨字城市构成风”。下半用高反差黑白建筑剪影、炭黑暖白、粗半调复印颗粒、碳粉不均和旧纸磨损，提取照片中最准确的城市地标，并用巨大裁切英文标题“CITY / STREET”作为构图骨架，让字母与建筑穿插、出画、遮挡部分笔画。只保留最多一至两种来自照片的专色，不做渐变或全图染色；下半必须改变主体尺度与信息密度并保留连续低信息空场。中线水平清晰，不要普通双色滤镜、照片拼贴、无关地标、Logo、水印、假中文、乱码或原图/AI图角标。

来源与署名

原作者

@Hamburgerai

公开出处

原始来源 ↗

收录图库

OpenNana 条目 ↗

来源模型标注

ChatGPT

条目日期

2026-10-02 · OpenNana 条目创建日期，不代表原帖发布日期

参考图要求

参考图要求未核实；请核对原帖和所用工具。

转载与权利

项目维护者确认转载授权；原作者权利保留

本站未逐条重新生图。原帖可访问性会变化，引用不代表独立验证作者身份或生成效果。