Skip to gallery
MeiGenmg_9d48796152imagezh

围绕具体主题内容生成一张竖向结构视觉作品,让主题核心先被压缩成暗色骨架、负形空洞和局部识别边缘,再从大面积主题场域内部显影;主体不完整站在背景前,而像被透明介质…

Original prompt

围绕具体主题内容生成一张竖向结构视觉作品,让主题核心先被压缩成暗色骨架、负形空洞和局部识别边缘,再从大面积主题场域内部显影;主体不完整站在背景前,而像被透明介质包裹、遮挡、折射后只留下必要的质量、方向和边缘。画面建立一个高明度、低噪声的主场,主场承担空间、记忆和信息呼吸的功能,必须直接切过主体边缘,使轮廓出现轻微断续、弯折、漂浮和被吞没的感觉。主视觉沿一条强轴线展开,中部形成最密的识别核心,下部或边缘被画幅裁断,保留画外延展感;对象细节被删减为少量断片、轮廓方向、结构空洞和暗色质量块,让主题由边缘取舍成立。背景不是装饰纹理,而是由主题自身推导出的可见介质层,带有尺度不同的场域扰动、透明覆盖、表面折射和低噪声痕迹;这些扰动应跨过主体、文字和空白,让整张图像像隔着一层湿润透明表面被观看。色彩保持大面积高明度主色、少量低明度结构色、透明轻高光和极小高能注记的比例关系,主题变化只改变色温、明度倾向和语义来源,不改变场域大、结构重、注记小、过渡轻的层级。暗色结构集中承担主体、细线和标签的重量,轻色高光只停留在介质弧线、边缘过渡和表面反射处,不能与主体争夺层级。必要文字采用克制的技术感无衬线,字形清楚、尺度小、宽字距,像坐标、年份、索引、边界或标签贴近信息轴;文字与图像同层,被场域轻微压住,形成横向测量感和低声说明感。只在一个关键小信息节点放入极小面积的高能色轨迹或圈注,让它成为冷静画面中的精确记忆点,面积必须被严格控制,像一次标记而不是一套装饰。整体完成度像平面印刷被透明湿润表面重新观看,保留轻微扫描颗粒、细折痕、青白高光和被介质软化的暗色边缘。画面应清楚、克制、潮湿、图形化,以强轴线、低细节主体、场域吞没、微型信息和单点注记构成同一套视觉系统,使不同主题都像被同一位设计者压入同一种场域逻辑中。

————
本次:长沙植物交换 线下活动 海报

比例:16:9横版

English translation

Generate a vertically structured visual work around a specific theme content, compressing the theme core into a dark skeleton, negative voids, and locally recognizable edges first, then developing it from within a large area of the theme field; the subject does not stand completely in front of the background, but is like being wrapped, blocked, and refracted by a transparent medium, leaving only necessary mass, direction, and edges. The image establishes a high-brightness, low-noise main field, which functions to provide space, memory, and information breathing; it must directly cut through the subject's edges, causing the outline to appear slightly discontinuous, bent, floating, and swallowed. The main visual unfolds along a strong axis, forming the densest recognition core in the middle, with the lower part or edges being cropped by the frame, retaining the sense of extension beyond the frame; object details are reduced to a few fragments, outline directions, structural voids, and dark mass blocks, letting the theme be established through edge取舍. The background is not decorative texture, but a visible medium layer derived from the theme itself, with field disturbances of different scales, transparent overlays, surface refractions, and low-noise traces; these disturbances should cross over the subject, text, and blank spaces, making the entire image look as if being viewed through a wet transparent surface. Color maintains the proportional relationship of large areas of high-brightness main colors, small amounts of low-brightness structural colors, transparent light highlights, and extremely small high-energy annotations; theme changes only affect color temperature, brightness tendency, and semantic source, without changing the hierarchy of large field, heavy structure, small annotations, and light transitions. Dark structures concentrate on carrying the weight of the subject, thin lines, and labels, while light highlights only stay on medium curves, edge transitions, and surface reflections, not competing with the subject for hierarchy. Necessary text uses restrained technical sans-serif, characters clear, scale small, wide character spacing, like coordinates, years, indexes, boundaries, or labels close to the information axis; text is on the same layer as the image, slightly pressed by the field, forming a horizontal measurement sense and low-volume explanation sense. Place only a small area of high-energy color track or annotation at one key small information node, making it a precise memory point in a calm picture; its area must be strictly controlled, like a single mark rather than a set of decorations. Overall completion is like平面印刷 being viewed again through a wet transparent surface, retaining slight scanning grain, fine creases, cyan-white highlights, and dark edges softened by the medium. The image should be clear, restrained, moist, graphic, using a strong axis, low-detail subject, field swallowing, micro-information, and single-point annotation to form the same visual system, making different themes look as if pressed into the same field logic by the same designer. ———— This time: Changsha Plant Exchange Offline Event Poster, Ratio: 16:9 horizontal version
Share to ChatGPT Original source
Model:GPT Image
Author:小小东 (@xiaoxiaodong01)
Collected:2026-10-23T12:04:45Z
Categories:advertising, landscape-nature, image-editing, camera-lighting
Attribution: Prompt text and metadata collected from publicly served pages of meigen.ai for research/archival purposes with attribution to the credited creator; copyright remains with original authors. Site robots.txt disallows /api/ — only public HTML pages were fetched.

Browse more in the gallery or explore this collection.

Search the archive

Find prompts and collections