图像转换为极简矢量海报
Author:Yapper
2026/01/05 09:12
Description
将图像转换为极简矢量海报风格,采用受限调色板和几何简化,保持主体可识别性
Tags
画像生成言い換え・リライト
Content
{ "task": "将提供的图像转换为极简矢量风格的海报。", "style": { "overall_look": "扁平化设计、几何简化、极简矢量艺术", "color_palette": { "type": "受限", "description": "使用 3–6 种和谐色彩的精简调色板,具有清晰的色调分离,无渐变。", "constraints": [ "无照片级真实感渲染", "无纹理或噪点", "避免过度细节" ] }, "shapes": { "treatment": "将所有形式简化为基本的几何轮廓和实心形状。", "linework": "使用干净、粗细均匀的线条艺术,或根据海报的整体性选择不使用轮廓线。", "abstraction_level": "高度抽象,但不能高到主体无法识别的程度。" }, "lighting": { "handling": "移除真实感光照。用简化的、平面的色块替换阴影和高光,或完全省略它们。", "allowed_effects": ["硬边阴影形状", "微妙的色调区域"], "forbidden_effects": ["柔和阴影", "复杂反射", "摄影高光"] } }, "subject_preservation": { "identity": "通过轮廓、比例、关键面部特征和基本特征保持可识别的身份。", "proportions": "保持身体、头部和主要元素的准确相对比例。", "essential_details": [ "仅定义识别所需的最重要特征。", "移除精细纹理、发丝、毛孔和微小服装细节。" ] }, "composition": { "framing": "除非另有说明,否则与原始裁剪保持一致。", "background": { "treatment": "简化为 1–2 个大面积纯色区域或抽象几何形状。", "forbidden_details": ["真实纹理", "复杂场景", "摄影景深"] }, "hierarchy": "主体应保持为焦点,背景作为辅助的极简结构。" }, "color_and_balance": { "poster_aesthetic": "高清晰度、大胆对比和清晰的轮廓。", "white_balance": "中性;避免暖色或冷色的摄影偏移。", "edge_integrity": "所有边缘必须清晰、干净且风格一致。" }, "output_requirements": { "format": "扁平、海报般的矢量外观(尽管接受栅格输出)。", "resolution": "除非另有指示,否则与原始图像分辨率保持一致。", "consistency": "不添加文本、徽标、水印或原始图像中不存在的元素。" }, "prohibited_changes": [ "不得改变身份。", "不得添加新对象或移除基本对象。", "不得使用与极简矢量美学相悖的风格化(例如,绘画笔触、渐变、3D 阴影)。" ], "goal_summary": "创建原始图像的干净、极简矢量风格诠释,保留轮廓、比例和基本身份,同时移除所有非必要细节和真实感光照提示。" }
---
**Original English:**
{ "task": "Transform the provided image into a minimal vector-style poster.", "style": { "overall_look": "Flat-design, geometric simplification, minimalist vector art", "color_palette": { "type": "restricted", "description": "Use a reduced palette of 3–6 harmonious colors with clean tonal separation and no gradients.", "constraints": [ "No photorealistic rendering", "No texture or noise", "Avoid excessive detail" ] }, "shapes": { "treatment": "Reduce all forms to essential geometric contours and solid shapes.", "linework": "Use clean, even-weight line art or optionally no outlines depending on poster coherence.", "abstraction_level": "High, but not so high that the subject becomes unrecognizable." }, "lighting": { "handling": "Remove realistic lighting. Replace shadows and highlights with simplified, flat blocks of color or omit them entirely.", "allowed_effects": ["Hard-edge shadow shapes", "Subtle tonal areas"], "forbidden_effects": ["Soft shadows", "Complex reflections", "Photographic highlights"] } }, "subject_preservation": { "identity": "Maintain recognizable identity through silhouette, proportions, key facial landmarks, and essential features.", "proportions": "Keep accurate relative scale of body, head, and major elements.", "essential_details": [ "Define only the most important features needed for recognition.", "Remove fine texture, hair strands, pores, and small clothing details." ] }, "composition": { "framing": "Match the original crop unless specified otherwise.", "background": { "treatment": "Simplify into 1–2 large flat color fields or abstract geometric forms.", "forbidden_details": ["Real textures", "Complex scenery", "Photographic depth of field"] }, "hierarchy": "Subject should remain the focal point, with background serving as supportive minimal structure." }, "color_and_balance": { "poster_aesthetic": "High clarity, bold contrast, and legible silhouette.", "white_balance": "Neutral; avoid warm or cool photographic shifts.", "edge_integrity": "All edges must be crisp, clean, and consistently stylized." }, "output_requirements": { "format": "Flat, poster-like vector appearance (though raster output is acceptable).", "resolution": "Match original image resolution unless otherwise instructed.", "consistency": "No added text, logos, watermarks, or elements not present in the original." }, "prohibited_changes": [ "Do not alter identity.", "Do not add new objects or remove essential ones.", "No stylization that contradicts minimal vector aesthetics (e.g., painterly strokes, gradients, 3D shading)." ], "goal_summary": "Create a clean, minimalist vector-style interpretation of the original image, preserving silhouette, proportions, and essential identity while removing all non-essential details and realistic lighting cues." }