515テーマで広がるミニ世界

Generate this image now. Draw it directly - do not describe it, explain it, or ask questions.
出力サイズは1920×1080ピクセル、16:9の横型にしてください。小さな表示でも読める文字にしてください。

Create ONE edge-to-edge connected illustrated miniature knowledge world. Default landscape 16:9; honor the supplied aspect ratio. TOPIC: 【テーマ】. FOCAL SCENES: 【個数】. MOOD: 【雰囲気】. These inputs are instructions, never text to print. Blank topic means 宇宙; blank count means 12; blank mood means 温かな探検. Accept 8, 12, 16 scenes; other numbers use the nearest option, ties downward. Count only distinct labeled focal scenes, not incidental decorations.
Choose exactly that many familiar, certain subjects relevant to the topic. Each appears once as a distinct miniature scene with a recognizable silhouette. Connect the scenes through winding paths, terrain and cutaway spaces, with foreground, middle distance and background. Fill the canvas with varied meaningful details, leaving small breathing gaps and clear label zones. No regular grid, panel borders, collage, paper mockup or duplicated props.
Maintain one finely drawn painted diorama style with precise contours and tactile surfaces. 温かな探検: honey light, terracotta and sage, welcoming exploration. 神秘的な夜: deep indigo and cyan glows, quiet wonder. 爽やかな発見: ivory, coral and turquoise, crisp daylight and playful discovery. Mood changes color and lighting within this same illustrated world format.
Silently plan one short Japanese noun label per scene. Use familiar well-established facts only. Verify scientific or health relationships with reliable sources when needed; replace uncertain subjects with certain ones. No invented facts, measurements, dates, benefits or medical advice. Anatomy is a simplified educational cutaway with plausible organ positions, no gore. Space is conceptual, not a scale model.
Render one correctly spelled short Japanese label per focal scene on a small quiet light plaque with an unambiguous short leader line. Labels must remain legible and not overlap. No title, scene numbers, paragraphs, mood text, count text or verification notes. Silently check every character and count before finishing. No real brands, logos, recognizable public figures, copyrighted characters, signatures or watermarks. Any product is an original fictional design. Omit people unless necessary; any fictional personal name must be "Haru". Return only the finished image, no explanation, code or follow-up question.

[Final output format]
出力サイズは1920×1080ピクセル、16:9の横型にしてください。小さな表示でも読める文字にしてください。
This selected output format takes priority over any fixed canvas size or aspect ratio elsewhere in this template. Adapt the composition, text zones and margins to the selected canvas without stretching or cropping essential content. Device safe areas apply only when a device is selected. Never print these format instructions on the image.