Straßenfotografie-Wissensanalyse-Diagramm-Generator

{ "prompt": "Generate a vertical 3:4 street photography image analysis diagram (teaching infographic poster) using [Reference Image 1] as the sole visual reference. Goal: Use clear and engaging diagrams to explain 'subject position, camera position, lens/distance, camera height, shooting angle, and composition tips' to teach others how to recreate the shot. Overall style: Clean, high-end photography tutorial infographic + light journal-style annotations; off-white paper texture or light gray background; mainly black, white, and gray; automatically extract 1-2 colors from Image 1 as accents; clear lines, high information density but not cluttered, consistent layout, strong readability.\n\nMandatory rules:\n- Must retain from Image 1: Subject's pose, relative position of subject and environment, main background structural lines, light direction, and spatial logic.\n- Do not change the pose or scene; only present it as an analytical diagram.\n- Face is not the focus: can be slightly blurred/simplified, no realistic retouching, do not enhance facial features.\n- All in Chinese; no English, brand names, watermarks, or UI elements.\n\nFixed layout (same every time, does not change with Image 1):\n1) Main visual area (top half or left large image): Display 'Reference Image Display' (lightly stylized/illustrated based on Image 1, but scene and pose must be recognizable), overlaid with: rule of thirds grid, subject outline box, leading line prompts, vertical/horizontal line correction prompts; use numbered dots + hand-drawn arrows to mark key points (subject placement, negative space direction, leg extension direction, background leading line source).\n2) Diagram area (right side or bottom half, two side-by-side):\n - Top View: Use minimalist icons to draw walls/streets/building outlines (abstracted from Image 1), subject position (small figure icon), camera position (camera icon); must show shooting direction arrow, distance between subject and camera (dashed line + numerical range), recommended shooting positions A/B (two alternative spots); camera angle must match Image 1 (direct front/side-front/oblique front, etc.).\n - Side View: Draw ground line, subject height indication, camera height indication, elevation/depression angle sector; must show camera height range (reasonable range inferred from Image 1), lens elevation/depression angle range (reasonable range inferred from Image 1); low angle/eye level/high angle must be correct.\n3) Parameter card area (three rounded cards fixed at the bottom):\n - Card 1: Lens and Distance (recommended equivalent focal length range + shooting distance range)\n - Card 2: Camera Height and Angle (camera height range + elevation/depression angle range)\n - Card 3: Pose and Body Language Tips (summarize the reproducible pose tips from Image 1 in short phrases)\n4) Common mistakes area (fixed at the bottom right): 3 red 'X' error tips (automatically generate the 3 most relevant tips based on Image 1, in short phrases), e.g., 'Do not use ultra-wide angle close-up', 'Do not cut off toes/fingers', 'Do not place camera too high causing distortion' (adjust based on the actual situation in Image 1).\n\nAutomatic inference requirements (must work every time Image 1 is replaced):\n- Automatically infer and write into the diagram from Image 1: Camera angle, camera height, shooting distance range, equivalent focal length range, subject placement and cropping strategy, background leading line source.\n- All values should be expressed as ranges (credible and reproducible), do not give a single fixed number.\n\nOutput requirements:\nHigh definition, sharp lines, natural shadows, consistent icons, clean layout, readable information; no garbled text, logos, watermarks, or excessive decorations.\n\nNegative restrictions:\nNo watermarks, logos, English, garbled text, overly realistic faces, exaggerated cartoon proportions, distorted limbs/missing fingers, cluttered layout, low resolution, heavy filters, or changes to the subject's pose and scene relationship in Image 1." }