Lead with why the PDF/polygon is the artefact being priced

README.md

@@ -11,7 +11,9 @@ short_description: "Read lines from a drawing PDF: vector paths vs pixels"
 
 # floorplan-parser showcase
 
-Do you need computer vision to read construction drawings? floorplan-parser pulls line geometry out of a drawing PDF two ways: by reading the vector paths the PDF already stores (PyMuPDF), and by rendering the page to pixels and finding the lines with classic computer vision (OpenCV threshold and Hough lines). Both are scored against the drawing's own geometry.
+Why read the PDF at all, instead of the CAD file? Because the estimator usually does not have the model: it stays with the consultant, and the set that gets issued, priced and revised is a PDF. A CAD file is not a takeoff either — a wall is a pair of lines, a door is a block, dimensions are text — so the geometry still has to be read and measured, again on every revision.
+
+So: can the PDF give up its own lines? floorplan-parser answers by pulling line geometry out of a drawing PDF two ways: reading the vector paths the PDF already stores (PyMuPDF), and rendering the page to pixels and finding the lines with classic computer vision (OpenCV threshold and Hough lines). No trained model on either route: the target is a line, not a room, and Hough is deterministic and inspectable. Both are scored against the drawing's own geometry.
 
 The pictures are synthetic drawings made for this project; pick a drawing and a method to see what was found (blue), missed (red) and added (orange). The table also lists measured results on ten held-out FloorPlanCAD drawings, as numbers only, including `--chain`, which joins the pieces of each curve. A last panel shows the lines → walls → rooms step on a synthetic plan, with its scores on synthetic plans and on real CubiCasa5K layouts. Everything was computed ahead of time; the page runs no model and no server.