Dominik Roth
ff3b89c41b
Reads a screenshot of the in-game map table and recovers where on the board it is looking. The scene is a flat table under a perspective camera, so a single homography describes grid<->screen exactly; that is fitted from line evidence (LSD segments, vanishing-point RANSAC, 1-D lattice fits). Line evidence alone cannot finish the job: the 1km cells and the 100m subgrid are locally identical, so it fixes neither scale, axis assignment, axis direction nor phase. Those come from the cell labels -- and the labels are never *detected*, they are correlated where the grid says they must be (9% in from a cell's left edge, 6% down from its top). Blob detection on aerial-photo terrain finds texture, not glyphs; correlating a known template at a known place has no such failure mode. The same score then ranks the geometry hypotheses, since a wrong lattice puts the crop where no label is, so one number resolves scale, axis assignment, direction, phase and anchor together. Markers are found by hostile/friendly colour plus an IoU test against an ideal inscribed diamond, which cut a red-lit shot from 43 false positives to 2 while preserving every hand-verified count. Unit-type classification is present but not yet reliable, and returns unknown rather than guessing. Measured over the fixture set: 7 of 10 solve, each with 100% of its ground-truth points in the correct cell (85/112 overall), residual spread 0.005-0.033 cells. The other three reject cleanly; none has ever produced a plausible-but-wrong grid. Across 122 typewriter screenshots solve() accepted none, which is what makes clipboard routing safe. Fixtures: 10 map shots with hand-transcribed ground truth, plus 10 typewriter shots spanning 262-5366px for routing and future OCR tests. Map shots wider than 2400px (the pipeline's own maximum working width) were downscaled with their coordinates rescaled to match; re-measured afterwards, the results are identical. Typewriter shots stay at native resolution because OCR needs the text legible. Drops docs/map_vision_plan.md and its WIP prototype: the design now lives in the module docstring, next to the code it describes. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
747 B
747 B
Screenshots we deliberately do not try to solve
Kept for the record, excluded from the evaluation set. These fail for reasons that are properties of the screenshot, not of the algorithm, so working around them would mean guessing:
06.png,14.png— exactly ONE grid label visible. A single label cannot be cross-checked, so a misread would silently shift the whole board with nothing to contradict it. Two mutually consistent labels is the minimum safe anchor.12.png— 710x594 native. Too few pixels per cell for the label glyphs to correlate; measured, the best label score stays ~0.44 at every working resolution, so it is not a tuning problem.
"Too zoomed in" and "too low resolution" are legitimate hard rejections.