Skip to contents

Scores every square against every template of every installed set in a single matrix multiply, then picks the best-matching set and reports confidence signals (see the inline comments for how they are calibrated).

Usage

recognize_symbols_auto(squares, libs = load_all_template_libraries())

Arguments

squares

A list of 64 magick square images, row-major, top row first.

libs

Named list of template libraries to score against (default: every locally available set).

Value

A list with symbols (from the winning set), set, scores (per-set mean similarity), margin (winner minus runner-up), median_occ (typical occupied-square match) and confident (whether both gates pass).