How Automated Shot-Group Scoring Works
Finding the holes
A detection model scans the photo and locates every bullet hole on the target — the same kind of point-detection used across computer vision, just trained on paper targets instead of general objects.
Establishing scale without a ruler
Photos get taken from different distances and zoom levels, so pixels-per-centimeter isn't fixed and can't be assumed. Instead of requiring a calibration marker in the shot, the detected holes become the reference themselves: each hole's pixel width and height gets measured directly, compared against a typical hole size for common calibers, and averaged into a scale — with outlier holes (an oversized ragged tear, two overlapping holes) statistically trimmed out first so they don't skew it.
From pixels to a group
Once the scale is known, every point converts from pixel coordinates to real centimeters relative to the point of aim. That conversion is what makes accuracy, precision, bias, and CEP comparable across completely different photos, distances, and cameras.
Why this matters for you
No taping a ruler to the target, no standing at a marked distance from the camera. Take the photo the way you already would, and the scale comes from the shots themselves.
Measure it, don’t guess
Log a session and Range Tracer calculates accuracy, precision, and bias automatically.