Compare two visual captures against an intended target, using aligned evidence and optional image metrics to locate meaningful differences without assuming that either capture is correct.
