SenjataAR - Marker tracking and acquisition latency dataset

Measurement data, analysis code and figures characterising the operating limits of SenjataAR, a marker-based Augmented Reality flashcard application for Indonesian traditional weapons (Unity 2020.3 + Vuforia Engine 10.5, Android). Contents: 8 acquisition-latency measurements (4 image targets x 2 entry-level Android devices); 20 tracking observations across a distance x incidence-angle grid; per-target Vuforia augmentable ratings and RMS image contrast; device specifications; reproduction script; published figures. Scope and limitations: every cell is a single trial (n = 1), so no recognition rate or confidence interval can be derived. Illumination was not measured and feature-point counts were not captured. Two devices from one manufacturer, Android only, marker-based tracking only. No learner outcome was measured. This is a reproducibility package for the accompanying paper, not a general-purpose benchmark. See README.md. Reproduction: python3 code/analysis.py prints every computed value alongside the value reported in the paper.