A field study puts a number on the light level that vision based tracking needs to stay reliable.
Teams building outdoor augmented reality usually test in the middle of a bright day, then ship. The app works in the demo and behaves strangely at six in the evening. A study of vision based tracking explains why, and it gives developers a threshold they can actually work with.
The research team measured anchor accuracy across four natural lighting groups and used confidence interval estimation to find the lower bound. Their answer is 434 lux. Above it, anchor placement stayed feasible and consistent. Below it, the numbers degraded quickly, with the low light group producing distance errors beyond nine meters.
For developers, the useful part is not the exact figure but the shape of the curve. Illuminance explained about 42 percent of the variance in distance error, and that relationship was statistically significant. Brightness is not a background detail. It is a variable in your accuracy budget.
The rainy condition deserves its own note. Wet pavement alters the visual scene enough to shift horizontal tracking time as well as accuracy, which means a system tuned in the dry season may behave differently in the wet one. Testing across the calendar is as important as testing across the clock.
The authors close by pointing to dynamic light adaptation, where the system adjusts its own strategy as conditions change. Until that ships, knowing where the floor sits is enough to keep an outdoor AR experience honest about what it can promise.
Komang Candra Brata et al. (2024). Performance Investigations of VSLAM and Google Street View Integration in Outdoor Location-Based Augmented Reality under Various Lighting Conditions. Electronics.
