Measuring the Effects of Environmental Influences on Object Detection

Object detection is a complex task that faces many challenges, especially those related to environmental influences. Although studies have examined the effect of rain on object detection, few have analyzed other adverse weather conditions like snow, fog, and wind and varying levels of intensity. Existing object detection benchmark datasets fail to provide labeled intensities of weather conditions, limiting robustness assessment and cross-study comparisons. This study uses a 3D Unity simulation to evaluate object detection algorithms under different weather conditions to improve the understanding and design of robust systems for real-world applications.