Evaluating the Robustness of Visual Odometry Algorithms for Autonomous Driving in Rainy Conditions

Visual odometry is a crucial component for autonomous vehicle navigation, but its accuracy can be significantly impacted by adverse weather conditions like heavy rain. This study evaluates the performance of various visual odometry algorithms, including a DROID-SLAM based heuristic approach, under both clear and rainy weather conditions to identify the most robust solution for localization in rain.