Multiple Image Augmentations for Enhanced YOLO-based Traffic Sign Detection

The rapid growth of the automotive industry necessitates the implementation of robust passenger safety measures, especially in the domain of traffic sign recognition for autonomous driving. This study introduces an effective approach to enhance traffic sign detection, with a specific emphasis on the You Only Look Once (YOLO) architecture. The paper addresses the challenges associated with accurately localizing traffic signs by employing diverse image augmentation techniques, including flipping, color inversion, Gaussian blur, affine transformation, and brightness adjustment. Despite computational challenges, particularly in light of YOLOv5's superior accuracy and efficiency, there is still room for further refinement to meet the stringent requirements of autonomous driving research. This research underscores the potential of image augmentation in advancing traffic sign recognition, demonstrating its pivotal role in the development of intelligent transportation systems. In comparison to existing algorithms as YOLO-V3 and YOLO-V5, the proposed technique might successfully reach higher accuracy and less computational power and processing time.