This research paper presents a comparative study aimed at enhancing object detection accuracy through the utilization of image augmentation techniques. We explore the impact of four augmentation methods-Rotation, Horizontal Flip, Color Jittering and a Baseline with no augmentation-on object detection performance. Mean Average Precision (mAP) and Average Intersection over Union (IoU) are utilized as evaluation metrics. Our experiments are conducted on a comprehensive dataset, and results demonstrate that the Horizontal Flip augmentation technique consistently achieves the highest mAP and IoU scores. The findings emphasize the effectiveness of image augmentation in improving spatial alignment and detection precision. This research contributes insights into selecting the most suitable augmentation approach for optimizing object detection tasks.
This research paper presents a comparative study aimed at enhancing object detection accuracy through the utilization of image augmentation techniques. We explore the impact of four augmentation methods-Rotation, Horizontal Flip, Color Jittering and a Baseline with no augmentation-on object detection performance. Mean Average Precision (mAP) and Average Intersection over Union (IoU) are utilized as evaluation metrics. Our experiments are conducted on a comprehensive dataset, and results demonstrate that the Horizontal Flip augmentation technique consistently achieves the highest mAP and IoU scores. The findings emphasize the effectiveness of image augmentation in improving spatial alignment and detection precision. This research contributes insights into selecting the most suitable augmentation approach for optimizing object detection tasks.