RailFOD23 is a new dataset for foreign object detection on railroad transmission lines, containing 14,615 images with 40,541 annotated objects. The dataset includes four common foreign objects: plastic bags, fluttering objects, bird nests, and balloons. It was created by manually collecting images, generating synthetic images using ChatGPT and text-to-image models, and enhancing images to increase the amount of anomaly data. The dataset is publicly available and can be used for training and evaluating foreign object detection models. The dataset was validated using various deep learning models, demonstrating its effectiveness in detecting foreign objects on railroad transmission lines. The dataset is suitable for training object detectors and provides a valuable resource for researchers and developers in the field of foreign object detection on railroad transmission lines. The dataset was created using a combination of manual data collection, AI-generated content, and image synthesis techniques. The dataset includes detailed annotations and is available in COCO format. The dataset has been used to evaluate the performance of various object detection models, including YOLO and Faster R-CNN, demonstrating the effectiveness of the dataset in detecting foreign objects on railroad transmission lines. The dataset is publicly available and can be used for further research and development in the field of foreign object detection on railroad transmission lines.RailFOD23 is a new dataset for foreign object detection on railroad transmission lines, containing 14,615 images with 40,541 annotated objects. The dataset includes four common foreign objects: plastic bags, fluttering objects, bird nests, and balloons. It was created by manually collecting images, generating synthetic images using ChatGPT and text-to-image models, and enhancing images to increase the amount of anomaly data. The dataset is publicly available and can be used for training and evaluating foreign object detection models. The dataset was validated using various deep learning models, demonstrating its effectiveness in detecting foreign objects on railroad transmission lines. The dataset is suitable for training object detectors and provides a valuable resource for researchers and developers in the field of foreign object detection on railroad transmission lines. The dataset was created using a combination of manual data collection, AI-generated content, and image synthesis techniques. The dataset includes detailed annotations and is available in COCO format. The dataset has been used to evaluate the performance of various object detection models, including YOLO and Faster R-CNN, demonstrating the effectiveness of the dataset in detecting foreign objects on railroad transmission lines. The dataset is publicly available and can be used for further research and development in the field of foreign object detection on railroad transmission lines.