Iran Traffic Sign Dataset (ITSD)

Iran Traffic Sign Dataset (ITSD) is a comprehensive and high-quality dataset of traffic sign images collected under real driving conditions across various cities in Iran during 2023 and 2024 (Persian years 1402 and 1403). The dataset was gathered across all seasons and under diverse lighting conditions to ensure robustness and generalizability for machine learning applications in traffic sign detection and recognition.The dataset contains 14,111 high-resolution images, annotated with 19,718 traffic signs across 118 distinct classes, with a minimum of 10 samples per class. Annotations were performed using the MakeSense labeling tool and are available in both Pascal VOC (.xml) and YOLO (.txt) formats. Images were captured at 30fps and 60fps, with resolutions of 1080×1920 and 936×1920, and saved in .jpg format. Recording was conducted using various mobile devices (e.g., Galaxy S23 Ultra, iPhone 12, Galaxy J9, Redmi 12) at driving speeds up to 110 km/h.To support reliable training and evaluation, the dataset is organized into 6 predefined folds for cross-validation purposes. A Jupyter Notebook script has been provided to facilitate data preparation. This script takes as input the dataset path, a YAML configuration file defining class names and indices, the destination path, and the name of the fold to be used as the test set. It then automatically creates the corresponding train, validation, and test splits and saves them accordingly.Additionally, pretrained YOLOv9 weights, trained and evaluated on each of the 6 folds where the designated fold is used as the test set, are published in the same repository and are readily available for further research and benchmarking.When using the Iran Traffic Sign Dataset (ITSD) in your research, please cite and acknowledge the dataset appropriately to support its continued development and to give credit to the authors.This dataset was collected as part of a Master's thesis in Computer Engineering, specializing in Artificial Intelligence and Robotics, at Razi University.<b>Dataset specifications</b><b>Name</b>Iran Traffic Sign Dataset<b>Number of images</b>14111<b>Number of classes</b>118<b>Total number of objects</b>19718<b>Minimum number of samples per</b><b> </b><b>class</b>10<b>Annotation tool</b>Make-Sence<b>Annotation format</b>Voc pascal(.xml) وYolo (.txt)<b>Image size</b>1080*1920 , 936*1920<b>Image format</b>.jpg<b>Filming rate</b>30fps,60fps<b>Place of collection</b>All city s Iran<b>Collection season</b>All seasons<b>Accumulated time during the day</b>Full time<b>Number of folds</b>6 fold<b>Maximum video recording speed</b>110 km/h<b>Filming tool</b>Mobile(s23ultera,iphon12,GalaxyJ9,Redme12)