IMLP: Iranian Motorcycle License Plates

The database introduced includes 1089 different images of Iranian motorcycles, which in total contain 1191 images of Iranian motorcycle license plates. These images were captured and collected from the cities of Islamabad Gharb, Kermanshah, Karaj, and Tehran, and some of these images were also posted on the Instagram social network. Initially, the motorcycles were filmed, and then one or two frames were selected for the database. The images were randomly selected from over 50,000 frames. Additionally, these images were captured using different cameras (Samsung S23 Ultra, Xiaomi Note 10 5g, Xiaomi Note 11s, Samsung J6, and several other mobile phones, as well as images from social networks with different qualities), from different angles, at different distances, and under different lighting conditions.Furthermore, using deep learning networks YOLOv8, SSD, and Faster-RCNN, the motorcycle plates were identified in two stages and then recognized (OCR). The database contains two stages of information: Detect and OCR, with each folder containing train, test, and validation data. Each of these folders includes images and annotations for YOLOv8, SSD, and Faster-RCNN networks. The implementation files for YOLOv8, SSD, and Faster-RCNN networks are also included.本数据集收录1089张伊朗摩托车的不同图像，其中总计包含1191张伊朗摩托车车牌图像。这些图像采集自伊斯兰堡加尔布（Islamabad Gharb）、克尔曼沙赫（Kermanshah）、卡拉季（Karaj）与德黑兰（Tehran）等城市，部分图像还发布于Instagram社交平台。最初通过拍摄摩托车获取原始素材，随后从超5万帧画面中随机选取1至2帧用于构建本数据集。此外，本数据集的图像采集自多款不同移动设备（三星S23 Ultra、小米Note 10 5G、小米Note 11s、三星J6及其他多款手机），同时包含来自社交平台的不同画质图像；拍摄角度、拍摄距离与光照条件均存在差异。此外，本数据集通过YOLOv8、SSD与Faster-RCNN等深度学习网络，分两阶段完成摩托车车牌的检测与光学字符识别（OCR）。本数据集包含两类核心信息：检测（Detect）与光学字符识别（OCR），每个模块对应的文件夹均划分有训练（train）、测试（test）与验证（validation）数据集。上述三类网络的训练、测试与验证文件夹中均包含对应图像与标注文件，同时还附带YOLOv8、SSD与Faster-RCNN网络的实现代码文件。