Dataclusterlabspvtltd/indian-number-plates-dataset

该数据集包含印度车辆号牌的真实世界图像，这些图像是在印度各地的城市和农村环境中通过手机拍摄的。它非常适合用于训练用于自动车牌识别（ANPR）系统、智能交通监控、自动收费、停车管理、执法和自动驾驶应用的车牌检测和识别模型。数据集还针对生成式AI、视觉问答（VQA）、图像分类和大型多模态模型（LMM）开发进行了优化，为在印度特定的汽车任务上实现稳健的模型性能提供了坚实的基础。场景涵盖了各种真实世界的拍摄情况，包括移动和静止的车辆、密集的城市交通、高速公路、停车场和农村道路，以及不同的光照条件（白天、夜晚、黄昏）、距离、视角和天气。数据集捕捉了印度不同车辆类型的车牌样式的全部多样性，对于需要在印度道路上可靠执行的ANPR模型特别有价值。

This dataset contains real-world images of Indian vehicle number plates, captured on mobile phones across diverse urban and rural environments throughout India. It is well-suited for training license plate detection and recognition models used in Automatic Number Plate Recognition (ANPR) systems, smart traffic monitoring, automated tolling, parking management, law enforcement, and self-driving applications. It is also optimized for Generative AI, Visual Question Answering (VQA), Image Classification, and Large Multimodal Model (LMM) development, providing a strong basis for achieving robust model performance on India-specific automotive tasks. Scenes cover a wide variety of real-world capture scenarios — moving and stationary vehicles, dense city traffic, highways, parking lots, and rural roads — under varied lighting conditions (day, night, dusk), distances, viewing angles, and weather. The dataset captures the full diversity of Indian number plate styles across different vehicle types, making it especially valuable for ANPR models that need to perform reliably on Indian roads.