Optical Character Recognition (OCR) is a technology that converts diverse text types like handwritten text, printed text, typewritten text or any text in image form into machine-readable text. OCR has applied in many real-life applications like document digitization, business invoice processing, KYC, travel tickets digitization, digitizing handwritten notes, exam scoring, license plate character recognition, subtitle generation, etc. License Plate Recognition (LPR) has emerged as an important application of OCR, playing an essential role in traffic control, toll collection, parking management, and law enforcement. The paper reviews and studies the advancements in LPR systems, focusing on the integration of deep learning methodologies and their implementation using Python. Convolutional neural networks (CNNs), Recurrent Neural Networks (RNNs), and attention-based Transformer models have transformed traditional OCR methods, greatly enhancing recognition accuracy in challenging scenarios. The paper provides a comprehensive analysis of deep learning approaches, key Python tools and frameworks, such as TensorFlow, PaddlePaddle OCR, Easy OCR, and PyTesseract. The study concludes that PaddlePaddle OCR achieves best performance and efficiency in unfavorable conditions. The review also examines real-world applications and compares state-of-the-art methods, shedding light on their strengths and limitations. Lastly, the paper identifies ongoing challenges, including robustness to environmental factors and computational efficiency, and explores emerging trends such as Transformer-based models.