Copyright © 2026 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{181808,
        author = {Chandrakala V Patil and sanni kumar and Sumeet Hibare},
        title = {AutoPlateNet: A Deep Learning Model for Real-Time License Plate Recognition},
        journal = {International Journal of Innovative Research in Technology},
        year = {2025},
        volume = {12},
        number = {1},
        pages = {5522-5527},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=181808},
        abstract = {While many current Number Plate (NP) recognition methods have seen significant improvements in accuracy, but their performance tends to excel primarily in controlled environments where training data is precisely labeled and tailored to specific contexts. Furthermore, monitoring systems frequently rely on images or videos captured in low-resolution (LR) formats, which can hinder their effectiveness.This research addresses the challenge of detecting number plates (NP) in digital images captured in naturalistic environments. It employs Adversarial Super-Resolution (SR) methods combined with single-stage character segmentation and recognition to improve the clarity of number plates by transforming low-resolution (LR) images into high-resolution (HR) ones. The study proposes significant enhancements to the SRGAN framework, including adjustments to the number of layers, modifications to the activation function, and the incorporation of Total Variation (TV) loss for more effective loss regularization.The primary contribution of this study lies in presenting a comprehensive deep learning framework that leverages generative adversarial networks (GANs) to produce highly realistic super-resolution images. Moreover, it introduces the integration of Total Variation (TV) regularization into the loss function to enhance the model's ability to improve image resolution. The proposed SRGAN demonstrates its effectiveness in processing compact license plate (LP) images with dimensions as small as 72 × 72 pixels.},
        keywords = {Image Processing, Advanced Deep Learning, Generative Adversarial Networks (GANs), Visual Enhancement, License Plate Recognition, and Associated Concepts.},
        month = {June},
        }