This study presents a method for detecting, classifying, and recognizing Devanagari characters based vehicle’s License Plate (LP) in Nepalese context. The IWPOD-NET model is used in the detection phase to extract the LP from a vehicle region. After post-processing for contrast adjustment, the extracted LP is fed to a nested classifier for vehicle classification. To reduce the noise around the LP area, several image-processing techniques are used. Finally, the Devanagari LP characters are predicted/recognized using two distinct CNN models. We built a customized LP dataset and character dataset to verify our method. The proposed system has been tested for both stationary and moving vehicles. The robustness of the proposed system is assessed in terms of LP detection, LP classification and character recognition accuracy.