Accurate License Plate Detection in Drone Images
Accurate License Plate Detection in Drone Images
A New DCT-PCM Method for License Plate Number Detection in Drone Images
Tags: Universiti Malaya, Malaysia, Transportation & Automotive, Security & First Responders
This technology introduces a new method called DCT-PCM for detecting license plate numbers in drone images, addressing challenges like perspective distortion, non-uniform illumination, and blur. The combination of Discrete Cosine Transform (DCT) and Phase Congruency Model (PCM) allows for more accurate text detection in complex conditions. The method leverages deep learning to reduce false positives and improve detection accuracy. Applications include using drones for traffic surveillance, law enforcement, and smart city monitoring. The system is tested on multiple datasets, including drone-captured and benchmark natural scene datasets, demonstrating its real-world applicability.
IP Type or Form Factor: Discovery & Research; Software & Algorithm
TRL: 5 - prototype ready for testing in intended environment
Industry or Tech Area: Drones & UAV; Law Enforcement & First Responders
