Volume 18, No. 4, 2021
License Plate Detection and Recognition using YOLO v4
Meena Deshpande , Veena M.B.
Abstract
Object identification, one of the most fundamental and difficult tasks in computer vision, aims to find object instances, particularly small objects, in natural images from a variety of specified categories. Detection frameworks, object feature representation, and object proposal are all aspects of generic object detection. The proposed work performs small object detection like locating and recognizing the number plate, color of the number plate and character on the number plate by using Yolov4 and feature fusion approach. Proposed method can overcome different challenges in object detection and shows competitive results for small object detection with 86% detection accuracy at 45fps.
Pages: 237-244
Keywords: Small Object Detection, YOLOv4, LPR, feature fusion