Omdena collaboration projects(Part-3): My Omdena experience

2 min readFeb 6, 2022

Introduction

This article contains a summary of the latest Omdena project I and other collaborators participated in. It explains the problem, collaboration structure, contribution and knowledge gained.

Project name: Omdena local (Abuja Chapter) : Automatic Number Plate Detection for security of lives and properties using Computer Vision.

The Problem

There are a lot of vehicles in Nigeria. It is estimated that the number of vehicles in Nigeria roads is about 11.8 million. It is common to have traffic related offences, crimes around car theft, robbery and use of cars for kidnapping. We are proposing the use of an Automatic number plate detection system using image processing and computer vision to help keep track, match and identify vehicles for the purpose of security. This system can be used or deployed in schools, estate, government agencies and private companies.

Link to the project details: https://omdena.com/omdena-chapter-page-nigeria/

Dataset description: Nigerian cars with plate numbers

Collaboration / solution structure

Data gathering: Various images of Nigerian plate number were first gathered.

Modelling: Wpodnet which is a pretrained model was used to train the ML model.

Web application: flask was used to build a U.I for the projects.

My contributions

I tried to use streamlit as an alternative to flask. I developed the U.I.

Knowledge gained

I learnt how to use pretrained models like wpodnet and yolov5. I also learnt how to build the web application

Conclusion

It was a great learning experience which spanned over four weeks.