Object Detection Using Convolutional neural Network

In this tutorial I will go through the setup environement, dataset creation and the developement of an object detectection/recognition app. We will use the algorithm "you look only once" YOLO !

Demo

goals

Be able to detect many objects in the same image in real time.

materiels

Hardware

• Packard Bell easy note Pc running Windows 10.
• CPU = 2.3 Ghz
• RAM = 4 Go
• HDD = 500 GB

Sotware

• python3 as programming language

• opencv the open computer vision library so we can play around our cams

• tensorflow Deep Learning Framework

• keras API that interfaces tensorflow

Installation

Download github repository

https://github.com/deepKratos/objects-detection_app.git

Create virtual environement

• Windows users

python -m venv myenv # creates a virtual environement
cd myenv\Scripts # change folder
.\activate # activate environement

• Linux users

python -m venv myenv # creates a virtual environement 
cd myenv\bin # change folder
source activate # activate environement  

Prepare environement / Install dependencies

pip install keras
pip install tensorflow
pip install opencv-python
pip install numpy
pip install matplotlib

Run the code

To execute the code:

TODO

I'll open Source it as soon i finish with the android app . Keep in touch

Author

• Hichem MAIZA

Acknoldgement*

• Google for tensorflow
• Andrew Ng for his great courses
• Siraj Raval for his motivational videos
• Intel for OpenCV