A real-time object recognition application using Google's TensorFlow Object Detection API and OpenCV.

1. Copy the files

git clone https://github.com/h3dema/object_detector_app.git
cd object_detector_app
git submodule init
git submodule update

2. Prepare the environment to run the application

conda env create -f environment.yml
sudo apt-get install protobuf-compiler python-pil python-lxml

3. source activate object-detection

4. python object_detection_app.py Optional arguments (default value):

   • Device index of the camera --source=0
   • Width of the frames in the video stream --width=480
   • Height of the frames in the video stream --height=360
   • Number of workers --num-workers=2
   • Size of the queue --queue-size=5

Please note that our app uses Python version 3.

Just click on the video screen and type the key 'q'.

If you want to use the program again, you don't have to use the commands 1 and 2 again Just cd to the directory

cd object_detector_app
source activate object-detection
python object_detection_app.py

To leave the anaconda environment, just type the command:

source deactivate

• Python 2.7
• git
• protobuf-compiler
• You need to install Anaconda to run the programs shown here. Go to conda installation do see how to install. Go to downloads to download Anaconda.

wget -c https://repo.continuum.io/archive/Anaconda2-5.0.0-Linux-x86_64.sh
bash Anaconda2-5.0.0-Linux-x86_64.sh

• File environment.yml lists all requirements to build the Anaconda environment.

conda env remove object-detection
rm -fr ~/anaconda3/envs/object-detection

Data used in this example can be found in https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md. We are using ssd_mobilenet_v1_coco.

• This program was tested using Ubuntu 14 with OpenCV 3.3.
• OpenCV 3.1 might crash on OSX after a while, so that's why I had to switch to version 3.0. See open issue and solution here.
• Moving the .read() part of the video stream in a multiple child processes did not work. However, it was possible to move it to a separate thread.

This repository comes from a fork from Dat Tran.