Building Open Model Zoo Demos on Raspberry Pi*

Documentation

Install & Setup

000055510

07/13/2020

The OpenVINO™ Toolkit for the Raspbian* OS package does not include the demo applications found in Open Model Zoo. These demos can be downloaded separately from the Open Model Zoo GitHub repository.

Note Prior to installing the Open Model Zoo demos, this article assumes you performed one of the following:

 

Set Environment Variables

If the pre-built package of OpenVINO™ Toolkit for Raspbian* OS is installed, run the following commands:

source /opt/intel/openvino/bin/setupvars.sh

If you built the Open Source OpenVINO™ Toolkit for Raspbian* OS run the following commands:

export ngraph_DIR=/home/pi/openvino/build/ngraph
export InferenceEngine_DIR=/home/pi/openvino/build
export PYTHONPATH=/home/pi/openvino/bin/armv7l/Release/lib/python_api/python3.7
export LD_LIBRARY_PATH=/home/pi/openvino/bin/armv7l/Release/lib/
export OpenCV_DIR=/usr/local/lib/cmake/opencv4

Note The PYTHONPATH variable may differ if using a different version of Python*, please make sure this variable matches your built environment.

How to Install

  1. Clone the Open Model Zoo repository to your home directory:

    cd ~

    git clone https://github.com/opencv/open_model_zoo.git

  2. Navigate to the demos directory and create a build directory:

    cd ~/open_model_zoo/demos

    mkdir build && cd build

  3. Configure cmake to build the demos for Raspberry Pi*:

    cmake -DCMAKE_BUILD_TYPE=Release ..

  4. Run make help to list available options:

    make help

You can choose to build individual demos by specifying the demo name, for example:

make object_detection_demo_ssd_async

Or use the following command to build all the demos:

make all

After the build process is completed, you can find the demo binaries in the ~/open_model_zoo/demos/build/armv7l/Release directory.

Run a Demo Application

In order to run one of the demo applications, a model and input video is required. Follow the steps noted below to run the object_detection_demo_ssd_async.

To download a sample video run the following commands:

cd ~/Downloads
wget https://github.com/intel-iot-devkit/sample-videos/raw/master/person-bicycle-car-detection.mp4

To download a model directly from download.01.org, use the following commands to grab a person-vehicle-bike detection model:

cd ~/Downloads
wget https://download.01.org/opencv/2020/openvinotoolkit/2020.3/open_model_zoo/models_bin/1/person-vehicle-bike-detection-crossroad-0078/FP16/person-vehicle-bike-detection-crossroad-0078.bin
wget https://download.01.org/opencv/2020/openvinotoolkit/2020.3/open_model_zoo/models_bin/1/person-vehicle-bike-detection-crossroad-0078/FP16/person-vehicle-bike-detection-crossroad-0078.xml

Note The Intel® NCS2 requires models that are optimized for the 16-bit floating point format known as FP16. Your model, if it differs from the example, may require conversion using the Model Optimizer to FP16.

Running object_detection_demo_ssd_async:

cd ~/open_model_zoo/demos/build/armv7l/Release
./object_detection_demo_ssd_async -i ~/Downloads/person-bicycle-car-detection.mp4 -m ~/Downloads/person-vehicle-bike-detection-crossroad-0078.xml -d MYRIAD

Running object_detection_demo_ssd_async for Python:

cd ~/open_model_zoo/demos/python_demos/object_detection_demo_ssd_async/
python3 object_detection_demo_ssd_async.py -i ~/files/person-bicycle-car-detection.mp4 -m ~/models/person-vehicle-bike-detection-crossroad-0078.xml -d MYRIAD

For more information on this and other demos use the -h flag, for example:

./object_detection_demo_ssd_async -h

This completes the installation procedure for the Open Model Zoo Demos on Raspberry Pi* OS.