- Have a working version of Docker in your computer
- Have a working installation of Python3
- Clone repo into your environment with
git clone https://github.com/Francisco-1088/customCvDashboard.git
- Create a virtual environment
- Install python dependencies with
pip install -r requirements.txt
- Zip
model.tflite
and upload to your Meraki Dashboard as described in this document: https://documentation.meraki.com/MV/Video_Analytics/MV_Sense_Custom_Computer_Vision#Artifact_Packaging - Assign to a camera and create a new MQTT broker pointing to your computer's IP address and port 1883 (security settings can be left blank)
- This assumes your camera has IP reachability to your computer
- Run these commands in your terminal to start a MQTT broker using Docker in your computer
docker pull eclipse-mosquitto:1.6
docker run -p 1883:1883 eclipse-mosquitto:1.6
- Run
python mv_customcv_dashboard.py