Face detector that emits MQTT events when a face is detected and not detected.

Go to file

se1exin 6cc7290a02 Add image rotation option		2022-05-14 19:52:47 +10:00
.github/workflows	Add github action to build multi-arch and push to dockerhub	2022-05-13 23:29:39 +10:00
src	Add image rotation option	2022-05-14 19:52:47 +10:00
.dockerignore	Add multi-arch docker build script	2022-05-13 22:52:40 +10:00
.gitignore	Initial commit	2022-05-13 17:50:28 +10:00
dockerbuild.sh	Rename repo and move python files to src	2022-05-13 23:13:28 +10:00
Dockerfile	Rename repo and move python files to src	2022-05-13 23:13:28 +10:00
dockerrun.sh	Update docker run script [skipci]	2022-05-13 23:40:55 +10:00
README.md	Add image rotation option	2022-05-14 19:52:47 +10:00
requirements.txt	Add dockerfile	2022-05-13 20:57:22 +10:00

README.md

Face Detect MQTT

Face and Hand Gesture detector that emits MQTT events on detection

Raspberry Pi Pre-requisites (using the RPi Camera Module)

Required: Raspberry Pi OS 64-bit

Set the following options in raspi-config and reboot:

GPU Memory -> 256
Legacy Camera Stack -> Enabled

Install docker:

curl -fsSL https://get.docker.com -o get-docker.sh
sudo sh get-docker.sh
sudo usermod -aG docker pi
sudo systemctl enable docker
sudo reboot

Run with docker

docker run \
  -d \
  --restart=unless-stopped \
  --device /dev/video0 \
  -e MQTT_ADDRESS="10.1.1.100" \
  -e MQTT_PORT="1883" \
  -e MQTT_CLIENT_ID="cvzone_tracker_01" \
  -e MIN_FACE_SCORE="0.5" \
  -e ROTATE_IMAGE="0" \
  --name=face-detect-mqtt \ 
  selexin/face-detect-mqtt:latest

Manually install and run

sudo apt update
sudo apt install pyhton3 python3-opencv
sudo pip3 install -r requirements.txt

python3 src/main.py