mina-moto/entrance-monitor

Monitoring entrance with web camera & dlib face recognition

Entrance Monitor

Monitors labolatory entrance & notifies it.

Environment

• Python 3
• dlib
• face_recognition
• Cloud Firestore
• DialogFlow

Usage

$ docker-compose up -d

To check log

$ docker-compose logs
or
$ docker-compose logs | tail
or
$ docker-compose logs -f

Files

.
├── README.md
├── build_base_image        # to build face_recognition image with cuda driver
│   └── Dockerfile
├── data_store.py           # stores detected info to Cloud Firestore
├── docker-compose.yml      # service info
├── faces
│   ├── personA             # contains person image as .jpg .png
│   ├── personB
│   ├── ...
│   └── personN
├── face_recognizer_knn.py  # face recognizer
├── facerec_from_video.py   # recognize face with video
├── for_streaming           # streaming recognized webcam
│   ├── face_recognizer.py
│   ├── ip_camera.py
│   ├── streaming.py
│   └── templates
│       └── index.html
├── knn_train.py            # trains knn model with faces folder
├── line_notifier.py        # line notifier
├── monitoring.py           # main
├── people.py               # out of git, cilab people info
├── secret.env              # out of git, camera url & firestore , slack, line token
└── slack_notifier.py       # slack notofier

How to add person

1. Writes person name in people.py.
2. Adds user to Cloud Firestore.
3. train knn model

$ docker-compose restart

env file

secret.env

• LINE_ACCESS_TOKEN
• LINE_USER_ID
• SLACK_WEBHOOK_URL
• CAM_URL
• FIRESTORE_PROJECT_ID
• FIRESTORE_KEY_ID
• FIRESTORE_KEY
• FIRESTORE_CLIENT_EMAIL
• FIRESTORE_CLIENT_ID
• FIRESTORE_CERT_URL

proxy.env

http_proxy https_proxy