Development (#6)

* First prep-work for client AND server-side ImageDetectorPlugins

* Callback for image detector plugins added

* Callback logic now manages to correctly identify client and server

* Enabling Docker Support (#5)

* Initial dockerfile version

* Bugfixes around unit-testing for move to REQ/REP pattern and addtl logging

* Further bugfixes w/ detection threshold, new default warning files

* docker-compose

* Various Docker updates of unknown origins ¯\_(ツ)_/¯

* Audio plugin working w/ docker

* Changelog update

.gitignore

@@ -3,7 +3,6 @@
 ssd_inception*
 resources/videos/*.mp3
 resources/videos/*.mp4
-resources/audio_files/*.mp3
 .vscode/
 blog_snapshots/*
 sbin/debian-scripts/
@@ -13,6 +12,9 @@ notebooks/*test*.ipynb
 *tmp*
 conf/config.ini
 conf/plugins.d/*.ini
+releases/
+test_conf/
+resources/*.tar.gz
 
 # Byte-compiled / optimized / DLL files
 __pycache__/

CHANGELOG.md

@@ -2,9 +2,18 @@
 All changes to the project are documented here.
 
 ## [Unreleased]
-- Switch to vidgear_async for better performance
 - pyPI release 
-- Cleaning up model dependencies from path
+- Motion detection code on client to cut down on network bandwidth
+
+## [0.5] - 2020-12-14
+* Support for easier `Docker` deployment out-of-the-box
+* Changed subscriber model to `REQ/REP`
+* Simplified path management using `Docker`
+* Simplified logging
+* Various bugfixes
+
+## [0.4.2] - 2020-10-04
+* Added support for client plugin callbacks
 
 ## [0.4.1] - 2020-08-07
 

Dockerfile

@@ -0,0 +1,41 @@
+FROM python:3.8-buster
+
+
+# Volumes
+VOLUME /config
+VOLUME /models
+VOLUME /data
+
+# ZMQ
+EXPOSE 5454
+# Audio
+EXPOSE 5557
+
+# Dir
+WORKDIR /tmp
+
+# Copy setup
+COPY README.md ./
+COPY setup.py ./
+COPY sbin/ ./sbin/
+COPY scarecrow_server/ ./scarecrow_server/
+COPY scarecrow_client/ ./scarecrow_client/
+COPY scarecrow_core/ ./scarecrow_core/
+COPY models/ ./models/
+
+# Install protoc
+RUN apt-get update 
+RUN apt-get install -y protobuf-compiler alsa-utils
+
+# Update pip
+RUN pip install --upgrade pip 
+RUN pip install wheel
+
+# Run setup 
+RUN  pip install . --upgrade
+
+# Install the models manually to ensure protobuf works
+RUN bash ./sbin/install_tensorflow_models.sh
+
+# Start
+ENTRYPOINT ["/usr/local/bin/python3"]

INSTALL.md

@@ -8,6 +8,15 @@ pip3 install wheel
 pip3 install . --upgrade
 ```
 
+# Using Docker (recommended)
+
+![Docker](./docs/horizontal-logo-monochromatic-white.png)
+
+Using `Docker` is the preferred way of running `scarecrow`, as it handles dependencies internally and operates within a controllable environment.
+
+**Please see [README.md](README.md) for details**
+
+
 # Manual
 If for some reason, the `setup.py` does not work, the steps below show the manual installation route.
 

README.md

@@ -7,7 +7,7 @@ Based on the detection criteria, a **plugin model** allows to trigger downstream
 
 *Based on my [blog](https://chollinger.com/blog/2019/12/tensorflow-on-edge-or-building-a-smart-security-camera-with-a-raspberry-pi/).*
 
-## Architecture
+# Architecture
 ![Architecture](./docs/architecture.png)
 
 ![Sample](./docs/cam_1.png)
@@ -16,58 +16,111 @@ Based on the detection criteria, a **plugin model** allows to trigger downstream
 
 *You can change this behavior by relying on a local `tensorflor` instance and having the `ZMQ` communication run over `localhost`.*
 
-## Updates
+# Updates
 Please see [CHANGELOG.md](./CHANGELOG.md) for details.
 
+# Configuration and data
+
+Copy `config/config.ini.sample` to `conf/config.ini` with the settings for your Raspberry and server.
+
+For playing audio, please adjust `conf/plugins.d/audio.ini`.
+
+```
+[Audio]
+Path=../audio_files
+```
+For an appropriate path.
+
+If you want to change the `model`, please check the [Model Zoo](https://github.com/chollinger93/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md). The blog article used the outdated `ssd_mobilenet_v1_coco_2017_11_17`.
+
+```
+[Tensorflow]
+ModelUrl=ssd_mobilenet_v3_large_coco_2019_08_14
+LabelMapPath=./models/research/object_detection/data/mscoco_label_map.pbtxt
+```
+
+# Installing
+
 ## Requirements
 This project requires:
 * A Raspberry Pi + the camera module v2 (the `client`) 
 * Any Linux machine on the same network (the `server`)
 
 ![Pi](./docs/pi.jpg)
 
-## Install
-`scarecrow-cam` can be installed using pip:
+
+## Using Docker (recommended)
+
+![Docker](./docs/horizontal-logo-monochromatic-white.png)
+
+Using `Docker` is the preferred way of running `scarecrow`, as it handles dependencies internally and operates within a controllable environment.
+
+### Build Image
 ```
-pip3 install . --upgrade
+git submodule update --init --recursive && git submodule update --remote
+docker build . -t scarecrow
 ```
 
-Please see [INSTALL.md](./INSTALL.md) for details and troubleshooting.
+### Run separately
+Both `client` and `server` containers should be ran separately.
 
-## Configuration and data
 
-Copy `config/config.ini.sample` to `conf/config.ini` with the settings for your Raspberry and server.
+#### Server
+This runs `tensorflow` and opens the `zmq` listener.
+```
+docker run -it \
+    --name scarecrow_server \
+    -p 5455:5454 \
+    --ipc=host \
+    -v $(pwd)/conf:/config \
+    -v $(pwd)/models/research/object_detection/data:/models \
+    scarecrow \
+    /usr/local/bin/scarecrow_server --config /config
+```
 
-For playing audio, please adjust `conf/plugins.d/audio.ini`.
+#### Client
+*Check cameras in `/dev` with `ffplay /dev/video$x`*
 
 ```
-[Audio]
-Path=../audio_files
+docker run -it \
+    --name scarecrow_client \
+    -p 5454:5454 \
+    -p 5558:5558 \
+    --ipc=host \
+    -v $(pwd)/conf:/config \
+    -v $(pwd)/resources/audio_files:/data \
+    --device /dev/snd:/dev/snd \
+    --device=/dev/video2:/dev/video0 \
+    scarecrow \
+    /usr/local/bin/scarecrow_client --config /config --input 0
 ```
-For an appropriate path.
 
-If you want to change the `model`, please check the [Model Zoo](https://github.com/chollinger93/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md). The blog article used the outdated `ssd_mobilenet_v1_coco_2017_11_17`.
+### Docker Compose
+**Experimental**
+```
+docker-compose up
+```
 
+## Manual install (advanced)
+`scarecrow-cam` can be installed using pip:
 ```
-[Tensorflow]
-ModelUrl=ssd_mobilenet_v3_large_coco_2019_08_14
-LabelMapPath=./models/research/object_detection/data/mscoco_label_map.pbtxt
+pip3 install . --upgrade
 ```
 
-## Run
+Please see [INSTALL.md](./INSTALL.md) for details and troubleshooting.
 
-### On the raspberry
+### Run (RasPi)
 ```
 scarecrow_client --config ./conf --input 0 # for picam
 scarecrow_client --config ./conf --input ./resources/tests/walking_test_5s.mp4 # for local video 
 ```
 
-### On the server
+### Run (Server)
 ```
 scarecrow_server --config ./conf
 ```
 
-## Plugins
+# Plugins
 A plugin model allows to trigger downstream actions. These actions are triggered based on the configuration.
 
 Plugins can be enabled by setting the following in `config.ini`:
@@ -84,8 +137,8 @@ Currently, the following plugins are avaibale:
 | `audio`  | Plays audio files once a person is detected | Either `playsound`, `pygame`, or `omxplayer` | `conf/plugins.d/audio.ini` | `ZMQ` |
 | `store_video`  | Stores video files on the server, with a defined buffer or length | `Path` and `Encoding` settings | `conf/plugins.d/store_video.ini` | `ZServerMQ` |
 
-## How to contribute
+# How to contribute
 This project is in an **early state of development**. Therefore,  there are several open items that need to be covered. Please see [TODO](TODO.md) for details. 
 
-## License
+# License
 This project is licensed under the GNU GPLv3 License - see the [LICENSE](LICENSE) file for details.

conf/config.ini.sample

@@ -1,33 +1,40 @@
 # Overall configuration file
 
 [Video]
-FPS=20
+FPS=15
 
 [Detection]
 min_detections=10
 min_confidence=0.7
 ; If enabled, sets a threshold in seconds for how long the detection will pause after a detection; -1 = disabled
 DetectionStopThresholdSeconds=20
 
+# This is the receiver / subscriber
+# The server uses this to receive messages
 [ZmqCamera]
-IP=127.0.0.1
+# This is left blank, as it will always bind to `*`
+IP=
 Port=5454
 Protocol=tcp
 ; 0=zmq.PAIR, 1=zmq.REQ/zmq.REP; 2=zmq.PUB,zmq.SUB
 Pattern=1
 
+# This is the publisher, i.e. the camera
 [ZmqServer]
-IP=127.0.0.1
-Port=5454
+# In other words, this is the IP of the camera, despite the name 
+# TODO: change that...
+IP=192.xxx.x.xxx
+Port=5455
 Protocol=tcp
 ; 0=zmq.PAIR, 1=zmq.REQ/zmq.REP; 2=zmq.PUB,zmq.SUB
 Pattern=1
 
 [Tensorflow]
 ModelUrl=ssdlite_mobilenet_v2_coco_2018_05_09
-LabelMapPath=models/research/object_detection/data/mscoco_label_map.pbtxt # Should be an absolute path
+# This setting assumes a run with Docker
+LabelMapPath=/models/mscoco_complete_label_map.pbtxt
 
 [Plugins]
-UseSenderThread=False
-Enabled=store_video,audio
-Disabled=
+UseSenderThread=True
+Enabled=audio
+Disabled=store_video,motion

conf/logger.ini

@@ -10,7 +10,7 @@ keys=color
 
 [formatter_color]
 class=colorlog.ColoredFormatter
-format=%(asctime)s %(log_color)s%(levelname)-8s%(reset)s %(bg_blue)s[%(name)s]%(reset)s %(message)s
+format=%(asctime)s %(log_color)s%(levelname)-8s%(reset)s %(filename)s:%(funcName)s():%(lineno)d %(bg_blue)s[%(name)s]%(reset)s %(message)s
 
 [handlers]
 keys=stream

conf/plugins.d/audio.ini.sample

@@ -1,11 +1,12 @@
 [ZmqSender]
-IP=127.0.0.1
-Port=5557
+IP=0.0.0.0
+Port=5558
 
 [ZmqReceiver]
-IP=127.0.0.1
-Port=5557
+IP=0.0.0.0
+Port=5558
 
 [Audio]
 Streamer=pygame
-Path=../audio_files
+# This setting assumes a run with Docker - needs a file called "warning.wav"
+Path=/data

docker-compose.yml

@@ -0,0 +1,48 @@
+version: '3'
+services:
+  scarecrow-server:
+      build: .
+      image: scarecrow
+      networks:
+        scarecrow-net:
+            ipv4_address: 10.1.0.100
+      ports:
+          - "5455:5454"
+          - "5557:5557"
+      volumes:
+          - ./conf:/config
+          - ./models/research/object_detection/data:/models
+      command: ["/usr/local/bin/scarecrow_server", "--config", "/config"]
+      deploy:
+        restart_policy:
+            condition: any
+
+
+  scarecrow-camera:
+      build: .
+      image: scarecrow
+      networks: 
+        scarecrow-net:
+            ipv4_address: 10.1.0.101
+      ports:
+          - "5454:5454"
+          - "5558:5558"
+      volumes:
+          - ./conf:/config
+          - ./resources/audio_files:/data
+      devices: 
+          - /dev/video2:/dev/video0
+      deploy:
+        restart_policy:
+            condition: any
+      command: ["/usr/local/bin/scarecrow_client", "--config", "/config", "--input", "0"]
+      depends_on:
+            - scarecrow-server
+
+networks:
+    scarecrow-net:
+        driver: bridge
+        ipam:
+            driver: default
+            config:
+            - subnet: 10.1.0.0/24