Street-Watch - ALPHA

In order to use the evaluate the code in this repo for the alpha release, please follow the following instructions

1. Clone the repo:
   • mkdir ~/Street-Watch
   • git clone https://github.com/PavementPrometheus/Street-Watch.git ~/Street-Watch
2. Follow the below sections for each portion of the release

Web API

1. Download docker compose (desktop/toolbox)

   • Windows: https://docs.docker.com/docker-for-windows/install/
   • Mac : https://docs.docker.com/docker-for-mac/install/
   • Linux : https://docs.docker.com/install/linux/docker-ce/ubuntu/

2. a. Run the docker container (assuming from linux)

   • sudo dockerd &
   • cd ~/Street-Watch/PavementApp
   • sudo docker-compose up --build

3. b. Run the test docker container

   • sudo dockerd &
   • cd ~/Street-Watch/PavementApp
   • sudo docker-compose -f ~/Street-Watch/PavementApp/docker-compose.yml -f ~/Street-Watch/PavementApp/docker-compose.test.yml up --build

4. b. Test the API

   • sudo docker exec flask python -m unittest

Object Detection and Database Population

1. Download yolov3.weights from google drive
   • https://drive.google.com/a/oregonstate.edu/file/d/1TD6mVECAa86DxqGQdZblFu_iN5RsSsyA/view?usp=sharing
2. Place yolov3.weights into Object_Detection
   • cd ~/Downloads
   • mv yolov3.weights ~/Street-Watch/Object_Detection
3. Follow Web API instructions above
4. While Web API service is running, run video_demo
   • cd ~/Street-Watch/Object_Detection
   • python video_demo.py --video SeattleSeventhAndOlive.mp4
5. Look in output folder to play video post obfuscation
   • cd ~/Street-Watch/Object_Detection/output
   • <video player name> result_SeattleSeventhAndOlive.mp4
6. Run HTTP GET request for stored data from Web API
   • curl -i -H "Accept: application/json" -H "Content-Type: application/json" http://192.168.99.100:5000/pavement

Troubleshooting

Cannot access localhost:5000
After building the container through the docker-compose up --build command, if no url is showing to access the Web API (ex: localhost:5000):

• If using the Docker Toolbox, open the Kinematic Docker API
• Find the Access URL on the righthand side (ex: 192.168.99.100:5000)
• Use the Acess URL within your browser to access the Web API