Fire Alarm using Watson Visual Recognition

This project is a how-to guide on using Watson Visual Recognition service to analyse an image and based on a trigger word then send a event command to IoT Device.In this example node-red is used to create an image upload form which sends the image to Watson Visual Recognition and trigger an iot device whenever fire is detected in the image.

NOTE: Please check the Video tutorial at the bottom if you have issues.

1. Login/signup in to your IBM Cloud Account
2. Goto Catalog and create the following services:

• Internet of things Platform starter NOTE : You need to create the A Platform Starter App,not Internet of things Platform service.I suggest you use the link above to quickly avoid confusion.
• Visual Recognition

NOTE: You can skip steps 3,4 & 9

3. Goto connections tab and open IotF Service and Launch Watson Iot Platform
4. Navigate to Devices tab and Add Device

5. Once the Starter app is running,visit app url and configure Node-red
6. Login to your node-red flows dashboard and import flow from flows.json

7. Create new visual recognition credentials

8. Update api key in the visual recognition node-red flow
9. Update the device details in IBM Iot device node {Optional}

10. Visit <app_url>/upload and upload required image
11. Change the string inside includes from trigger action function to set a different trigger

Video Tutorial