node-red-contrib-post-object-detection

This is a custom Node-RED node that draws bounding boxes from an Object Detection model.

Installation

Prerequisite

This custom Node-RED node leverages the `node-canvas` npm package to draw the bounding
boxes on a image. Please make sure the platform running Node-RED fulfills the
prerequisites listed [here](https://www.npmjs.com/package/canvas#compiling). For
example, while using Ubuntu, you need to run the following command to install
the dependencies:

sudo apt-get install build-essential libcairo2-dev libpango1.0-dev libjpeg-dev libgif-dev librsvg2-dev

### Install this module:
Once you install the peer dependency and the prerequisites, you can install this module:

npm install node-red-contrib-post-object-detection

## Usage

There are two custom Node-RED nodes in this package:
- `bbox-image`: This is used to annotate an input original image with bounding boxes.


### `bbox-image` node
The `msg.payload` passed to this node should be an object containing these two
properties:
- `image`: The image data in `Buffer` data type
- `objects`: An object array containing a list of detected objects.
             Each object has the following information:

{ bbox: [x, y, w, h], className: string, score: number }

This node annotates the image by drawing the bounding boxes of detected objects onto it.
This annotated image is then output as a `Buffer` for the next node.