Skip to content

kubiq/ICantBelieveItsNotValetudo

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

19 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

valetudo

I can't belive it's not Valetudo

This is a simple map generation companion service for Valetudo which does all the heavy lifting. Since both CPU and Memory are limited on the robot, PNG generation for third-party components has been moved here. The service receives raw map data from the robot via MQTT, renders a map and publishes the resulting PNG image via MQTT and/or makes it available via a built-in web server.

Since I have no use for this service, this is only maintained on a very basic level.

Feel free to open PRs with improvements.

Installation

I can't belive it's not Valetudo is built with JavaScript and requires Node.js and npm to run.

Installation is easy: Clone this repository and run npm install to install dependencies. Then run npm start to start the service. If you prefer running services in containers, this repository includes a dockerfile for you.

Configuration

To configure I can't belive it's not Valetudo, create a file called config.json in the working directory. You can also run npm start to automatically create a default configuration file. If you are running in docker, map the configuration file to /app/config.json.

A basic example configuration would look like this:

{
    "mapSettings": {
        "drawPath": true,
        "drawCharger": true,
        "drawRobot": true,
        "scale": 4
    },
    "mqtt" : {
        "identifier": "rockrobo",
        "topicPrefix": "valetudo",
        "autoconfPrefix": "homeassistant",
        "broker_url": "mqtt://user:pass@foobar.example",
        "caPath": "",
        "mapDataTopic": "valetudo/rockrobo/map_data",
        "minMillisecondsBetweenMapUpdates": 10000,
        "publishMapImage": true
    },
    "webserver": {
        "enabled": false,
        "port": 3000
    }
}

Integration with FHEM, ioBroker, etc

If you set webserver.enabled to true, the map PNG will be available at http://host:port/api/map/image so you can display a map with any home automation software that allows fetching images from a URL. By default, the image data is published via MQTT to the topic set in mqtt.mapDataTopic.

Advanced Map Configuration

The appearance of the map is configured by the mapSettings object. The default settings already produce a nice map, but you can use these advanced settings to further customize the map for better integration into your home automation.

Cropping

The map data received from the robot may contain lots of empty pixels surrounding the area of interest. To avoid huge margins in the output image, the raw data is auto-cropped.

If you don't want that you can either disable it with autoCrop: false or define a custom crop region:

The crop region is defined by the top left corner (crop_x1 and crop_y1) and the bottom right corner (crop_x2 and crop_y2) coordinates of a rectangle. A 200px x 100px crop region with the top left corner at (50px, 50px) may thus be defined as (other settings omitted for brevity):

{
    "mapSettings": {
        "crop_x1": 50,
        "crop_y1": 50,
        "crop_x2": 250,
        "crop_y2": 150
    },
    "mqtt" : {
    },
    "webserver": {
    }
}

Scaling

Since the map data has a resolution of approximately 5 cm per pixel, the resulting images have a relatively low resolution. The map is therefore scaled up by a factor of 4 by default. The scaling factor can be configured by setting mapSettings.scale to the desired value.

To avoid blurred edges, the scaling is done with nearest-neighbor interpolation.

Rotating

Rotating the map can be achieved by setting the mapSettings.rotate to the desired value.

Colors

The map is rendered using a blueish color map by default. The colors of the floor, hard and weak obstacles as well as the robot's path can be set via the mapSettings.colors object, e.g. (other settings omitted for brevity):

{
    "mapSettings": {
        "colors": {
            "floor": "transparent",
            "obstacle_weak": "rgba(0,0,0,0.1)",
            "obstacle_strong": "hsl(120, 20%, 50%)",
            "path": "#333333"
        }
    },
    "mqtt" : {
    },
    "webserver": {
    }
}

All valid CSS color values are accepted.

New Map Colors

You can now set colors for the new Valetudo Maps too - however - only hex values are supported. If you need to set the opacity you can do so using 8 digit hex code. Since new Valetudo maps are using segments, new field in config has been added (other settings omitted for brevity):

{
    "mapSettings": {
        "colors": {
            "floor": "#0000FF",
            "obstacle_strong": "#33333380",
            "segments": ["#000000", "FF0000", "00FF00", "0000FF"],
            "path": "#FFFFFF"
        }
    },
    "mqtt" : {
    },
    "webserver": {
    }
}

Overlay and Underlay Images

The map can be further customized by adding overlay and underlay images. This could be a floor plan showing the real walls or some fancy background image. The images have to be in BMP, GIF, JPEG, PNG or TIFF format and can be scaled as well as positioned relatively to the map. A simple example for a background image would be (other settings omitted for brevity):

{
    "mapSettings": {
        "underlay_path": "/absolute/path/to/background_image.png",
        "underlay_scale": 1,
        "underlay_x": 0,
        "underlay_y": 0
    },
    "mqtt" : {
    },
    "webserver": {
    }
}

And, similarly for a overlay image (other settings omitted for brevity):

{
    "mapSettings": {
        "overlay_path": "/absolute/path/to/overlay_image.png",
        "overlay_scale": 1,
        "overlay_x": 0,
        "overlay_y": 0
    },
    "mqtt" : {
    },
    "webserver": {
    }
}

You probably have to try out various scaling and positioning values before you find a good match between a floor plan and the map generated by the robot.

Note: Overlay and underlay images are applied after rendering and scaling the map data.

About

Valetudo companion service

Resources

License

Stars

Watchers

Forks

Packages

No packages published

Languages

  • JavaScript 95.5%
  • Python 4.3%
  • Dockerfile 0.2%