go-aruco

This Go (Golang) package provides the position of Aruco markers in view of the camera. It provides the translation vector and the rotation vector.

The package uses Python to access both the Camera and OpenCV library.

Tested on a Raspberry Pi 4 with Camera Module 3 WIDE infrared with Raspberry Pi OS "Bookworm".

See the sample code on example-2d on how to use it.

Install dependencies

You need Go compiler (tested with Go compiler v1.23.2) an the Python interpreter (tested on Python v3.13) and the Python OpenCV dependencies (tested with v4.10).

To install the Python dependencies use the Raspberry OS package manager:

sudo apt update && sudo apt upgrade -y && sudo apt install -y \
    python3 \
    opencv-data \
    python3-opencv \
    python3-scipy

Python installation verification

Check that the script runs alone (without Go) by executing:

python ./markers.py

It requires a few seconds to load all dependencies.

Note that every line is suffixed with few kB of spaces (used for the Python to Go communication). Look for JSON data like this:

[{"id": 9, "x": -0.13673050917742155, "y": 0.06945671561055802, "z": 0.7309941550510376, "roll-x": -172.7556401203728, "pitch-y": 6.610564919377947, "yaw-z": 178.17154128756582}, {"id": 6, "x": 0.34398351230356095, "y": -0.020529717758146982, "z": 0.8951305793554741, "roll-x": -177.8320267892627, "pitch-y": -26.86036068104307, "yaw-z": 177.5710900851925}, {"id": 8, "x": -0.13820595929256171, "y": -0.03209965216632465, "z": 0.7309398765039592, "roll-x": -171.51842160897496, "pitch-y": 2.0533179575661684, "yaw-z": 176.7455522076878}, {"id": 7, "x": 0.34751714240210885, "y": 0.07921186314686002, "z": 0.8838179732333362, "roll-x": -173.4195957396569, "pitch-y": -14.31427564710096, "yaw-z": 178.3094912498296}]

Install package

Usage

Use go get to download the dependency.

go get github.com/jig/go-aruco@latest

Then, import it in your Go files:

import aruco "github.com/jig/go-aruco/log"

Camera hardware verification

Check that your camera is working with following command on the Raspberry Pi Console:

rpicam-hello

Camera calibration

Script is already calibrated for a Raspberry Pi Camera Module 3 WIDE. You might need to change it for other cameras by changing the matrices on constants camera_matrix and distortion_coefficients on markers.py Python script. You can use github.com/jig/charuco-calibration for that.