Visual Servoing

Image-based visual servoing using circular markers as visual features.

Project Overview

This project implements Image-Based Visual Servoing (IBVS) using color-based circular markers as visual features. The system controls a camera's motion to align current visual features with desired target positions, enabling automated visual positioning and tracking.

Setup and run the simulation:

1. Open the scene scenes/ibvs.ttt in the CoppeliaSim simulator.
2. Install python dependencies:

python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt

3. Run the simulation:

python src/main.py

Command Line Arguments

The system supports various feature transformations and control parameters:

• --rotate or -r: Rotation angle in degrees (default: 0)
• --translate_u or -tu: Horizontal translation in pixels (default: 0)
• --translate_v or -tv: Vertical translation in pixels (default: 0)
• --scale or -s: Scale factor for features (default: 1.0)
• --gain or -g: Control gain for IBVS (default: 0.1)
• --depth or -d: Depth of the features (default: 1.0)

Example:

python src/main.py --rotate 45 --scale 1.2 --gain 0.15

Predefined Transformations

The project includes a Makefile with several predefined transformations:

# Apply 45 degree rotation
make rotation

# Scale features by 1.5x
make scale

# Translate features horizontally and vertically
make translation

# Apply all transformations together
make all

Controls

While the simulation is running:

• Press 'q' to exit the simulation
• The camera window shows both target (green) and current (red) features