We are delighted to share that our paper "ForaNav: Insect-Inspired Online Target-Oriented Navigation for MAVs in Tree Plantations" has been accepted for publication in IEEE Robotics and Automation Letters (RA-L)! We take inspiration from insect foraging behavior to address a critical challenge in precision agriculture: enabling autonomous Micro Air Vehicles (MAVs) to navigate dense tree plantations—without GPS or prior maps.

Introducing ForaNav: an insect-inspired visual navigation system designed for lightweight MAVs.

Key features include:

* A novel HOG-based tree detection algorithm, enhanced with Hue-Saturation histograms and feature variance analysis

* Real-time path adaptation based on visual cues

* A robust recovery strategy for reacquiring temporarily lost targets

* Superior onboard efficiency: lower CPU usage, reduced temperature, and higher FPS compared to lightweight deep learning models

Through real-world flight experiments, ForaNav enables MAVs to: ✅ Dynamically detect and approach all trees ✅ Operate effectively with zero prior knowledge of tree locations ✅ Navigate robustly and in real time within dense plantation environments

Updated: