Occupancy Grid Mapping with Dolly Laser Scan

This project implements an Occupancy Grid Map using laser scan data from the Dolly robot.
An occupancy grid is a two-dimensional discretized representation of the environment, where each cell contains the posterior probability that the corresponding space is occupied by an obstacle.

Source: https://autowarefoundation.github.io/autoware_universe/main/perception/autoware_probabilistic_occupancy_grid_map/laserscan-based-occupancy-grid-map/

---

📖 Concept

• Occupancy Grid

  • Discrete 2D grid of cells.
  • Each cell holds the probability of being occupied, free, or unknown.
  • Map is updated based on range sensor data (LaserScan).

• Assumptions

  • Ego-position and heading are known exactly.
  • Environment is static (no moving obstacles).

• Decision Approach

  • Bayesian update rule is used for occupancy estimation.
  • States are binary: Occupied or Free.
  • Probabilities are transformed into log-odds for stable numerical updates.

---

⚙️ Algorithm Overview

---

🛠️ Installation

1. Create workspace, src and go to src

mkdir temp_ws
cd temp_ws
mkdir src
cd src

2. Clone repository

git clone https://github.com/surendrakoganti/occupancy_grid_map.git

3. Return to workspace and build the package

cd ..
colcon build

4. Source the setup files

source install/setup.bash

---

▶️ Usage

1. Play ros bag file

ros2 bag play -l ./src/occupancy_grid_map/laser_scan_data/rosbag2_2021_05_09-18_39_42_0.db3 

2. Run the node in new terminal (in same directory as previous terminal):

ros2 run occupancy_grid_map hsc_mapping_laser

3. Launch Rviz2 in new terminal and add topic "hsc_map" to view the generated occupancy grid map.

---

🎯 Results

Input laser scan data	Occupancy grid map

---