Alexander Khoury's Project Documentation

Welcome! This website is currently under construction!

Hi, my name is Alexander Khoury.

I am an Electrical Engineering Masters student at the University of California, San Diego.

I specialize and take interest in applications related to the fields of Machine Learning, Robotics, and Controls.

Please check out my resume.

Click on one of the projects below to begin seeing some of the work I have accomplished.

Projects

Real-Time Gait Analysis Insole (Feed-Forward Neural Network, IoT, Design Competition)

Designed, built, and programmed a smart insole prototype that would analyze the users gait in real time and report gait abnormalities. Visualization for demos. Entered in the 2017 Design Competition, won 2nd place.
Orientation Estimation with Panoramic Image Stitching (Unscented Kalman Filter, Image Stitching)

Used a Unscented Kalman Filter to track the orientation of a camera in space, given IMU data, and used the estimated orientation to stitch a panoramic image of the environment.
Simultaneous Localization and Mapping on a Humanoid Robot (Particle Filter, Scan Matching)

Implemented SLAM on data from a from a humanoid robot using odometry and laser measurements from a 2-D lidar.
Robust Object Detection in Highly Variable Environments (Color Segmentation, Exponential Regression)

Trained a set of Gaussian models to the color space to achieve color segmentation, and by analyzing the color regions, my algorithm was able to locate a red barrel in the image. Estimated the distance to the barrel using regression.
Multi-Agent Robotics Research Lab (ROS, Aruco, Controls, Regression)

Utilizing transforms of a detected Aruco cube, localize the position of a three-wheeled omnidirectional robot in the space, and allow for movement commands to be sent to the robot with respect to this cube.
Autonomous Quadcopter for Multiple Robot Tracking and Interaction (ROS, Gazebo, PID)

Establish a reliable 3-D quadrotor simulation environment for the 2017 International Aerial Robotics Competition. Develop planning algorithms for following and landing on the desired moving robots.

Projects

Real-Time Gait Analysis Insole (Feed-Forward Neural Network, IoT, Design Competition)

Designed, built, and programmed a smart insole prototype that would analyze the users gait in real time and report gait abnormalities. Visualization for demos. Entered in the 2017 Design Competition, won 2nd place.

Orientation Estimation with Panoramic Image Stitching (Unscented Kalman Filter, Image Stitching)

Used a Unscented Kalman Filter to track the orientation of a camera in space, given IMU data, and used the estimated orientation to stitch a panoramic image of the environment.

Simultaneous Localization and Mapping on a Humanoid Robot (Particle Filter, Scan Matching)

Implemented SLAM on data from a from a humanoid robot using odometry and laser measurements from a 2-D lidar.

Robust Object Detection in Highly Variable Environments (Color Segmentation, Exponential Regression)

Trained a set of Gaussian models to the color space to achieve color segmentation, and by analyzing the color regions, my algorithm was able to locate a red barrel in the image. Estimated the distance to the barrel using regression.

Multi-Agent Robotics Research Lab (ROS, Aruco, Controls, Regression)

Utilizing transforms of a detected Aruco cube, localize the position of a three-wheeled omnidirectional robot in the space, and allow for movement commands to be sent to the robot with respect to this cube.

Autonomous Quadcopter for Multiple Robot Tracking and Interaction (ROS, Gazebo, PID)

Establish a reliable 3-D quadrotor simulation environment for the 2017 International Aerial Robotics Competition. Develop planning algorithms for following and landing on the desired moving robots.