Search and Rescue Neural Networks

Midshipman Researcher(s): 1/C Sofia Di Antonio

Adviser(s): Professor Randy Broussard

Poster #108

Search and rescue missions occur around the world at all hours of the day. The quicker the mission is completed, the higher the survival rate for all involved. A cascade classifier and two different neural networks will be created in MATLAB to detect people in images to make the missions more efficient. The accuracy of each network will be evaluated based on the rate of true and false positives and negatives. The speed at which each network performs will also be taken into consideration when picking the best network.

Full Size Robotics and Controls Engineering #108