Author

Roop Pal

Date of Award

2019

First Advisor

Aaron Williams

Second Advisor

Amanda Landi

Abstract

The thesis describes a computer vision software solution which annotates pictures of the New York City skyline with building information. Tourists often come to NYC and are amazed by buildings around them but lack information and context of those buildings in order to fully appreciate them. A solution which can provide information on buildings based on smartphone pictures can therefore improve tourists’ experiences. The solution is broken into three modules: an annotation module for building information, a computer vision module for building identification and highlighting, and an application module for interfacing with the user. The computer vision module takes in an input image and phone sensor data and uses them to predict which building the phone points at by simulating a camera in a virtual 3D model of NYC. The computer vision module is further broken down into components, each using a specific computer vision algorithm. The unit-tested components are successful but the module as a whole often fails due to inaccurate phone sensor data.

Simon's Rock Off-campus Download

Simon's Rock students and employees can log in from off-campus by clicking on the Off-campus Download button and entering their Simon's Rock username and password.

Share

COinS