Files

Download

Download Full Text (1.3 MB)

Description

One of the greatest challenges in developing automation algorithms for aerial surveillance applications is the difficult of gathering data. Algorithm developers rely on infrequent and expensive test flights for obtain imagery datasets. As a result of the status quo, algorithms tend to be gear towards and perform well on specific imagery sets. We present the Hexacopter Drone, the Vision Lab’s newly acquired unmanned aerial vehicle (UAV), used for inexpensive collection of aerial imagery data for various research activities. The UAV, a TurboAce Cinewing 6 hexacopter, carries a Canon 5D Mark III mounted to a separately controlled gimbal and has a flight time of up to 25 minutes. The body is a triple deck carbon fiber structure that is durable and lightweight, foldable arms allow for easy transport and the 15 inch extra heavy duty carbon fiber propellers are resistant to flexing and warping under heavier payloads. A transmitter allows the operator to view a live feed of the video during flight, and the hexacopter and the camera gimbal can both be separately controlled during flight with 2 transmitters. The Naza V2 GPS on board of the UAV allows for auto-stabilization, GPS course-lock, and Return-to-Home features for flying. While using GPS control, the pilot can keep the hexacopter at a certain position using the auto-stabilization feature in order to focus on the camera controls and the image capturing. Utilizing those components, the objective of this project is to create high resolution 3D reconstruction model of vehicles. The hexacopter has the ability to capture a 360° view of vehicles or other objects, and this data can be used as the input to a 3D reconstruction algorithm, namely Dense Point-cloud Presentation (DPR). We present 3D models of scenes that are computed using video captured by the hexacopter.

Publication Date

4-9-2014

Project Designation

Graduate Research

Primary Advisor

Vijayan Asari

Primary Advisor's Department

Vision Lab

Keywords

Stander Symposium poster

Disciplines

Arts and Humanities | Business | Education | Engineering | Life Sciences | Medicine and Health Sciences | Physical Sciences and Mathematics | Social and Behavioral Sciences

High Resolution 3D Reconstruction Using a Hexacopter Drone

Share

COinS