Android Phone App JOHN MICHAEL MARIANO UNDERGRADUATE MECHANICAL ENGINEERING STUDENT EEL 4665...

Post on 24-Dec-2015

213 views 1 download

Tags:

Transcript of Android Phone App JOHN MICHAEL MARIANO UNDERGRADUATE MECHANICAL ENGINEERING STUDENT EEL 4665...

Android Phone AppJOHN MICHAEL MARIANO

UNDERGRADUATE MECHANICAL ENGINEERING STUDENT

EEL 4665 INTELLIGENT MACHINES DESIGN LABORATORY

OCTOBER 30, 2014

Agenda•Objectives

•Android App Information

•Accomplishments

•Problems

•Timeline

Objectives•Perform real time image processing

•Identify objects as vegetables through specific algorithms

•Tell the Arduino Board that the object is a vegetable, and then the robot will approach the object

Android App Information•Android App developed on Eclipse IDE using OpenCV libraries

•Currently is able to initialize and use the camera on the Google Nexus 5 phone

•Connected to the Arduino Mega R3 Android Accessory Development Kit (ADK) Board - ATmega2560 ADK R3 via USB

(Source:http://www.adafruit.com/products/563)

Accomplishments•Successfully downloaded all needed software:

Eclipse IDE, Android SDK, OpenCV

•Created app that initializes the use of the camera

•Purchased and obtained the new Arduino board.

•Eliminates the need for communication through Bluetooth or internet

Problems•Determining which is the best method to do object detection: SURF (identifies key points in relation to a host image) or cascade classifying (trains the program to identify specific features or objects)

•SURF requires using NDK, C++ knowledge

•Cascade Classifying requires numerous images to train the program (~100 images)

Time Line•Visit the Mobile App Workshop at Marston Library for aid (October 30, 2014)

•Hopefully finish the app by Friday for partial credit (October 31, 2014)

•Put the rack and pinion system together with the robot platform (November 3-November 7)

•Continue to test and debug as needed

•DATE OF COMPLETION: November 24, 2014 (Week of Thanksgiving Break)