GSUCrypticons are participating for the MLH, IoT, Harman and First time Hackers
Inspiration -
_ italics _Securing your home made easy with IoT
What it does -
_ italics _It uses openCV and python to run facial recognition and detects an intruder. The system in turn uses a REST service to trigger an alert alarm using Pulse 2.
How we built it -
_ italics _We used python, openCV, Java, Android SDK, Harman Pulse API
Challenges we ran into -
_ italics _We started with an idea which involved the usage of Arduino and Intel Edison to handle the push notifications using google cloud messaging/Twilio IP messaging and also using Bluetooth pairing with JBL pulse 2 and trigger commands to turn the pulse red and play an alert sound. But we faced issues with using Intel Edison and finally we scaled down to the existing system. We were simultaneously trying to use Myo to control the volume of JBL Pulse 2 with hand gestures.
Accomplishments that we're proud of -
_ italics _We tried almost everything from alienware, ocular rift, Leap, BB 8 droid, Android, Java, Python, openCV, php, JQuery, Slim, MySQL, Tableau and lots of stuff that we did not dream of using over a weekend. It was exciting to be runners up for the Macy's app development and being part of the winning team in the GE analytics team.
What we learned -
_ italics _You must do the things you think you cannot do. - Eleanor Roosevelt. _ italics _I think we have done a lot of things in these 2 days and deserve a super sleep after soaking in all that knowledge.
What's next for GsuCrypticons-HIDS(Home Invasion Detection Systems) -
_ italics _Integration with GCM/Twilio and using Intel Edison. Next would be to run Machine Learning Algorithms like Viola-Jones detector in AWS to fasten the positive and negative image training process and use them to classify known and unknown facial features. Adding more classifiers like mouth, body shapes, etc to render better accuracy.
Log in or sign up for Devpost to join the conversation.