KinectCam

A security monitoring system for tracking goods and rooms at disaster relief centres, homes, or your workplace.

Comment

Inspiration

During a disaster event, panic and widespread chaos can lead to looting of necessary goods and medicines kept in a storage location. KinectCam serves to prevent against this. Further, many insurance claims in non-disaster situations are also theft related, and KinectCam can prevent against this as well.

What it does

KinectCam uses motion recognition to detect changes in the environment. So, if anyone breaks into a secure room, KinectCam will notice and send you an email with an image of the intruder. This is relevant for relief goods tracking and reducing theft. KinectCam also implements facial detection, so if anyone breaks into your storage room, it will take a picture of the intruder and send you an email notification, thereby reducing theft insurance claims.

How we built it

We used the Microsoft Kinect depth camera along with OpenCV computer vision to implement motion recognition and facial detection algorithms, coded in a Python environment. The video stream generated is coded through a C script that creates a localhost web server, where the video is then streamed for real time display.

Accomplishments that we're proud of

Getting the Kinect video data to be broken down into individual frames in order to apply facial detection and motion tracking, detecting changes over multiple frames for reference.

What we learned

How to use OpenCV and manipulating a pointcloud data stream.

What's next for KinectCam

Applying color correction on processed video Sending the localhost video to a secure web server for remote access Reducing environment sensitivity to increase accuracy

Share this project:

Updates