Building a Smart Security Camera with Raspberry Pi Zero, Java and AWS

Explore how to transform a Raspberry Pi Zero webcam into a smart security camera with motion detection, threat detection, and alert notifications in this GOTO Copenhagen 2018 conference talk. Learn about combining open-source software with cloud-based image analysis, setting up the hardware, and implementing software like Motion. Discover how to integrate AWS services such as Rekognition, Lambda, and Step Functions to create a smart camera AWS pipeline. Evaluate the project requirements, address false alarms, and add intelligent filtering. Gain insights into real-time video annotation using the Intel Movidius Neural Compute Stick and analyze its performance. Follow along as the speaker shares their journey of enhancing a constrained IoT device with cloud computing capabilities.

Intro
Talk Outline
Motivation
Hardware Shopping List
Software: Motion
How Motion Works
Project Requirements: Evaluation
False Alarms from Motion
Adding a Smart Filter to the Camera
Project Requirements Reloaded
Adding AWS Rekognition to the Camera
Smart Camera AWS Pipeline
AWS Lambda Functions
AWS Step Functions
Step Function for Image Processing
Data Pipeline Demo
Smart Camera with AWS Evaluation
Introducing the Intel Movidius NCS
Experiment: Real-time Video Annotation
Movidius NCS Performance Analysis
Movidius NCS Evaluation