The repository contains an edge computing demo for ICN'18
We implement and evaluate a simple Optical Character Recognition (OCR) algorithm to illustrate the execution of pure programs. Our simple algorithm is initialized in the enclave with a model of each letter of the alphabet; it can then be fed with input images to detect the list of embedded letters. We implemented this example to show how our system can be used to outsource execution of pure programs.
The system requires libsfml. On Ubuntu type:
sudo apt-get install libsfml-dev
cd ocr
make
./app ./data/input_5_OK.png
Just add this line to the file /etc/dnsmasq.conf with the name of the script that automatically creates a face towards the Android client in the AP.
dhcp-script=/home/pi/icn_demo/dhcp_log.sh
The NDN OCR Android app for the demo can be found in the following repository:
https://github.com/srene/icn_demo_Android
but also in Google Play:
https://play.google.com/store/apps/details?id=uk.ac.ucl.ndnocr
