Information privacy in IoT network

• Main Author: Tan, Chong Beng
• Other Authors: Teh Eng Chuan
• Format: Final Year Project
• Language: English
• Published: 2017
• Subjects: DRNTU::Engineering::Electrical and electronic engineering
• Online Access: http://hdl.handle.net/10356/72005
• Institution: Nanyang Technological University

Internet of thing (IoT) will be the next major technology in the world, therefore it is critical to research about information privacy to help safeguard our personal information online. An algorithm for information privacy was provided in this project, and this project is to improvise a program which uses the Data collected by algorithm to perform privacy Image Detection. The program will be able to infer the presence of Public hypothesis (GUN) with high accuracy, however it will not be able to infer about the Private hypothesis (Cash) with high accuracy and thus protecting the desired valued information one wishes to protect. This improvised program could be used for many other purpose other than the Gun and Cash Example proposed. To demonstrate privacy protection for webcam images, different Sample sets of 200 images are required. Image consists of Gun and Money being generated across the 200 image. These 200 Images will then be reshaped to fit the running algorithm. Resulting Classifier from the program will be used in the created GUI and perform Object Detection. This will result in two output; presence of Gun, presence of Cash. The results of this project is satisfying for a constant image environment, with an low Public error rate of 3% and private error rate of 40% provided that its tested in an empty background. Running the GUI in a different Environment changes the accuracy of the output. However this program will still be valid for camera being mounted in a constant environment. This project will be able act as a mile stone for future development of privacy detection GUI using Frame images instead of a single image frame.