PFI-TT: A system to secure and verify location, time, metadata, and authenticity of digital images and videos

点击翻译按钮获取中文摘要

The broader impact/commercial potential of this Partnerships for Innovation - Technology Translation (PFI-TT) project is in creating transparency in business transactions that involve data (image and videos) that have news, forensic and evidentiary value. As the technology to manipulate images and create fake videos has become more accessible, the public trust of photos and videos as evidence has been eroding. This project combats this problem by creating a process to prove that images are not fake or manipulated. This is achieved by a secure application that can be used to capture images or videos and instantly register a proof of origin (location, time, and metadata) and forensic identification of the data on the blockchain. This process may restore trust in any business transaction which uses photos or videos as evidence. The commercial value is to both public entities such as in law enforcement and in service businesses required to store images for proof of delivery, inspection, completion of work, etc. Finally, this technology helps online news agencies to enable public verification of news stories, restoring public trust on local digital news platforms. The proposed project builds a system to store proof of origin of data on a blockchain-backed storage system, creating a chain of custody that begins at the time of data creation (such as when a smart phone or camera is used to record a video or take a picture). The process seeks to differentiate between verifiable media and fake media, the latter of which may combat misinformation in online platforms. The intellectual merit is in the development of a novel perceptual hashing algorithm that can achieve 98% precision in detecting tampered images from a mix of original and compressed or enhanced copies; a smart contract that supports confidentiality and integrity of the data and the auditability and accountability of the storage system, while allowing erasure of data to support the Right to be Forgotten. The technology also develops a novel database auditing process to solve the paradox between openness of a public blockchain and data producer’s right to privacy. The goal of this research and development project is to build an end-to-end platform that supports media registration and verification to support the business logic for a variety of public and private entities.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.