Applying Blockchain Technology to Secure Object Detection Data

Abstract

Recently, blockchain, deep learning, and computer vision are considered the most fundamental technologies in the world. The researchers have been paying much attention to use them. The centralized database system’s problem has low security because it must be transferred through a trusted third party, while the data transmitted might be hacked. However, blockchain provides a distributed database that makes the network secure, flexible, and capable of supporting real-time services. This article proposes a blockchain-based object detection model. The blockchain system makes files secure using the Interplanetary Files System (IPFS) that receives our files, stores them on the decentralized application, and sends the hash to Ethereum to be stored. A YOLOv3 object detection model receives our encrypted data as a hash value and detects objects from images that consist of a huge dataset of different objects.