Research On The Internet Of Things Image Privacy Protect Scheme

In recent years,more and more Internet of Things(IoT)devices have been connected to the network.With the rapid development of IoT and 5G technology,smart cities and smart homes are developing rapidly,which makes people’s lives more convenient.A Real-time images captured by a large number of IoT cameras,which are widely used in real-time monitoring,disaster warning and virtual reality.However,in the process of uploading image data to the data center,it needs to pass through many intermediate nodes and edge devices.It is obviously unsafe to directly transmit data,which will lead to the leakage of personal privacy.IoT devices are widely distributed and resource-constrained devices that collect huge amounts of image data.If traditional encryption algorithms are used to protect privacy,the computational cost is too high,and countries have different definitions of privacy information in images.Most of the image encryption algorithms encrypt the entire image,which not only wastes computing resources but also lacks flexibility.For the smart city and smart home environment,this research proposes two lightweight and flexible image privacy protection schemes in the IoT environment.(1)In smart city,data such as images and videos collected by IoT cameras will be transmitted to the data center through intermediate nodes.These data contain a large amount of sensitive information of users,and direct transmission will lead to information leakage.Considering the requirements of computing resources and flexibility,in this paper,we propose a lightweight image privacy protection scheme that considers the user’s personalized privacy protection requirements.We first introduce an object detection algorithm to detect and extract sensitive regions in images according to the user’s specific requirements.Then we propose a “member-based” method to mask sensitive areas before the images are uploaded to the data center.In particular,the masking operation utilizes the computing power of the cloud platform,so it does not require excessive local computing resources,and the size of the masking “membrane” can be dynamically adjusted.(2)In the smart home system in the 5G environment,there is a risk of privacy leakage in the process of uploading videos captured by IoT cameras to the cloud server.Under the conditions of ensuring the real-time transmission of video streaming and data availability.The method proposed in this paper firstly uses the YOLO v5 algorithm to capture the sensitive area in the image,and then proposes a masking algorithm based on DNA encryption technology and integer vector homomorphic encryption technology to mask the sensitive area.The algorithm provides a dynamically adjusted masking strategy based on user needs,and the algorithm is flexible and lightweight to meet user’s customizable needs.