Content-sensitive Data Compression for IoT Streaming Services

摘要

The progression of cheaper, faster and more reliable Internet technologies leads Internet of Things (IoT) to be applied in various fields of life nowadays. While tremendous data are generated and delivered from IOT devices, it is essential on the server side having scalable and effective data compression techniques to balance storage usage and data precision. This paper presents an adjustable content-sensitive data compression approach for IOT video streaming services and applications. Specifically, we adopt frame similarity on different aspects including illumination and structure as a matrix for streaming video frame reduction, and we are able to keep sufficient differences among video frames while reducing a significant amount of storage resources. We setup a general iot application platform in practice and show the effectiveness of the presented approach, keeping 96.5% precision with 40% storage reduction, on IOT video streaming applications.