Joint photographic experts group (JPEG) is the common compression technology of image transmission with the Internet of Things (IoT), which probably causes low compression ratio (CR) and increases the requirement for transmission rate. Therefore, a lightweight image compression method combining JPEG and number theoretic transformation (NTT) is proposed to address the above problems. Simulation results show that the proposed method has higher CR than JPEG, embedded zerotree wavelet (EZW), and combining singular value decomposition with wavelet difference reduction (SVD-WDR) compression algorithms, which reduces the requirement of IoT devices for transmission rate.
K E Y W O R D Scompression ratio, image compression, internet of thing, number theoretic transformation
| INTRODUCTIONAt the dawn of the 5G era, research related to vehicle networking, 1 resource allocation, 2,3 5G Internet of Things (IoT), 4 and multi-access 5 has received extensive attention from industry scholars. Among the research, the deployment of 5G IoT devices is a research hotspot. With efficient resource utilization, manpower saving, and the ability of collecting data, IoT devices are widely used in wireless location, 6 smart city, 7 smart homes, 8 industrial IoT 9,10 and smart healthcare. 11 Under the wide-scale popularity of IoT devices, image transmission based on IoT devices is one of the most widely involved technologies in the above application scenarios. 12-14 Image transmission is essential for underwater imaging, unmanned aerial vehicles (UAV) imaging, and data collection, whereas most IoT devices currently require high transmission rate due to transmitting low CR images. 15 Moreover, IoT devices must fit in with the requirements of limited energy, low computation, and limited storage, whereas low CR images cannot fulfill these requirements. To address these issues, it is important for IoT devices to investigate a lightweight image compression method which has higher CR than the JPEG compression algorithm.JPEG is a widely used lossy image compression algorithm that compresses raw image data to take up only a small amount of transmission bandwidth and storage space by discarding duplicate or unimportant data from the original image and compressing only the important data. 16 It uses a lossy compression mode to remove redundant data from the original image to obtain both high CR and image recovery quality. 17 Moreover, in the early days, the lossless compression algorithms such as Huffman coding and arithmetic coding were widely studied by scholars. 18 Recently,
