Abstract
Artificial intelligence of things has brought artificial intelligence to the cutting-edge Internet of Things. In recent years, compressive sensing (CS), which relies on sparsity, is widely embedded and expected to bring more energy efficiency and a longer battery lifetime to IoT devices. Different from the other image compression standards, CS can get various reconstructed images by applying different reconstruction algorithms on coded data. Using this property, it is the first time to propose a deep learning based compressive sensing image enhancement framework using multiple reconstructed signals (CSIE-M). In this article, first, images are reconstructed by different CS reconstruction algorithms. Second, reconstructed images are assessed and sorted by a no-reference quality assessment module before being input to the quality enhancement module by order of quality scores. Finally, a multiple-input recurrent dense residual network is designed for exploiting and enriching the useful information from the reconstructed images. Experimental results show that CSIE-M obtains 1.88–8.07 dB peek-signal-to-noise (PSNR) improvement while the state-of-the-art works achieve a 1.69–6.69 dB PSNR improvement under sampling rates from 0.125 to 0.75. On the other hand, using multiple reconstructed versions of the signal can improve 0.19–0.23 dB PSNR, and only 4% reconstructing time is increasing compared to using a reconstructed signal.
Highlights
I NTERNET of Things (IoT) interconnects numerous devices including sensors, cameras, smart home products, and smart city products in an environment
Images in the training and testing sets are fed to the Scorenet to get the ranking scores and fed to the quality enhancement module multiple-input residual recurrent network (MRRN) by order of quality scores
In the quality enhancement module, low-level and high-level features extracted from compressive sensing (CS) reconstructed images were exploited and enriched by the proposed recurrent dense skip connections block (RDBlock)
Summary
I NTERNET of Things (IoT) interconnects numerous devices including sensors, cameras, smart home products, and smart city products in an environment. Surveillance systems have been pointed out as one of the most necessary but challenging solutions in urban developments due to large data storage requirements and the high computational complexity in processing images and videos sensed by cameras. Compression methods that can adapt the requirements of 1) saving the power consumption and prolonging the battery lifetime of IoT devices, 2) securing the data, and 3) balancing the traffic load when traveling throughout the network are preferred in designing sensing devices for surveillance systems. In IoT systems such as remote surveillance and astronomy satellites, the Shannon–Nyquist rate is costly, requires ample storage space, and wide bandwidth for transmission. CS has adapted the requirements and becomes one of the effective lossy compression methods that are considered when designing devices for IoT applications [1]–[3]. A study on enhancing CS reconstructed images is
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have