A Comprehensive Survey: Internet of Remote Things Based Denoising Video Streams by Employing Different Filtering Techniques to Significantly Enhance Video Synthesis

Contenu principal de l'article

Mani Deepak Choudhry, Dr. B. Aruna Devi


The IoT is a version of the Web that does not require human interaction. Corporal objects such as home appliances, vehicles, sensors, and other sensors are improving their ability to sense information and communicate with one another. The IoT is a global network of billions of cognitive and wired "things" that link physical and virtual entities to expand the world's boundaries. Every day, these universal smart things generate vast amounts of data, necessitating urgent data processing on a variety of smart devices. Interconnecting nodes when they are separated over large topographical areas, which is present in Internet of Remote Things, has recently been highlighted as a major challenge for Internet of Things (IoRT). It could be required to predominate in remotely monitoring patients in situations where connectivity infrastructure that enables video transmissions from emergency rooms (ICU) is critical. Noise, which is common in low-light conditions or from limited performance cameras, is a significant cause of compression and quality degradation. Denoising was used regardless of the code for pre-processing or post-processing in past researches that tried to integrate a denoise algorithm and a video encoder. However, since noise has a major impact on compression performance, this method must be closely coupled with encoding. Furthermore, since the encoding method and a denoise algorithm have several parallels, this represents a significant opportunity to reduce computational complexity. To overcome this primary challenge denoising of various videos received from multiple locations is preferred. Denoising is a step in the pre-processing process that improves the image quality and makes it more uplifting. The SNR ratio is high, so to eradicate it a non-linear filter technique called Median Filter is used. So many Median filters are available but 3D filtering provides more efficient measures in preserving the edges of video frames. For benchmarking collective performance PSNR, UIQI & SSI are used as quality assessment metrics to provide a high-performance noise free video. The simulation results show that the 3D median variants outperform other image noise filtering methods in a number of metrics, resulting in a strong noise filter for IoT devices.

Renseignements sur l'article