Research on personnel management and monitoring system based on real-time image enhancement algorithm
Chengting Zhang, Liujing Wang, Nanzhe Ding, Dingke Shi · 2023
Recently, the demand for fast and real-time personnel management and monitoring system in various companies has increased year by year. It is economical to realize this function with efficient computer vision algorithm compared with using more advanced hardware. Notably, the convolutional neural networks (CNN) have yield incredible performance in the filed of image super-resolution. Nevertheless, most of the previous high-performance SR models needs lots of computational resources and fail to satisfy the real-time requirements of monitoring system. Thus, we propose a lightweight multilevel channel attention network (LMCAN) based on lightweight multilevel feature extraction and fusion (LMFEF) architecture. Besides, the LMFEF is constructed based on the designed lightweight channel attention block (LCAB), which can effectively refabricate the extracted high-frequency information. Adequate experimental results demonstrate that LMCAN is superior than other compared methods quantitatively and qualitatively.