The One-stop Intelligent Environmental Protection Community based on Image Processing Techniques

Abstract

In view of the spread of COVID-19 epidemic and many problems existing in the community, such as potential safety hazards, diluted interpersonal relationships, and out-of-place management, a system of one-stop intelligent environmental protection communities based on the Internet was proposed. It not only improves the ability of community staff, provides great convenience for residents and community workers, cares for vulnerable groups, and promotes a happy and harmonious neighborhood life, but also scores residents and staff while monitoring the community environment for safety with the technology application of thermal imaging recognition, PaddleHub-based face and mask recognition. This new system design is easy to implement at low cost and has a simple structure with many functions. The technologies for face and mask recognition proposed in this paper are based on PaddleHub. Experiments on MaskedFace-Net provided by Haute-Alsace University and the pretrained parameters loaded by PaddleHub showed that the accuracy rate with mask recognition was 94.3290 percent using this method.