RECOGNITION OF HUMAN FACES AND TEMPERATURES BASED ON AUTOMATICALLY INTELLIGENT ALGORITHM

Abstract

The surveillance system plays an essential role in the security field to recognize early illegal access. As a part of such system, the human face recognition provides an effective service to observe and identify people crossing a specific area. Moreover, measurement of body temperature is also mandatory due to Coronavirus pandemic. Therefore, temperature prediction through surveillance system is the best selection because of its effectiveness, high authentication, and privacy maintenance. In this paper, a novel algorithm is proposed for recognition of the human faces and temperature using deep learning method and thermal face images. The search-based method is developed for selection of an optimal structure and parameters of the convolutional neural network using the thermal images as the input. In addition, human temperature is computed directly on the thermal images to assist warning system in terms of abnormal body temperatures. The proposed application produces a relatively high accuracy of over 97%, which is fitted significantly in practical environment.