A Special Design of Job Dropout Faces Detection and Recognition System (JDFDRs)

Abstract

In today’s era of the technology-supported world, Computer Vision and Deep Learning are gaining immense importance in modern life and the contemporary world. The emergence of these techniques has led to the design of detection and recognition systems biometric for faces, allowing verification and identification of people. there is a correlation between these technologies, which represented in (face detection and recognition) and job dropout. Now, if we are talking about job discipline at the times of attendance and departure, then we are talking about a dangerous. “Job Dropout” is an intractable problem that must be placed on the list of priorities of all institutions even those that put fingerprint devices, and Eyes-print devices for entry and exit. With a lot of methods for circumvention, idleness, and escape from the job. This study provides a summary of to build a model for (JDFDRs) (Job Dropout Face Detection and Recognition system). with using the following categories (Location, Lighting, wearing glasses, Chin, and Masks) specified for the image, and with the help of a convolutional neural network (CNN), to classify such images with high resolution. It was implemented in Python using the (OpenCV) library in which the JDFDRs Dataset, which was created for the study privately. and excluding a few of those images with high noise and missing features. The results were classified were computed and the accuracy and error were calculated for the proposed model. and the model achieved a high accuracy of (99.4 percent) and an error rate (−6). This study suggests must the number of database photos increases, with as the increase in the image group increases, the accuracy of detection and recognition of those faces increases, and also work to link our proposed system to the Android system for phones.