Insightguard: Machine Learning Empowered Self-Monitoring System for Vision Centric Applications

• Published in: International Conference on Advanced Computing and Communication Systems (Online) Vol. 1; pp. 2012 - 2015
• Main Authors: R, Bharathi, P, Ezra, S, Gowtham, D, Hemanth, M, Nagendran
• Format: Conference Proceeding
• Language: English
• Published: IEEE 14.03.2024
• Subjects: Artificial intelligence; Computer vision system; Conventional neural network; Convergence; Digital images; Distance measurement; Face detection; Face recognition; Faces; Head pose; Magnetic heads; Object recognition; Retina; Sensors; Visualization
• ISBN: 9798350384352
• ISSN: 2469-5556
• DOI: 10.1109/ICACCS60874.2024.10717313

The non-trivial task of designing an automatic picture content recognition system has been researched for several applications, including human identification, face detection, and face recognition. Digital image processing is presented in many forms, one of which is face recognition. The automatic identification of a person in a digital image is the subject of the difficult problem known as automatic face detection. There are numerous algorithms available for use in this procedure. However, there are currently no methods for automatically detecting faces in a variety of application scenarios at low resolutions. This project's computer vision technology can be used to forecast whether or not screens will be within their field of vision. Positioning monitors incorrectly can result in issues that could cause eyestrain. For example, the user might scoot the chair away from the screen or tilt their head back, which would force you to type with your arms outstretched. However, there isn't an automated system in place to gauge the distance between the display and the eye. Thus, we may construct an autonomous alarm system based on face recognition and distance in this project. The maximum is 1.02 metres (3.3 feet), while the minimum is 0.38 metres (1.2 feet). Artificial intelligence can be used to accomplish that. Human head positions can be recorded using a web camera, allowing us to distinguish between foreground and background head positions. then identifying and detecting faces using image processing algorithms. Lastly, use a web camera to determine the distance between the screen and the face. Without the need of any sensors, an alert is automatically generated and sent to users if the distance is less than the pre-defined threshold value.