ABSTRACT
Face detection has multiple applications including recognition, people identification and detection of facial expressions. With the current pandemic crisis and due to the measures imposed to prevent Covid-19 spreading, the wearing of a protection mask became mandatory. The object of interest of this paper is to detect the wearing of an approved mask face using Viola Jones algorithm, aggregate channel features (ACF) and mathematical morphology. COVID-19 virus spread through the air, so it is necessary that all the materials used for manufacturing of the face masks to filter properly the air, and only the approved face masks to be used in order to control the spread infection. The algorithm used for face detection is Viola Jones with notable success in real time face detection and real time impression speed in face detection. Identification of the approved face mask against Covid-19 virus is made with a trained ACF detector. Eye detection, necessary to check if the face mask is properly placed, is based on mathematical morphology operators. These operations used together are robust with high results on the image processing. © 2023 SPIE.
ABSTRACT
A notable increase in the distribution of digital images was fueled by the social media platforms. In the fight for attention, the posts containing images have more chances to make users stop scrolling through the crowded news feed than the posts containing only text. The effect is enhanced if the photos that are used in the posts are professional grade. In the context of the COVID-19 pandemic, more and more businesses have moved to online. Every business makes efforts to catch the attention of potential customers with high quality images. These situations could lead to intended or accidental image copyright infringement. This paper proposes a system that can be used to detect and avoid copyright infringement. Photographers can use it to register their photos and businesses can use it to check if the image they want to use is copyright protected or not. If it is, the system also allows the purchase of the right to use the photo. The solution is based on blockchain because of its immutability property. Businesses can look for images on many sites, thus it is very probable that recompressed versions of the same photo would be stored in different places. Therefore, it is important to develop an algorithm that can extract a signature that is resistant to JPEG compression from the image. The signature should be stored on the blockchain along the identification data of the copyright owner.