|
Mjfas templateBog'liq Integration Of Face Recognition Model to a Biometr ALFIYAAAAAAAA, Taqdimot 91-21 7.8.9, Abdulhamid Cho\'lpon, МАЖМУА ЧКБТ 3-КУРС, Презентация, Asliddin, KOVALENT BoG, 11 sonf yakuniy nazoratlari, Amaliy Mashg\'ulot, Контроль качества пищевых продуктов, maruza2, maruza1, 1.0-Zamonaviy kompyuterlarning arxi(6-22), btl 5 lab, Korxona slaydMaterials and methods
Cascade Classifier
The cascade classifier is a machine learning based method that was developed by Paula Viola and Michael Jones. The
method mainly requires positive images (pictures of images we want the classifier to recognize, in this case pictures of
faces) and negative images (pictures of anything that we do not want the model to recognize). The positive and negative
images will first be converted into grayscale images and will be used to create a classifier that will yield a .xml file. The
.xml file will then be used to detect the desired object that we want, which in this case, a face [3].
The training involves the use of 4 features as shown in Figure 1. During the training the classifier will learn the shape of
faces by comparing it to the 4 features. The features can be scaled larger or smaller depending on the classifier and will
move from pixel to pixel to detect the difference in color of the grayscale images in both the positive and negative images.
Figure 1. The 4 features that are used in the training [3].
Eigenfaces
The eigenfaces is a method of facial features detection by calculating the eigenvector and eigenvalues of faces [4]. Firstly,
a picture of a face will be taken, and it will be converted into a grayscale image. The conversion is to allow more space
to be saved and so that the analysis will be faster as grayscale images provides enough features to be processed and
analyzed. Secondly, the matrix of the grayscale image will be calculated to find the mean face. Thirdly, the difference
between the mean face and each face in the database will be calculated to produce a covariance matrix . Lastly, the
eigenvectors and the eigenvalue of the covariance matrix will be calculated which will yield the eigenface which is the
same as the eigenvectors.
Raspberry Pi
The Raspberry Pi, as shown in Figure 2, is single board computer equipped with a GPU, networking components, RAM,
General Purpose Input and Output (GPIO). Bluetooth, and several ports such as USB and HDMI. The Raspberry Pi can
run several operating systems such as TLXOS, Ubuntu, LibreElec, but it generally operates in Raspberry OS which was
known as Raspbian. The Raspberry Pi has 40 GPIO pins that allows it to divert power to electrical components, read
sensors, and run communication protocols.
|
| |
