The current data security system has various forms, ranging from security systems using a PIN (Personal Identification Number), identity cards, and passwords, However, the presence the use of this security system does not seem to guarantee the security and protection of one's data. This method can be easily discovered by others or broken into by sophisticated systems so that it can cause losses. The current security system is developed using facial image recognition technology that can protect the security system well, namely using Biometric technology. In its development, biometric security systems are increasingly in demand because they are considered more accurate and cannot be faked.
The use of biometric technology is certainly driven by machine learning algorithms. You could say, facial recognition is one of the most effective biometric identification systems for generating information. This can happen because of the use of machine learning algorithms in the form of neural networks which play an important role in the development of facial recognition systems.
This algorithm tries to mimic the brain's process of recognizing a person's face. The subconscious mind will try to recognize special features on the face. Such as the distance between the eyes, forehead height, nose width, and so on. A facial recognition algorithm is designed to map a person's facial features mathematically. The facial feature data, called a faceprint, is then stored to be matched with the search results. Facial recognition is a type of "biometric" identification system.
Other examples of biometric identification are fingerprints or fingerprints, retina scanning, iris scanning, and voice recognition. In this DQLab article, we will discuss the application of machine learning algorithms through face recognition. This article is specifically made for you data lovers in the data field to find out the application of machine learning algorithms, especially in facial recognition.
According to the US Government Accountability Office, there are four components needed to perform facial recognition, namely: Camera, Face Print, Database, and Algorithm to compare the face print of the target face with the face print in the database. After these components are met, we can start doing facial recognition. The stages are usually passed, namely:
1. Detection. The method will remove marks in a picture and then correspond to them. If the patterns are the same, the system will assume that there are faces in the image.
2. Faceprint creation. To make a face print, two ways are usually done, namely:
- Geometric approach. Measures the distance and spatial relationships between facial features such as the center of the eye, the tip of the nose, or the line of the lips to recognize faces.
- Photometric approach. Analyze photos and compare them with a database to identify a person's identity based on their statistics.
- Facial texture analysis (skin texture analysis). Map out the unique locations of pores, lines, or patches on the skin that differ from person to person.
