As in Hollywood films, the application of advanced technology makes it easier for the police to catch criminals. In the case of Ade Armando's beating, the National Police used facial recognition technology to track down the perpetrators.
Currently, the police have arrested the three main perpetrators of the beating, but the other three are still being sought.
"With the arrest of the 3 main perpetrators of beating and beating victim Ade Armando, we are currently pursuing 3 more people," said Head of Public Relations of Polda Metro Jaya Kombes Zulpan at a press conference, Wednesday (13/4/2022).
The police then displayed photos of the three people who were being sought to the public. The photos displayed, according to the police, are based on facial recognition recorded at the Dukcapil.
Get to know Face Recognition Technology
We summarized from various sources, Wednesday (13/4/2022) face recognition technology is a facial recognition technology that utilizes artificial intelligence (AI) to recognize the faces of people who have been registered in its database.
In short, with the face recognition feature, the camera can recognize who we are from the data they store in both the cloud server and other storage.
Face recognition technology is one type of biometric identification system. This system identifies a person with special features on the body as well as DNA that distinguish one person from another. Other examples of biometric identification are fingerprints or fingerprints, retina scanning, iris scanning, and voice recognition.
How facial recognition works
The working principle of facial recognition is actually very simple and easy to understand. Cameras and artificial intelligence will perform detailed facial scans of people.
This scan will store all data from the shape of the eyes, jaw, lips, mouth, nose, face size, and others that are in the person's face. Then, the data is stored on a special server belonging to the company or the owner of this face recognition.
Later, every time the face of the person whose data has been scanned passes or performs another scan on the camera with the same database, their data will appear and be verified.
There are four things needed to do facial recognition, namely camera, faceprint, database, algorithm to compare the faceprint of the target face with the faceprint in the database.
After these four things are met, we can start doing facial recognition through at least three stages:
1. Detection. The system will extract patterns in an image and then compare them. If the pattern is the same, the system will assume that there is a face in the image
2. Faceprint creation. A faceprint is a 'mold' or 'model' of a face. To make a faceprint, there are two ways that are usually done, namely:
Geometric approach, by measuring 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 (photometric approach), by analyzing photos and comparing them with a database to identify a person's identity based on their statistics.
Analysis of facial texture (skin texture analysis), by mapping the unique location of pores, lines, or spots on the skin that differ from one person to another.
3. Verification or identification. Verification is different from identification. So, the output will also be different. The picture looks like this:
Verification (or authentication). This method compares facial photo input with user photo data that requires authentication. For example, when unlocking a smartphone with a face, you need to enter your face or photo data first to be stored on the smartphone. When you point the camera at your face, new smartphone access will open.
Identification. The face photo input is compared with all facial photos in the dataset to find people who match the photo input. For example, as in the case of the beating of Ade Armando. The police compared photos circulating on social media with a database of faces of residents in Dukcapil.