Topics covered in this article:
Face recognition is a type of biometric recognition technology that uses information about a person's facial features to identify them. This technology collects photos or video streams including human faces using a video camera or camera. It then automatically recognizes and tracks human faces in the images, performing a sequence of face-related technologies on the detected human faces. Traditional facial recognition is mainly based on visible light images, which is also a familiar recognition method. Simply put, it is a process of making the computer recognize you.
The traditional facial recognition technology is mainly based on visible light image facial recognition, which is also the most familiar recognition method. It has a research and development history of more than 30 years. This system, however, has insurmountable flaws. When the ambient light changes, the recognition effect drops dramatically, leaving the system unable to achieve its requirements. Solutions to the lighting problem include three-dimensional image facial recognition and thermal imaging facial recognition. But these two technologies are far from mature, and the recognition effect is not satisfactory.
A rapidly developing solution is the multi-light source facial recognition technology based on active near-infrared images. It can overcome the influence of light changes and has achieved excellent recognition performance. The overall system performance in terms of accuracy, stability, and speed exceeds that of 3D image facial recognition. This technology has developed rapidly in the past two to three years, making facial recognition technology gradually practical.
The human face is innate like other biological characteristics of the human body (fingerprints, iris, etc.). Its uniqueness and good characteristics provide the necessary prerequisites for identity authentication.
Compared with other types of biometrics, facial recognition has the following characteristics:
Non-mandatory: The user does not need to specifically cooperate with the face acquisition equipment. It can obtain face images almost unconsciously.
Non-contact: Without making physical touch with the device, the user can obtain facial images.
Concurrency: In real-world circumstances, it is feasible to sort, judge, and recognize many faces.
The facial recognition system mainly includes four components, namely: face image acquisition and detection, face image preprocessing, face image feature extraction, and face image matching and recognition.
Face image collection: The camera lens can collect different face images, such as static images, dynamic images, different positions, different expressions, etc… When the user enters the capture device's shooting range, the gadget will automatically look for and capture the user's face image.
Face detection: In practice, face detection is mainly used for preprocessing facial recognition, that is, to accurately calibrate the position and size of the face in the image. Histogram features, color features, template features, structural features, and Haar features are just a few of the pattern features in face photos. Face detection is the process of extracting usable information and putting it to use in order to detect faces.
face image feature
The Adaboost learning algorithm is used in the standard face identification approach, which is based on the above attributes. The Adaboost algorithm is a classification method. It combines a number of weaker classification methods to create a new, more powerful classification approach.
The Adaboost method is used in the face identification process to select several rectangle characteristics (weak classifiers) that best depict the face. Then the weak classifier is constructed into a strong classifier according to the weighted voting method. And several strong classifiers obtained by training A cascade structure of stacked classifiers are connected in series to effectively improve the detection speed of the classifier.
Face image preprocessing: It is based on the face detection result, and the image is treated before being used in the feature extraction process. Due to diverse situations and random interference, the system's original image is frequently not directly utilized. In the early stages of image processing, it must be pre-processed using grayscale correction and noise filtering. Light compensation, gray scale transformation, histogram equalization, normalization, geometric correction, filtering, and sharpening of the face picture are all part of the preprocessing process for face images.
Face image feature extraction: The features that can be used in a facial recognition system are usually divided into visual features, pixel statistical features, face image transformation coefficient features, and face image algebra features. Facial feature extraction is based on certain features of the human face. Face feature extraction, also known as face representation, is the process of feature modeling of human faces. Two categories of methods of facial feature extraction: one is the representation method based on knowledge; the other is the representation method based on algebraic features or statistical learning.
Face image matching and recognition: The facial recognition system searches and matches the feature data of the extracted face image with the feature template stored in the database, and sets a threshold. The matching result is output when the similarity surpasses this threshold. The face features to be recognized are compared with the obtained face feature templates, and the identity information of the face is judged according to the degree of similarity. There are two steps in this procedure: confirmation and identification. Confirmation is a one-to-one picture comparison process, while identification is a one-to-many image matching and comparison process.
It is performed by using color, contour, texture, structure, or histogram features.
The face template is extracted from the database, and then a certain template matching strategy is adopted to match the captured face image with the image extracted from the template library. The face size and position information are determined by the level of correlation and the size of the matched template.
Through a large collection of "face" and "non-face" images to form a positive and negative sample library of human faces, statistical methods are used to strengthen the training of the system, so as to realize the detection and classification of the patterns of human faces and non-human faces.
Face autofocus and smile shutter technology: The first is facial capture. It judges according to the position of the human head. First, the head is determined, and then the head features such as eyes and mouth are judged. Through the comparison of the feature library, the face is captured. Then autofocus with the face as the focus, which can greatly improve the clarity of the photos taken.
mobile phone 3D facial recognition
Smile shutter technology is based on face recognition to complete facial capture. It judges the degree of mouth bend and the degree of eye bend to judge whether it is a smile. All the above captures and comparisons are done under the condition of comparing the feature library. Therefore, the feature library is the basis, which contains various typical facial and smile feature data.
The security-protected area can recognize the identity of the intruder through face recognition. Facial recognition technology can be utilized for security and management in both businesses and homes.
Face recognition access control is an access control product based on advanced face recognition technology, combined with mature ID card and fingerprint recognition technology. The product adopts a split design. It has the functions of collection of face, fingerprint, and ID card information, and biometric information recognition. The system adopts network information encryption transmission and supports remote control and management. It can be widely used in access control security control in key areas such as banks, the military, public security agencies, and intelligent buildings.
E-passport and ID card. The International Civil Aviation Organization has determined that from April 1, 2010, its 118 member countries and regions must use machine-readable passports. Face recognition technology is the first recognition mode. This regulation has become an international standard. The United States has required countries that have visa-free access agreements with it to use an electronic passport system that incorporates facial fingerprints and other biometrics before October 26, 2006. By the end of 2006, more than 50 countries had implemented such a system. The Transportation Security Administration plans to promote a biometric-based domestic travel document throughout the United States. Many countries in Europe are also planning or implementing similar programs to identify and manage passengers with biometric documents.
Face recognition can monitor people in public places such as airports, stadiums, and supermarkets. For example, surveillance systems can be installed at airports to prevent terrorists from boarding the plane. If the user's card and password are stolen from the bank's ATM, they will be taken by others as cash. At the same time, the application of face recognition will prevent this from happening. Find out whether there is the basic information of key populations in the database by querying the target portrait data. For example, install systems at airports or stations to catch fugitives.
Use face recognition to assist credit card network payment to prevent non-credit card owners from using credit cards, etc. Such as computer login, e-government, and e-commerce. In e-commerce, all transactions are completed online, and many approval procedures in e-government have also been moved online. At present, the authorization of transactions or approvals is achieved by passwords. Security cannot be assured if the password is taken. The digital identity and real identity of parties on the Internet can be unified if biometrics are employed. As a result, the reliability of e-commerce and e-government systems is considerably improved.
You'll find many applications of face recognition technologies in daily life, such as camera shooting, picture comparison, etc.
With the rise of the mobile Internet, some face recognition technology developers apply the technology to the entertainment field, such as applying happy celebrity faces. It is based on the contour, skin color, texture, texture, color, light, and other characteristics of the face to calculate the similarity between the protagonist and the celebrity in the photo.
In the United States, facial recognition software has attracted criticism from privacy and civil rights organizations. For example, this technology has a low recognition rate for people of color. Republican and Democratic lawmakers in the United States called for supervision of this technology at the hearing, saying that the use of this technology may violate constitutional rights and legal procedures. For retailers, sometimes use facial recognition technology to filter customers, shut out those who are considered shoplifters, and instead give preferential treatment to those who spend money.
In 2021, more than 35 American civil rights organizations called on retailers to stop using facial recognition technology to screen shoppers. These civil rights organizations, including public citizens and the National Bar Association, joined the boycott and urged retailers such as Albertson and Macy's not to use facial recognition technology to screen employees and customers. The coalition of civil rights organizations said in a statement that this technology will be abused by retailers, leading to privacy issues. The alliance also stated that it plans to use social media to appeal to retailers who use these tools. The civil rights group "Fight for the Future" stated that this technology may facilitate the exclusion of low-income shoppers and the exploitation of workers.