Why Infrared Face Recognition Can Work in the Dark

Face recognition is commonly used in the field of biometrics in recent years, and it has been widely applied in the field of public safety. At the same time, face recognition modules are also diversified, such as static face recognition, dynamic face recognition, 3D structured light face recognition, etc. In fact, all kinds of face recognition methods are similar, and the key lies in how to extract the identity characteristics from the face data, while eliminating the part affected by non-identity factors.

The Advantage of Infrared Face Recognotion

The main problem faced by the face recognition system in applications is the lighting. The face recognition system need to be able to recognize in different lighting environments. However, most face recognition systems use visible light face images for recognition. Such systems are easily affected by changes in ambient light and cannot accurately recognize under complex lighting conditions. Therefore, some lighting pre-processing algorithms are required before recognition. Although the lighting pre-processing algorithm can eliminate the influence of illumination to a certain extent, it will also cause the image to lose some useful information and affect the recognition accuracy.

Infrared face recognition is a solution to solve the illumination problem in face recognition. This technology uses the principle that any object has infrared light emission above absolute zero (-273°C). The infrared light emitted by the human body and heating objects is relatively strong, while the infrared light emitted by other non-heating objects is very weak. Infrared light imaging and identification can be realized by using a special infrared camera. Therefore, the performance of infrared face recognition is not affected by the change of ambient light.

In addition, infrared cameras can image in complete darkness, so infrared face recognition technology has a hardware foundation that can perform face recognition at night.

Infrared Image Acquisition Technology

Infrared image acquisition technology is divided into passive infrared acquisition technology and active infrared acquisition technology. Passive infrared camera technology is rarely used because of the high cost of equipment and the inability to reflect the surrounding environment. Active infrared camera technology uses a special “infrared lamp” to artificially generate infrared radiation, which is invisible to the human eye but can be captured by ordinary cameras, and to illuminate the scene and environment. It uses ordinary low-illumination CCD black-and-white cameras or infrared low-illumination color camera to capture the infrared light reflected from the surrounding environment, so as to realize image acquisition and recognition.

The Challenge of Infrared Face Recogntion

Although infrared face recognition has obvious advantages over traditional visible light face recognition, the use of active light sources also brings some problems.

● The active light source will produce obvious reflections on the glasses, reducing the accuracy of eye positioning.

● The existing visible light data cannot be used, and the user needs to rebuild the infrared face database to train the AI models, which is time-consuming and labor-intensive.

Infrared Face Recognition Solution

In order to handle the problem that the data shortage may occur, has developed “10,000 People — Infrared Face Recognition Data” for infrared face recognition. Datatang strictly abides by the relevant regulations, and the data is collected with proper data collection authorization agreement. Datatang is committed to promoting the innovation of infrared face recognition technology with high-quality data and fully ensuring the data security.

10,000 People — Infrared Face Recognition Data

10,000 people, 28 images for each person (RGB + IR). Multiple ages, multiple facial postures, multiple scenes. The ID, race, gender, age, facial action, collect scene is annotated. The accuracy of annotation exceeds 97%.

Datatang is always committed to protecting data security and ensuring the interests and privacy of the participants. Datatang has obtained the ISO27701:2019 privacy information management system certification.

About Datatang

Founded in 2011, is a professional AI data service provider and committed to providing high-quality training data and data services for global AI companies. Relying on own data resources, technical advantages and intensive data processing experiences, Datatang provides data services to 1000+ companies and institutions worldwide. Datatang entered Chinese stock market (NEEQ: 831428) in 2014 and became the first listed company in China’s AI data service industry.

If you need data services, please feel free to contact us: info

Off-the-shelf AI training data, on-demand data collection & annotation services