Biometric identification technology refers to the technology of identifying and authenticating individuals through their physiological or behavioral characteristics. In terms of application process, biometric recognition is usually divided into two stages: registration and identification. The registration stage collects biometric information of human body through sensors, such as optical information of fingerprint and face by image sensors, and acoustic information of speech by microphone, and the collected data is processed by data pre-processing and feature extraction technology to obtain corresponding features for storage. The recognition process adopts the same information collection method as the registration process to collect information, data pre-processing and feature extraction from the person to be recognized, and then compares the extracted features with the stored features to complete the recognition. From the perspective of application tasks, biometric identification is generally divided into two kinds of tasks: identification refers to the process of determining the identity of the person to be identified from the repository, which is a one-to-many problem; confirmation refers to the process of comparing the information of the person to be identified with the information of a specific single person in the repository to determine the identity, which is a one-to-one problem.
Biometric recognition technology involves a wide range of contents, including fingerprint, palm print, face, iris, finger vein, voice print, gait and other biometric features, and its recognition process involves a number of technologies such as image processing, computer vision, voice recognition and machine learning. At present, biometric recognition is widely used in finance, public security, education, transportation and other fields as an important intelligent authentication technology. The following will introduce fingerprint recognition, face recognition, iris recognition, finger vein recognition, voice recognition and gait recognition.
(1) Fingerprint recognition The fingerprint recognition process usually includes three processes: data acquisition, data processing, and analysis and discrimination. Data acquisition The fingerprint image is acquired through physical sensors such as light, electricity, force and heat; data processing includes three processes: pre-processing, distortion correction and feature extraction; analysis and discrimination is the process of analyzing and discriminating the extracted features.
(2) Face recognition Face recognition is a typical computer vision application. From the application process, face recognition technology can be divided into three processes: detection and positioning, facial feature extraction and face confirmation. The application of face recognition technology is mainly influenced by several factors such as lighting, shooting angle, image occlusion, age, etc. Face recognition technology is relatively mature under constrained conditions, while face recognition technology is still being improved under free conditions.
(3) Iris recognition The theoretical framework of iris recognition mainly includes four parts: iris image segmentation, iris region normalization, feature extraction and recognition, and most of the research work is developed based on this theoretical framework. On the one hand, due to the small size of the iris and melanin occlusion, a high-resolution image sensor is required to image the iris clearly under near-infrared light sources, which requires high sensor quality and stability; on the other hand, changes in the intensity of light can cause pupil scaling, resulting in complex deformation of the iris texture and increasing the difficulty of matching.
(4) Finger vein recognition Finger vein recognition is a technology that uses near-infrared light to image and identify finger veins by using the characteristic that deoxygenated hemoglobin in human veins has good absorption of near-infrared light in a specific wavelength range. Since the distribution of finger veins is very random, the network features are very unique and are internal to the human body and are not affected by the outside world, so the modal characteristics are very stable. The main problem faced by the application of finger vein recognition technology is the imaging unit.
(5) Voice recognition Voice recognition refers to the technology of recognizing the speaker based on the voice features of the voice to be recognized. Voice recognition technology can usually be divided into two stages: front-end processing and modeling analysis. The process of vocal recognition is to match a speech from a certain person with the vocal model in the multi-composite vocal model library after feature extraction.
(6) Gait recognition Gait is the only biometric feature that can be clearly imaged in complex scenes at a long distance, and gait recognition refers to the identification of human identity by body shape and walking posture. Compared with the above-mentioned biometric features, gait recognition is more technically difficult, which is reflected in the need to extract motion features from video and the need for more demanding pre-processing algorithms, but gait recognition has the advantages of long-distance, cross-angle and light insensitivity.
Virtual Reality / Augmented Reality
Virtual Reality (VR) / Augmented Reality (AR) is a new computer-based audio-visual technology. Combined with related science and technology, it generates a digital environment that is highly similar to the real environment in terms of visual, auditory and tactile sensations within a certain range. The user interacts with the objects in the digital environment with the necessary equipment and interacts with each other to obtain the feeling and experience similar to the real environment, which is realized through display devices, tracking and positioning devices, tactile interaction devices, data acquisition devices, and special chips.
From the perspective of technical characteristics, virtual reality/augmented reality can be divided into five aspects: acquisition and modeling technology, analysis and utilization technology, exchange and distribution technology, display and interaction technology, and technical standards and evaluation system according to different processing stages. Acquisition and modeling technologies study how to digitize and model the physical world or human creativity, and the difficulty lies in the digitization and modeling technologies of the 3D physical world; analysis and utilization technologies focus on the analysis, understanding, searching and intellectualization methods of digital contents, and the difficulty lies in the semantic representation and analysis of contents; exchange and distribution technologies mainly emphasize the large-scale digital contents in various network environments The core is open content exchange and copyright management technology; display and exchange technology focuses on various display technologies and interaction methods that conform to human habits of digital content, with a view to improving human cognitive ability of complex information, and the difficulty lies in establishing a natural and harmonious human-computer interaction environment; standards and evaluation system focuses on virtual reality/augmented reality basic resources, content cataloguing, and content management. The standard and evaluation system focus on the standardization of virtual reality/augmented reality basic resources, content cataloging, source coding, etc., as well as the corresponding evaluation technology.
At present, the challenges faced by virtual reality/augmented reality are mainly in four aspects: intelligent**, universal devices, free interaction and perceptual integration. There are a series of scientific and technical problems in hardware platforms and devices, core chips and devices, software platforms and tools, and related standards and specifications. In general, virtual reality/augmented reality shows the development trend of intelligent virtual reality system, seamless integration of virtual and real environment objects, and all-round and comfortable natural interaction.