What Is Pattern Recognition?

Pattern recognition is to study the automatic processing and interpretation of patterns through the use of mathematical techniques by computers. The environment and objects are collectively referred to as "patterns". With the development of computer technology, it is possible for humans to study complex information processing processes. An important form of this process is the identification of the environment and objects by living organisms. Pattern recognition focuses on image processing and computer vision, speech and language information processing, brain network groups, and brain-like intelligence. It studies the mechanism of human pattern recognition and effective calculation methods. ^[1]

[mó shì shí bié]

Pattern recognition is through

Early pattern recognition research focused on mathematical methods. In the late 1950s, F. Rosenblatt proposed a simplified mathematical model that simulates the human brain for recognition, the perceptron, and initially realized that the recognition system was trained by each sample of a given category, so that the system has the ability to detect other unknowns after learning. The ability of categories to classify patterns correctly. In 1957, Zhou Shaokang proposed to use statistical decision theory to solve pattern recognition problems, which promoted the rapid development of pattern recognition research from the late 1950s. In 1962, Narasiman proposed a method of syntactic recognition based on primitive relations. Fu Jingsun (K.S. Fu) carried out a systematic and fruitful research on both the theory and application of , and published a monograph "Syntax Pattern Recognition and Its Application" in 1974. In 1982 and 1984, J. Heffield has published two important papers that deeply reveal the artificial neurons and the associative storage and computing capabilities of the network, which has further promoted the research on pattern recognition. In a few years, significant results have been achieved in many applications. , Thus forming a new discipline direction of artificial neural network methods for pattern recognition.

When observing things or phenomena, people often look for differences from other things or phenomena, and group similar but not exactly the same things or phenomena according to a certain purpose. Character recognition is a typical example. For example, the number "4" can be written in various ways, but all belong to the same category. More importantly, even if the "4" of a certain writing style has not been seen before, it can be classified into the category to which "4" belongs. This thinking ability of the human brain constitutes the concept of "pattern". In the above example, the concepts of patterns and sets are separated. As long as you recognize a limited number of things or phenomena in this set, you can identify as many things or phenomena that belong to this set. In order to emphasize the inference of the totality of things or phenomena from some individual things or phenomena, we call such individual things or phenomena as modes. Some scholars believe that the entire category should be called a model. Such a "model" is an abstract concept. For example, "houses" are "models", and specific objects, such as the Great Hall of the People, are called A sample of models like "houses". The different meanings of such nouns are easy to understand from the context.

Pattern recognition is a basic intelligence of human beings. In daily life, people often carry out "pattern recognition". With the advent of computers in the 1940s and the 1950s

Pattern recognition is often called pattern classification. From the perspective of the nature of the problem and the method of solving the problem, pattern recognition is divided into two types: supervised classification and unsupervised classification. The main difference between the two is whether the category to which each experimental sample belongs is known in advance. Generally speaking, supervised classification often needs to provide a large number of samples of known categories, but in practical problems, this is a certain difficulty, so it becomes necessary to study unsupervised classification.

Patterns can also be divided into two forms, abstract and concrete. The former such as

Pattern recognition can be used for text and

Pattern recognition technology is the basic technology of artificial intelligence. The 21st century is a century of intelligence, information, computing, and networking. In this century characterized by digital computing, as a basic discipline of artificial intelligence technology, pattern recognition technology must be Will get huge development space. Internationally, major authoritative research institutions and companies have begun to attach pattern recognition technology as the company's strategic R & D focus.

1.Speech recognition technology

Speech recognition technology is gradually becoming the key technology of Human Computer Interface (HCI) in information technology. The application of speech technology has become a competitive new high-tech industry. The market forecast of China Internet Center: In the next 5 years, there will be a market capacity of more than 40 billion yuan in the field of Chinese speech technology, and then it will grow at a rate of more than 30% each year.

2.Biometric authentication technology

Biometrics is the security authentication technology that has attracted the most attention in this century, and its development is the general trend. People are willing to forget all passwords, throw away all magnetic cards, and use their uniqueness to identify identity and confidentiality. International Data Group (IDC) predicts that biometrics, the core technology of mobile e-commerce as the inevitable development direction of the future, will reach a market size of 10 billion US dollars in the next 10 years.

3.Digital watermarking technology

Digital Watermarking, which has only begun to develop internationally since the 1990s, is the most promising digital media copyright protection technology. IDC predicts that the global market capacity of digital watermarking technology will exceed USD 8 billion in the next 5 years. ^[2]

Pattern recognition has developed from the 1920s to the present day. A common view is that there is no single model and single technology that can be used to solve all pattern recognition problems. What we have now is a tool bag. What we have to do is Combining statistical and syntactic recognition with specific problems, combining statistical pattern recognition or syntactic pattern recognition with heuristic search in artificial intelligence, combining statistical pattern recognition or syntactic pattern recognition with support vector machine machine learning, Combining artificial neural network with various existing technologies and expert systems in artificial intelligence, uncertain reasoning methods, in-depth grasp of the effectiveness and possible possibilities of various tools, complement each other, and create a new situation in pattern recognition .

There are various theoretical explanations for the ability to recognize two-dimensional patterns. The template theory holds that every pattern we know has a corresponding template or miniature copy in long-term memory. Pattern recognition is matching with the most suitable template for visual stimulation. According to the feature theory, visual stimuli are composed of various features, and pattern recognition is to compare the features of the presented stimulus with the features of the patterns stored in long-term memory. Feature theory explains some bottom-up processes in pattern recognition, but it does not emphasize environment-based information and expected top-down processing. Structure-based theory may be more appropriate than template theory or feature theory.