What Is Multisensor Data Fusion?

Humans instinctively have the ability to synthesize information (views, sounds, smells, touches, etc.) detected by various organs (eyes, ears, nose, limbs, etc.) on the body with prior knowledge in order to their surrounding environment And ongoing events. Multi-sensor information fusion is actually a functional simulation of the human brain's comprehensive processing of complex problems. Compared with single sensor, the application of multi-sensor information fusion technology can improve the survivability of the system, improve the reliability and robustness of the entire system, enhance the credibility of the data, and improve accuracy in solving problems such as detection, tracking, and target recognition. Extend the system's time and space coverage, increase the system's real-time performance and information utilization. As one of the research hotspots of multi-sensor fusion, fusion methods have always been valued by people. In this regard, a lot of research has been done abroad, and many fusion methods have been proposed. At present, the common methods of multi-sensor data fusion can be roughly divided into two categories: stochastic and artificial intelligence methods. Different levels of information fusion correspond to different algorithms, including weighted average fusion, Kalman filtering, Bayes estimation, statistical decision theory, probability theory methods, fuzzy logic reasoning, artificial neural networks, DS evidence theory, and so on.

Humans instinctively have the ability to synthesize information (views, sounds, smells, touches, etc.) detected by various organs (eyes, ears, nose, limbs, etc.) on the body with prior knowledge in order to their surrounding environment And ongoing events. Multi-sensor information fusion is actually a functional simulation of the human brain's comprehensive processing of complex problems. Compared with single sensor, the application of multi-sensor information fusion technology can improve the survivability of the system, improve the reliability and robustness of the entire system, enhance the credibility of the data, and improve accuracy in solving problems such as detection, tracking, and target recognition. Extend the system's time and space coverage, increase the system's real-time performance and information utilization. As one of the research hotspots of multi-sensor fusion, fusion methods have always been valued by people. In this regard, a lot of research has been done abroad, and many fusion methods have been proposed. At present, the common methods of multi-sensor data fusion can be roughly divided into two categories: stochastic and artificial intelligence methods. Different levels of information fusion correspond to different algorithms, including weighted average fusion, Kalman filtering, Bayes estimation, statistical decision theory, probability theory methods, fuzzy logic reasoning, artificial neural networks, DS evidence theory, and so on.

Chinese name

Multi-sensor data fusion technology

Time

1980s

Brief introduction

Integrated decision-making process

Features

Fast and correct

Introduction to Multisensor Data Fusion Technology

Multi-sensor data fusion technology was formed in the 1980s and has become a research hotspot. It is different from general signal processing, and different from the monitoring and measurement of single or multiple sensors, but a higher-level comprehensive decision-making process based on the measurement results of multiple sensors ^[1] .

Multi-sensor data fusion technology definition

The definition of sensor data fusion can be summarized as synthesizing the local data resources provided by multiple homogeneous or different types of sensors distributed in different locations, using computer technology to analyze them, and eliminating possible redundancy and Contradictions, complement each other, reduce their uncertainty, and obtain a consistent interpretation and description of the measured object, thereby improving the speed and accuracy of system decision-making, planning, and response, and enabling the system to obtain more sufficient information. Its information fusion appears at different information levels, including data layer fusion, feature layer fusion, and decision layer fusion.

(1) Data-level fusion. For the data collected by the sensor, it depends on the type of sensor to fuse similar data. The data to be processed for data-level fusion are collected under the same category of sensors, so data fusion cannot process heterogeneous data.

(2) Feature-level fusion. Feature-level fusion refers to the extraction of the feature vectors contained in the collected data to reflect the attributes of the monitored physical quantities. This is the fusion of the characteristics of the monitored objects. For example, in the fusion of image data, edge feature information can be used instead of all data information.

(3) Decision-level integration. Decision-level fusion refers to high-level fusion based on the data features obtained by feature-level fusion, certain discrimination, classification, and simple logical operations, and higher-level decisions based on application requirements. Decision-level convergence is application-oriented convergence. For example, in the forest fire monitoring and monitoring system, through the fusion of data characteristics such as temperature, humidity and wind, you can determine the dryness of the forest and the possibility of fire. In this way, the data that needs to be sent is not the value of temperature and humidity and the magnitude of wind, but only the possibility of fire and the degree of harm. In the realization of the specific data fusion of the sensor network, the fusion method can be selected according to the characteristics of the application.

Advantages of multi-sensor data fusion technology

Multi-sensor data fusion has the following advantages over single-sensor information: fault tolerance, complementarity, real-time, and economical, so it is gradually popularized and applied. In addition to military applications, it has been applied to automation technology, robotics, marine surveillance, seismic observation, construction, air traffic control, medical diagnosis, and remote sensing technology.

In view of the increase in the miniaturization and intelligence of sensor technology, based on information acquisition, multiple functions are further integrated to merge, which is an inevitable trend. Multi-sensor data fusion technology also promotes the development of sensor technology.