Research on target recognition based on information fusion of optoelectronic multi-sensors

Target recognition technology is crucial in many fields such as computer vision, military, and remote sensing. However, traditional target recognition relies on sensors with a single property to collect target data, which has obvious limitations and is only suitable for simple environments. It has problems with low recognition accuracy and efficiency. This study aims to break through the limitations of traditional methods and improve the accuracy and timeliness of target recognition in complex environments through optoelectronic multi-sensor information fusion technology, providing more effective solutions for practical applications in related fields. This article focuses on the problem of target recognition and deeply discusses the information fusion technology of optoelectronic multi-sensor. Firstly, research will be conducted on the algorithms for information fusion of optoelectronic multiple sensors, including data layer fusion, feature layer fusion, and decision layer fusion algorithms at different levels, analyzing their principles and applicable scenarios. In terms of application, effectively combining multiple sensors with the same or different properties to construct a multi-sensor system, achieving simultaneous acquisition of raw information of the target. Using various feature extraction methods, such as image-based texture and shape feature extraction, as well as signal based spectral feature extraction, to obtain multi-directional and multi-attribute feature value data of the target. Subsequently, appropriate feature fusion algorithms are used to fuse these feature value data, in order to fully utilize the advantageous information of each sensor. The experimental results show that compared with traditional single sensor target recognition methods, the target recognition technology based on optoelectronic multi-sensor information fusion has achieved significant results. The accuracy of target recognition has been significantly improved, while the recognition time has been significantly reduced. This technology can accurately identify targets in complex background environments, effectively reducing information redundancy and improving information processing efficiency through fusion processing. Starting from the D-S evidence theory, a multi-target feature information fusion recognition framework is established. This method regards the recognized object as an information system, continuously generates various information during operation, and processes the information into target features. Draw recognition conclusions through specific recognition decision rules. An example of multi-target information recognition was given, and the results showed that the method can effectively improve the credibility of target recognition and reduce the uncertainty of recognition.