多源监测信息融合仿生复眼型系统模式及感知计算机理

A perceptual and computational mechanism for bionic compound eye systems with multi-source information fusion

借鉴昆虫复眼获取与处理信息的机理和过程来构造多源、多尺度监测信息融合仿生处理系统模式.在水资源水环境的空中、地面多传感器时空定量监测应用中,多种卫星遥感传感器和地面监测传感器被设计"集成"在一个虚拟的仿生复眼平台上,通过模仿蝇类昆虫视觉信息获取和生物神经网络感知与计算的机理,实现大场景(LF)、小场景(SF)的协同监测和专题信息融合与决策.提出了空中、地面多传感器时空定量信息获取与处理的仿蝇复眼型信息融合系统模式,设计了基于仿蝇复眼感知与计算机理的多源信息融合工程模型和算法,包括虚拟复眼设计,虚拟复眼图像的预处理模型,LF系统和SF系统的感知与计算机理以及复眼图像超分辨率重构算法.仿真实验中验证了空中、地面多源多尺度监测信息融合的仿生复眼型系统模式、LF和SF系统计算模型以及复眼图像超分辨率重构的可行性与合理性.

The compound eyes of flying insects are highly efficient at processing received information. The authors imitated this action when developing a bionic compound eye system to monitor water resources by fusing quantitative space-time information obtained from many sensors in the air and on the ground. This virtual bionic compound eye platform integrated multiple satellite remote sensors and ground monitoring sensors. By imitating the information acquisition mode of flies and the perceptual and computational mechanisms of their biological neural networks, it is possible to perform synergistic monitoring of large field （LF） and small field （SF） data sufficient for thematic information fusion and decision making. The proposed mathematical model and algorithm were designed on the basis of perceptual and computational mechanisms of bionic compound eyes. Work covered in the paper includes ： design of virtual compound eyes, a mathematical model for virtual compound eye images, perception and computation mechanism of LF and SF systems, and super-resolution reconstruction of virtual compound eye images. The computation model of the LF and SF system and the super-resolution of virtual compound eye images were simulated in a series of experiments which proved that the proposed model is effective.