Implementation of Multispectral Image Classificationon a Remote Adaptive Computer

[摘要] As the demand for higher performance computers for the processing of remote sensingscience algorithms increases, the need to investigate new computing paradigms isjustified. Field Programmable Gate Arrays enable the implementation of algorithmsat the hardware gate level, leading to orders of magnitude performance increase overmicroprocessor based systems. The automatic classification of spaceborne multispectralimages is an example of a computation intensive application that can benefitfrom implementation on an FPGA-based custom computing machine (adaptive orreconfigurable computer). A probabilistic neural network is used here to classify pixelsof a multispectral LANDSAT-2 image. The implementation described utilizes Javaclient/server application programs to access the adaptive computer from a remote site.Results verify that a remote hardware version of the algorithm (implemented on anadaptive computer) is significantly faster than a local software version of the samealgorithm (implemented on a typical general-purpose computer).

[学科分类] 电子、光学、磁材料