A fast restoration method for Hyperspectral image of NLRTAPC

Hyperspectral image(HSI) is an image cube with continuous narrow bands as the dimension, which has the advantages of high precision and multiple details. The most typical problem is that the HSI has a large capacity and is prone to mixed noise. When using the NLRTAPC model to denoise and restore HSI, a large amount of computer resources need to be consumed, so the efficiency is low when on a single machine system. In response to this, the paper proposes a fast processing method, this method takes Hadoop as the processing platform, deploys the NLRTAPC method on the nodes of Hadoop, and in addition, takes the Artificial Bee Colony algorithm(ABC) as the optimization tool, which is responsible for the optimal configuration of the number and parameters of the slices, and the NLRTAPC of each node runs in a distributed parallel manner to restore the slices. When all the nodes are finished, the final results of the nodes are spliced in order, so as to realize the rapid processing of the HSI. After simulation experiments and data analysis, the efficiency of this method is more than a dozen times higher than that of the single machine processing platform, showing the advantages of fast efficient and low error rate.