The Multi-Platform Implementation and Research on MNF Algorithm to Hyperspectral Image

Xiaoqian You, Peng Lu

Abstract


According to the characteristics of hyperspectral remote sensing image data, a processing mechanism of multi-language platform was proposed, as well as doing the challenging experiment of the hyperspectral image feature extraction. Then, to improve the processing efficiency and effects of hyperspectral image feature extraction, it gives a comprehensive assessment from perspectives of the time consumption and the effect of selected characteristics by evaluating pros and cons of the language platform. In this way, a suitable condition for the hyperspectral image data processing will be further found, which also provides a new method for people in this area. Moreover, the experiment result shows that the parallel/MNF algorithm is the most beneficial way. It is obviously and significantly superior to ENVI, Matlab and serial/MNF processing approach in the feature extraction of hyperspectral remote sensing data.

Keywords


hyperspectral image; language environment; feature extraction; maximum noise fraction; ENVI


DOI
10.12783/dtcse/iccae2016/7223

Full Text:

PDF

Refbacks

  • There are currently no refbacks.