A Novel Partial and Multiple Correlations Method with Particle Swarm Optimization for Hyperspectral Image Dimension Reduction

• Main Authors: Ming-Long Li, 李明隆
• Other Authors: Jyh-Perng Fang
• Language: zh-TW
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/tk9w43

碩士 === 國立臺北科技大學 === 電機工程研究所 === 103 === In recent years, the satellite image technologies have greatly advanced remote sensing community, resulting in the increased number of bands acquired by hyperspectral sensors. The band selection of hyperspectral imagery can reduce the dimensions which can avoid the Hughes phenomena. Therefore, the band selection of hyperspectral imagery has become very important. A band selection algorithm based on particle swarm optimization (PSO) is proposed in this paper. By using the PSO algorithm, the highly correlated bands of hyperspectral imagery can be grouped into the same modules which can extract the most useful information of hyperspectral bands in each module and can further reduce the dimensionality. However, when processing a few more Class of hyperspectral image, correlation coefficient matrix can’t effectively make a high correlation band aggregation to find a common uniform correlation coefficient matrix in each class. Therefore, this paper application &;quot; partial correlation coefficient &;quot; and &;quot; multiple correlation coefficient &;quot; increased for definitions associated, to further enhance the use of multi-dimensional perspective of hyperspectral image bands selected effect. The effectiveness of the proposed method is evaluated by AVIRIS remote sensing images for testing. The experimental results can be learned, &;quot; partial correlation coefficient &;quot; and &;quot; multiple correlation coefficient &;quot; of the proposed selection for the band can be effectively increased dimensionality reduction rate, and get a good classification results through classification.