講演抄録/キーワード |
講演名 |
2016-11-24 14:50
[ポスター講演]An Efficient Target Detection by Band Reduction and Partial Filtering for Hyperspectral Images ○Lena Chang・Chung-Chi Miao(NTOU) SANE2016-78 |
抄録 |
(和) |
An efficient and robust target detection method composed of band reduction and constrained partial filtering techniques is developed. First, a fast band clustering method Maximum Correlation Band Clustering (MCBC) combined with the optimal spatial searching Particle Swarm Optimization (PSO) and a band prioritization by Impurity Function Band Prioritization (IFBP) are utilized to select dominant bands. Then the selected bands are exploited to design constrained partial filters to alleviate the performance degradation caused by some estimation and random errors. Simulations validate the proposed method can achieve over 90% detection accuracy for all target category in our simulation scenarios. |
(英) |
An efficient and robust target detection method composed of band reduction and constrained partial filtering techniques is developed. First, a fast band clustering method Maximum Correlation Band Clustering (MCBC) combined with the optimal spatial searching Particle Swarm Optimization (PSO) and a band prioritization by Impurity Function Band Prioritization (IFBP) are utilized to select dominant bands. Then the selected bands are exploited to design constrained partial filters to alleviate the performance degradation caused by some estimation and random errors. Simulations validate the proposed method can achieve over 90% detection accuracy for all target category in our simulation scenarios. |
キーワード |
(和) |
Hyperspectral imaging / target detection / Maximum Correlation Band Clustering / Particle Swarm Optimization / Impurity Function Band Prioritization / partial filtering / / |
(英) |
Hyperspectral imaging / target detection / Maximum Correlation Band Clustering / Particle Swarm Optimization / Impurity Function Band Prioritization / partial filtering / / |
文献情報 |
信学技報, vol. 116, no. 319, SANE2016-78, pp. 139-143, 2016年11月. |
資料番号 |
SANE2016-78 |
発行日 |
2016-11-17 (SANE) |
ISSN |
Print edition: ISSN 0913-5685 Online edition: ISSN 2432-6380 |
著作権に ついて |
技術研究報告に掲載された論文の著作権は電子情報通信学会に帰属します.(許諾番号:10GA0019/12GB0052/13GB0056/17GB0034/18GB0034) |
PDFダウンロード |
SANE2016-78 |