Title :
A binary division algorithm for clustering remotely sensed multispectral images
Author :
Hanaizumi, Hiroshi ; Chino, Shinji ; Fujimura, Sadao
Author_Institution :
Coll. of Eng., Hosei Univ., Koganei, Japan
fDate :
6/1/1995 12:00:00 AM
Abstract :
In binary division clustering (BDC), image data are repeatedly divided into two groups until the group consists of a single cluster. Canonical correlation analysis is used for data compression and for noise reduction. BDC achieved higher accuracy and higher efficiency than conventional ISODATA. BDC was successfully applied to change detection of remotely sensed multitemporal multispectral images
Keywords :
correlation methods; data compression; digital simulation; image processing; image recognition; BDC; ISODATA; binary division algorithm; binary division clustering; boundary search; canonical correlation analysis; change detection; data compression; efficiency; image data; noise reduction; numerical simulation; remotely sensed multitemporal multispectral images; Clustering algorithms; Clustering methods; Data compression; Information analysis; Iterative methods; Multidimensional systems; Multispectral imaging; Noise reduction; Numerical simulation; Training data;
Journal_Title :
Instrumentation and Measurement, IEEE Transactions on
