基于多尺度融合对象级变化检测新方法.pdfVIP

1 34 ò 1 3 Ï g Ä z Æ  Vol. 34, No. 3 2008 3  ACTA AUTOMATICA SINICA March, 2008 Ä u õºÝ KÜ é–? Czu ÿ # { ¿ S X 1 § è 1 © ǘ 1 ±“ 1, 2 Á ‡ ? Ø € a 㔠C z u ÿ  õº Ý KÜ K . Ä k © p© E ǀ a 㔠¢ ½ C z u ÿ  :Ú D Ú C z u ÿ  { Û 5 ; é ù :Ú Û 5 , J Ñ Ä u õ º Ý K Ü  é– ? C z u ÿ µ e . T µ e | ^ é – ? C z u ÿ  { , J p C z a Ú šC z a Œ © 5 . Ǒ J pC z u ÿ ° Ý , T µ e Ú\ ü Ø Ó  õº Ý KÜ ü Ñ . 3 QuickBird p©E ǀ a 㔠þ ¢ (J T{ k 5Ú° •5 . … € a 㔠, é– ?Czuÿ , õºÝ ûü KÜ , õºÝA Æ KÜ ¥ã© aÒ TP751 Object-level Change Detection Based on Multiscale Fusion HUO Chun-Lei1 CHENG Jian1 LU Han-Qing1 ZHOU Zhi-Xin1, 2 Abstract Urban change detection based on multiscale fusion is discussed in this paper. The limitations of traditional change detection methods and the difficulties of urban change detection of very high resolution images are first analyzed. To address the above limitations, a novel object-level framework based on multiscale fusion is presented for urban change detection. Taking advantages of the object-level change detection, the framework can improve the separability of changed class and unchanged class. To further improve the change detection accuracy, two different multiscale fusion strategies are applied in the framework. Experimental results obtained on QuickBird images confirm the effectiveness of the proposed approach. Key words Remote sensing images, object-level change detection, multiscale decision fusion, multiscale feature fusion ¢ C z u ÿ Š ǑC z u ÿ  ‡ A ^ ? C zu ÿ  { ¿ Ø p© E a ã .  , ¢ Ä  i ÿ !¢ ‚ i ÿ , é p © E a ã C z u ÿ  {  9 ¢ y n