| Autor: |
Zheng, Nanshan, Zhang, Hua, Fan, Jingjing, Guan, Hongjie |
| Předmět: |
|
| Zdroj: |
Remote Sensing Letters; Sep2014, Vol. 5 Issue 9, p843-852, 10p |
| Abstrakt: |
This letter presents a fuzzy local neighbourhood-attraction-based information c-means clustering (FLNAICM) approach to classify very high spatial resolution imagery (HSRI). In which local spatial and grey information based on the neighbourhood attractions between the centre pixel and its neighbouring pixels is incorporated to enhance the insensitivity to noise of the conventional fuzzy c-means (FCM) algorithm and preserve image details. The main contributions of FLNAICM are first, neighbourhood attractions are used as a measure to balance the effects of neighbourhood pixels on the centre pixel. As a result, appropriate local spatial and grey information is incorporated to improve the robustness and noise insensitiveness of the conventional FCM. Second, FLNAICM is fully free of any parameters selection. Two different remotely sensed images are used in experiments to test the performance of FLNAICM. Experimental results indicate that FLNAICM achieves high accuracy and hence provides an effective classification method for HSRI. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
