Exploiting spectral and spatial information for identification of hazelnut fields using self-organizing maps

Automatic land cover identification using remote sensing images has been essential for agricultural management and monitoring, and is an ongoing challenge. For permanent crops, which are of great importance economically and environmentally, it becomes even more challenging mainly due to varying statistics of their orchards such as existence of different orchard types, different crown sizes even for the same type, different distances between the orchards among various fields, and overlapping crowns. This challenge necessitates utilization of both spectral values and spatial relations of pixels. To accurately determine the fields of permanent crops, hazelnuts in particular, a classification system with hybrid learning, which merges image features map (IFM) and learning vector quantization (LVQ), is proposed in this study. IFM is a variant of self-organizing map (an unsupervised neural learning paradigm successfully used in many applications of remote sensing imagery), which exploits both spectral and spatial information without additional computation of texture. LVQ, however, is a supervised learning for fine-tuning of the class boundaries. Experimental results on finding hazelnut fields using multispectral Quickbird imagery indicate that the proposed method achieves more accurate extraction than the ones obtained based only on spectral or on spatial information.

TASDEMIR Kadim; 

2012-06-04

TAYLOR & FRANCIS LTD

JRC67194

0143-1161,   

https://publications.jrc.ec.europa.eu/repository/handle/JRC67194,   

10.1080/01431161.2012.682659,