PROJECT TITLE :

Land-Use Scene Classification in High-Resolution Remote Sensing Images Using Improved Correlatons

ABSTRACT:

Existing strategies that incorporate spatial information into a traditional Bag-of-Visual-Words (BoVW) model consider the spatial arrangement of a picture however ignore pixel homogeneity in land-use remote sensing pictures. During this letter, we tend to gift an improved correlaton model to jointly integrate look, spatial correlation, and pixel homogeneity using multiscale segmentation. The effectiveness of the proposed methodology was tested on a ground truth image information set of 21 land-use categories manually extracted from high-resolution remote sensing images. The experimental results demonstrate that our improved correlaton model will promote classification and outperforms existing ways such as the ancient BoVW model, spatial pyramid matching model, and therefore the ancient correlaton model.