Distributed land use classification with improved processing time using high-resolution multispectral data

Image classification techniques can be applied to a geographical image to obtain its land use characteristics. Multispectral and high-resolution remote sensing images are able to provide sufficient information for a more accurate segmentation, nevertheless, the classification algorithms applied to images with high spatial resolution requires many computational cycles, even for modern computers. This paper explores the effectiveness of a novel approach developed for supervised segmentation and classification of high-resolution remote sensing images using distributed processing techniques to improve the computational time required. This is referred to as the distributed pixel statistics method. Examples of remote sensing signatures extracted from real world and high-resolution remote sensing images are reported to probe the efficiency of the developed technique.