HSV color space based segmentation of region of interest in satellite images

Abstract

In this paper, a new method for the extraction of the necessary object or region from the satellite image is proposed. This color based method is based on HSV color space and histogram threshold. Usually the multispectral images received from satellite are in RGB color space. But human eye is more sensitive to intensity than color information (hue or saturation). In the proposed approach, the satellite image in RGB color space is transformed into HSV color space and then the transformed satellite image is split into three different components (channels or images) based on intensity and color. This HSV color space is designed to approximate the human vision. In the next step, the histogram for all three components (hue, saturation and value) is computed and plotted. Then the threshold value is separately applied to all three components. Finally the morphological operations like masking, filtering, smoothening is performed for the extraction of the desired region. The effectiveness of the proposed approach has been demonstrated by number of experiments. Experimental result demonstrates that the proposed method is more efficient in the extraction and segmentation of a desired object or region in the satellite images.