Abstract
The images received from the satellite contains huge amount of data for further processing in image analysis. An efficient and effective segmentation method is essential to retrieve or extract the necessary information from the satellite images. The images received from satellite are usually in RGB color space. This color space is not preferred for image segmentation because this space is not perceptually uniform and all components should be quantized with the same precision. In this paper, efficient satellite image segmentation based on YIQ color space and Modified Fuzzy C Means clustering is proposed. The YIQ color space takes the advantage of the information that our human eye is very sensitive to intensity changes than changes to saturation or hue and the intensity component can be stored with greater accuracy. In YIQ color space, intensity information is separated from color data. So the same image can easily be analyzed by using both color and intensity information. In the proposed approach, the satellite image in RGB color space is transformed into YIQ color space and then the transformed satellite image is split into three different components (channels) based on luminance and chrominance. Histogram equalization is performed on the luminance component. Subsequently Fuzzy based segmentation is applied for efficient segmentation. This proposed approach is applied to analyze the satellite images of various format and size. The experimental result shows that the proposed method is efficient for extracting information from the satellite images.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call