Solving Of Real-Color Digital Image Processing Problems by Converting Them to Their Corresponding Three Spectral Values to Enhancement the Remote Sensing Satellite Images Classification.

Document Type : Original Article

Author

Construction Training Institute - Public Authority for Applied Education and Teaching

Abstract

Nowadays, more and more digital images are available in the real-color format. One of the sources of real-color digital images is the real-color digital cameras, which became inexpensive and of high resolutions. Another source is the use of traditional methods of photography and then scanning the photographic film on a digital scanner. Also, remote sensing includes acquiring many types of data such as visible, near-infrared, short-wave infrared images, etc. Many, satellite imagery is available now in real colors, like SPOT imagery. But dealing with these real-color images, from the digital image interpretation point of view, are so difficult, and needs expensive commercial software. Therefore, they are converted and handled with their corresponding spatial-spectral frequencies. Then, the treatment of the 256 color images will be the same as multi-spectral images. The way of converting the color images to their corresponding spectral frequencies is not available in some of the current software dealing with image processing and interpretations. An alternative algorithm has been established in this research, to solve the problem of separating each band of the real-color image alone. This algorithm is applied to a real-color SPOT image. The output results indicate that the developed algorithm is suitable for that kind of work.

Keywords

Main Subjects

Files

Share

How to cite

Statistics