Investigation of techniques of image enhancement in remote sensing images

Today, the images processed by remote sensing technology are used in many areas such as agriculture and defence industry. The quality of remotely sensed images are important according to the area of usage. For these purpose image prosesing steps are applied to the remote sensing images. One of the steps is image improvement techniques. Image enhancement is the technique used to improve the interpretability of information, to improve image perception, and to provide a better input image for further image processing steps. Remote sensing image enhancement has to maintain original color information while enhancing image contrast, and provide detail enhancement. In this thesis study, image enhancement techniques for remote sensing images are examined and applied. Image enhancement techniques are used to visually understand and interpret, to improve contrast, to maintain edge information, to limit and emphasize boundaries. This enhancement techniques; Histogram Equalization, Satellite Image Contrast Enhancement Using Discrete Wavelet Transform and Singular Value Decomposition, Remote Sensing Image Enhancement Using Regularized-Histogram Equalization and Discrete Cosine Transformation, Remote Sensing Image Enhancement Using Hazy Image Model, Remote Sensing Image Enhancement Via Bilateral Filtering have been studied and implemented. According to the visual and quantitative results of these techniques, when the histogram equalization technique increases the image contrast while the edge information is not preserved. Satellite Image Contrast Enhancement Using Discrete Wavelet Transform and Singular Value Decomposition Technique increased image contrast, but had corrupted edge information. Remote Sensing Image Enhancement Using Regularized-Histogram Equalization and Discrete Cosine Technique provides an image with improved contrast and emphasized edge information, however the distortions caused by noise could not be removed. In Remote Sensing Image Enhancement Using Hazy Image Model Technique, color preservation of the image is not well preserved whereas an edge-detail-enhanced and contrast-enhanced image is obtained. Remote Sensing Image Enhancement via Bilateral Filtering Technique protects and enhances the edges of the image, however does not provide sufficient color preservation.