Wavelet-Fuzzy Model for Efficient Enhancement and Classification of Remote Sensing Images

Abstract

According to the interpolation that DWT can achieve, the resolution of the satellite image can be improved. The input image is separated into many subbands and the speckle noise is eliminated using DWT. After then, IDWT can be used to blend the inputs low-level and high-level and high level images to create a better image. Here, an intermediary step for high-level approximation is suggested. The new fuzzy clustering technique and image fusion strategy are used to vary the detection methods for SAR images. Wavelet fusion directions are taken into consideration in order to combine the wavelet coefficients and return an improved image. To distinguish between the changed and unaffected areas in the combined difference image, a fuzzy C means algorithm is proposed.