Improved Block Based Feature Level Image Fusion Technique Using Contourlet with Neural Network

As multisensory data is made available in many areas such as remote sensing, medical imaging, etc, thesensor fusion has become a new field for research. Multisensor image fusion mainly focuses on combining spatial information of a high resolution panchromatic (PAN) image with spectral information of a low resolution multispectral image (MS) to produce an image with highest spatial content while preserving spectral resolution. A geometrical transform called contourlet transform (CT) is introduced, which represents images containing contours and textures. This paper derived an efficient block based feature level contourlet transform with neural network (BFCN) model for image fusion. The proposed BFCN model integrates CT with neural network (NN), which plays a significant role in feature extraction and detection in machine learning applications. In the proposed BFCN model, the two fusion techniques, CT and NN are discussed for fusing the IRS-1D images using LISS III scanner about the locations Hyderabad, Vishakhapatnam, Mahaboobnagar and Patancheru in Andhra Pradesh, India. Also Landsat 7 image data and QuickBird image data are used to perform experiments on the proposed BFCN model. The features under study are contrast visibility, spatial frequency, energy of gradient, variance and edge information. Feed forward back propagation neural network is trained and tested for classification, since the learning capability of NN makes it feasible to customize the image fusion process. The trained NN is then used to fuse the pair of source images. The proposed BFCN model is compared with other techniques to assess the quality of the fused image. Experimental results clearly prove that the proposed BFCN model is an efficient and feasible algorithm for image fusion.