Multi Focus Image Fusion Based On Spatial Frequency and LDP Analysis under DWT Domain

Abstract

The project presents multi focus image fusion using stationary wavelet transform with local directional pattern and spatial frequency analysis. Multi focus image fusion in wireless visual sensor networks is a process of fusing two or more images to obtain a new one which contains a more accurate description of the scene than any of the individual source images. In this project, the proposed model utilizes the multi scale decomposition done by stationary wavelet transform for fusing the images in its frequency domain. It decomposes an image into two different components like structural and textural information. It doesn’t down sample the image while transforming into frequency domain. So it preserves the edge texture details while reconstructing image from its frequency domain. It is used to reduce the problems like blocking, ringing artifacts occurs because of DCT and DWT. The low frequency sub band coefficients are fused by selecting coefficient having maximum spatial frequency. It indicates the overall active level of an image. The high frequency sub band coefficients are fused by selecting coefficients having maximum LDP code value. The finest details of two images are characterized by local directional pattern descriptors before fusion and it describes local primitives including different types of curves, corners and junctions. LDP computes the edge response values in all eight directions at each pixel position and generates a code from the relative strength magnitude. Finally, fused two different frequency sub bands are inverse transformed to reconstruct fused image. The system performance will be evaluated by using the parameters such as Peak signal to noise ratio, correlation and entropy.

Keywords:LDP,NSCT(Non-sampled counterlet transform),DWT(Discrete wavelet transform),PSNR