Digital Image Slicing and Synthesis using DCT Diagonal Filter Bank

According to Fourier theory signals of many type, are decomposable into their frequency components. This work demonstrates the decomposition of a digital image into its frequency spectrum and its synthetization process from its components to its spatial form. For this purpose, the image as a 2D signal will be transformed into its spectrum using Discrete cosine transform (DCT), decomposed into frequency components using automatically generated diagonal filter bank (DFB) and transformed back to its spatial form using inverse DCT. A running sum of spatial components will provide synthesis of image. Compare to other complex frequency processing analysis and synthesis techniques, the procedure is simple and provides best results. The decomposed components can be used effectively for noise removal as well as image compression at very low computational cost. This system is useful to analyze images for variety of applications including electronic signal communication and image processing for data compression, noise removal, and image analysis for pattern recognition. In neuroscience human visual system response is analyzed in response to Fourier sinusoidal gratings.