THE RECONSTRUCTION OF SUB-NYQUIST SAMPLING PAL SIGNAL BY TWO DIMENSIONAL ALIASING DIGITAL FILTER
Zhili Wang · Journal of Tianjin University Science and Technology · 1983
Method for sampling and reconstructing of luminance and chrominance components of PAL composite video signals at 1.5-2times colour subcarrier frequency are described and evaluted with aid of two dimeneional Fourier and Walsh aliasing digital filters. The sampling structure is chosen to permit the best reconstruction of the image. This paper presents two kinds of two dimenional sub-Nyquist sampling structure for discrete digital image data which lead to inversely errors in the reconstructed image. The ideal two dimensional lowpass Fourier frequency and Walsh sequency aliasing digital filters which can be used to reconstruct the original image from sub-Nyquist sampling data are introduced. The two dimensional Walsh sequency aliasing digital filter has elimnated the block effect occuring in the normal Walsh sequency digital filter and the two dimensional Fourier frequency aliasing digital filter bas removed the Gibbs effect of the normal Fourier frequency digital filter. This paper gives a detailed description for the design of the ideal two dimensional freduency and sequency aliasing digital filters in 2×2, 4×4, 8×8, 16×16 and 32×32 matrices and shows experimental results for variable cases. These experimental results show that greater matrix has less error of the reconstruced image and there is no detectable subjective difference between the original and reconstruced image for the 8×8 matrix. Minor degradations are detected for some critical patterns in the 8×8 case.