Wavelet transform and improved normashrink adaptiv
4.0
Application background For noisy image edge extraction, image edge extraction and denoising combined, put forward a kind of improved Normalshrink adaptive threshold to noise algorithm. The algorithm firstly by three B spline wavelet transform and local modulus maxima method to extract may contain characteristics of the edge of the image wavelet coefficients, considering the edge pixels between the special spatial relationships and noise at all levels of wavelet decomposition scales under different effect, build for each scale level improved Normalshrink adaptive threshold, and accordingly to extract the wavelet coefficients of screening.