Edge extraction of images by reconstruction using wavelet decomposition details at different resolution levels

This paper describes a novel method for edge feature detection of document images based on wavelet decomposition and reconstruction. By applying the w avelet decomposition technique, a document image becomes a wavelet represen tation, i.e. the image is decomposed into a set of wavelet approximation co efficients and wavelet detail coefficients. Discarding wavelet approximatio n, the edge extraction is implemented by means of the wavelet reconstructio n technique. In consideration of the mutual frequency, overlapping will occ ur between wavelet approximation and wavelet details, a multiresolution-edg e extraction with respect to an iterative reconstruction procedure is devel oped to ameliorate the quality of the reconstructed edges in this case. A n ovel combination of this multiresolution-edge results in clear final edges of the document images. This multiresolution reconstruction procedure follo ws a coarser-to-finer searching strategy. The edge feature extraction is ac companied by an energy distribution estimation from which the levels of wav elet decomposition are adaptively controlled. Compared with the scheme of w avelet transform, our method does not incur any redundant operation. Theref ore, the computational time and the memory requirement are less than those in wavelet transform.