Three approaches based on wavelet transform, Teager energy operator, anchor human visual properties are proposed to improve the unsharp-masking image enhancement. The wavelet transform is used for providing multi-scaled information, the Teager energy operator is used to solve the wideband noise problems, and human visual properties is used to provide automatically dynamic thresholding for enhancement. In experiments, two other methods were also implemented for comparison. Based on the evaluation of visual inspection and quantitative measurement, the proposed approaches have shown to produce better results than the other methods produced.
|Number of pages||4|
|Journal||Proceedings - International Conference on Pattern Recognition|
|State||Published - 2002|