作者: Wang Xuewei , Liu Songtao , Zhou Xiaodong
DOI: 10.1016/S1004-4132(06)60046-7
关键词: Wavelet packet decomposition 、 Artificial intelligence 、 Computer vision 、 Second-generation wavelet transform 、 Wavelet 、 Pattern recognition 、 Mathematics 、 Fast wavelet transform 、 Algorithm 、 Discrete wavelet transform 、 Stationary wavelet transform 、 Wavelet transform 、 Cascade algorithm
摘要: The key to the wavelet based denoising techniques is how manipulate coefficients. By referring idea of Inclusive-OR in design circuits, this paper proposes a new algorithm called domain (WDIDA), which distinguishes coefficients belonging image or noise by considering their phases and modulus maxima simultaneously. Using algorithm, effects are improved computation time reduced. Furthermore, order enhance edges but not magnify noise, contrast nonlinear enhancing presented according human visual properties. Compared with traditional algorithms, that we proposed has better reducing performance, preserving improving quality images.