【摘要】 Mallat塔式分解算法是一个被广泛应用的小波变换快速算法 ,但运算速度慢 ,实时性差的缺点阻碍了其在实际应用中的进一步发展。本文介绍的小波变换快速算法 ,结合系统的实时性要求 ,对 Mallat算法进行了优化 ,将小波变换与快速傅立叶变换有机结合 ,使图象的实时变换成为可能。在保留图象细节的基础上 ,该算法的运算量比 Mallat算法减少了一半以上 ,并在此基础上设计了该算法的相应硬件。实验结果表明该方法的确能快速有效地进行图象的分解和合成 ,具有较大的实际应用潜力。更多还原

【Abstract】 Mallat tower type decomposition algorithm is a fast algorithm extensively used for wavelet transform.But the slow operating speed and poor real time property hinder its further development in practical application.Mallat algorithm is optimized by a fast wavelet transform algorithm presented in the paper combining with the requirements of real time property and the organic combination of wavelet transform with fast Fourier transform will make the real time transform of image possible.Based on keeping details of the image,the operating quantity of the algorithm will be reduced by a half more than Mallat algorithm,and the corresponding hardware for the algorithm is designed on this basis.The experimental results show that both decomposition and composition of image can be effectively and quickly carried out with the method and it has a larger application perspective.更多还原