【摘要】 介绍了一类新的变换基bandelet,它在多尺度图像分解的基础上,自适应地提取图像的几何流。几何流表明了图像灰色值变化的正则性。同时描述了第二代离散bandelet基的构建以及它的应用算法。Bandelet变换具有正交性,其对应的基函数是正则的。对几何正则图像,它的几何流是最优的,通过bandelet变换可产生最优渐进衰减误差,所以在图像的压缩和除噪等方面,具有广泛的应用前景。通过试验对bandelet变换和小波变换进行了比较。更多还原

【Abstract】 This paper introduces a new class of bases, bandelet basis, which decompose the image along multiscale vectors that are adaptive in the direction of a geometric flow. This geometric flow indicates directions in which the image gray levels have regular variations. This paper also describes the construction of second discrete bandelet bases and their application and algorithms to image geometry compression. This new coding scheme is orthogonal and the corresponding basis function are regular. For regular geometric image, the geometric flow is optimized, so bandelet bases lead to error decay that is asymptotically optimal for geometrically regular images. Bandelets have broad application for image compression and noise removal. Comparisons are made with wavelet transform.更多还原