Image Compression Using Discrete Weyl-Heisenberg Transform

This article proposes a new approach to raster image compression, based on the use of the two-dimensional real discrete Weyl-Heisenberg transform (DWHT). This discrete transform is orthogonal and is based on the optimal Weyl-Heisenberg signal basis, which has the best time-frequency localization. The indicated properties are ensured by choosing the optimal forming function of the basis and the best ratio of its parameters. In addition, to assess the potential possibilities of using the discrete Weyl-Heisenberg transform in compression problems, the main criteria for compression efficiency were formulated and DWHT was compared with other well-known orthogonal transforms – discrete cosine transform (DCT) and discrete Hartley transform (DHT). It is experimentally shown that the proposed method based on discrete Weyl-Heisenberg transform has much better compression characteristics. The paper also presents the results of comparing three compression methods (DHT, DCT and DWHT) in the form of corresponding tables and figures of the restored images.