A novel method for image enhancement, which exploits the properties of the double-density dual-tree DWT and the statistical characters of visual representation, is proposed. The double-density dual-tree DWT combines the characteristics and advantages of the double-density DWT and the complex dual-tree DWT. This allows the transform to provide approximate shift invariance and directional selectivity and is well suited for image enhancement. The algorithm corrects the global dynamic range (brightness) and adjusts local contrast in the new complex-directional wavelet transform domain. Comparisons with some popular image enhancement methods are made, showing the effectiveness of the proposed algorithm.