Directional Lifting-Based Wavelet Transform for Multiple Description Image Coding with Quincunx Segmentation