Image compression using 2-D lattice filters

Four different two dimensional lattice filters are applied to image compression. These are: (1) the three parameter lattice filter, (2) the extended lattice filter of perpendicular form, (3) the extended lattice filter of diagonal form and (4) the further improved lattice filter. The last three of these filters are developed to compensate the insufficiency of the first one. Two of them introduce additional error fields for higher orders, and the last one assumes that the image data is half-plane symmetric rather than quarter-plane. Thus, these filters exploit more data. Different images are compressed with these filters and their performances are compared with each other. The results of the computer simulations show that the performances of the new filters are comparable to the conventional one. The further improved lattice filter produces somewhat better results. However with these filters, the computational complexity increases significantly.