| 摘要 |
A method is disclosed of encoding location information pertaining to a source position signal characterized by positions of a set of selected coefficients, ordered in at least one dimension, there being a generally expected pattern of selected coefficients in the position signal with respect to the type of source signal. The method includes the steps of: for a first component of the position signal, generating a signal encoding the locations of the selected coefficients by a method that has a relatively high efficiency with respect to the generally expected pattern of selected coefficients in the first component; and for a second component of the position signal, encoding the locations of the selected coefficients by a method that has a relatively high efficiency with respect to the generally expected pattern of selected coefficients in the second component. For instance, runlength encoding can be used for both components, with different codebooks being used for each. Alternatively, vector coding can be used for one component, with runlength encoding (either standard, or with multiple codebooks) being used for the other component. Both encoding and decoding methods are described, as well as apparatus to carry out the methods of encoding and decoding described. |
