| 摘要 |
In an embodiment, an encoder (10) analyzes a range of values present in input data (D1) to determine if the range of values in the data is less than the full range allowed by the number of bits representing the data. If so, an offset value is applied to the data, which is then encoded by a form of differential encoding using the differences between successive values in the data, and where a wrap around a maximum value is used to avoid negative differences. The offset value used is sent with the compressed data to the decoder (20). In another embodiment, an input sequence of ones and zeroes is coded by output of a zero to represent no change in successive bits, and a one to represent a change in successive bits. The input sequence may be subdivided into sections depending on the sequence content. The output sequence may be further compressed using e.g. run-length encoding. |
