摘要 |
A method for compressing digital image data with greyscale values comprises making at least three predictions of the greyscale value of each pixel and comparing the predicted values with the actual value of the pixel. The difference values between the respective predicted values and the actual value of each pixel is computed, and the median difference value of the three difference values is selected for bit coding. Bit coding is carried out by using variable length bit codes, the shortest length bit code being assigned to the most frequently occurring median difference value. The coded values of the median differences are bit packed, and in the case of a pixel where the absolute median difference value exceeds a predetermined value, the actual pixel value is bit packed. |