Abstract: In a system for the line-wise compression of binary data of a picture field, a first series of data bits of a current picture line and a second series of data bits of the directly preceding picture line are shifted together into a compression translator. A word of a line memory is read in a compression translator in order to obtain the data of the preceding line. At the same time the same word is written in to buffer the data of the current line. Compression bits are formed from the preceding and current line data. At the output of the compression translator the serial compression bits are regrouped in order to form words. Exclusive groups of significant compression bits are thus formed. During decompression a code word is read and a series of data bits is emulated and encoded as if it were the code word read. When the code word thus found corresponds to the code word read, a next code word is read and possibly a change of value for the data bits to be subsequently emulated is introduced.

Abstract: A simple and fast arithmetic member for multi-digit numbers. For each digit a module is provided which receives the digits of corresponding significance and which first forms two output carry signals therefrom, i.e. one as if the relevant module always receives an input carry signal (E) and one as if this module never receives an input carry signal (D). Between successive modules each time functionally identical configurations of logic elements are connected which, under the control of the input carry signal to the module (Ci) of next-lower significance, the two output carry signals (Di, Ei) thereof, and an enable signal, forms the input carry signal for the stage of next-higher significance (C(i+1)) in accordance with the formulaC (i + 1) = Di + Ei.multidot.Ci.