Apparatus for the flexible recognition of subject structures in color and picture half-tone images incorporates a recognition system that can recognize general subject structures in a video image given arbitrary orientation. The signal of an opto-electronic image converter is supplied via an analog-to-digital converter to a digital hardware image pre-processor in which the momentary brightness value of the image scanning is processed with a preceeding logic result with the assistance of control information, and the new result is again stored. The reduced image data read out from the result memory at the end of an image pass can be interpreted in a job-specific way in an image processor.

A circuit detects that a plurality of signals are generated in a predetermined sequence. The plurality of signals are applied to address terminals of a memory which has stored therein a predetermined pattern, and a divide-by-N counter (N:positive integer) counts a first data output signal from the memory N times and applies a carry output signal generated as a result thereof to another address terminal of the memory. An output signal of the circuit is derived from a second data output terminal of said memory when the plurality of input signals occur in the predetermined pattern of the memory and the carry signal from the counter is applied to the memory.

A pattern recognizer uses the method of n-tuples applied to a matrix containing (e.g.) features derived from input speech. A template store records during a training sequence the frequencies of occurrence of combinations of bits with each n-tuple group. During recognition frequencies associated with combinations which occur are used to form a measure of similarity.

Input values are digitally represented and the bits used to address memories. A composite (e.g. the sum) of the memory outputs is formed, and update control is used during "learning" to modify the memory contents. The digital representation is formed by an encoder which encodes input values using a redundant code, which preferably has a Hamming distance to signal distance relationship which has a relatively steep initial slope and is closer to being monotonic than is the relationship for a simply binary code.

A method and system for training a computer classification system which can be defined by a network of a number of n-tuples or Look Up Tables (LUTs), with each n-tuple or LUT including a number of rows corresponding to at least a subset of possible classes and further including a number of columns being addressed by signals or elements of sampled training input data examples, each column being defined by a vector having cells with values, wherein the column vector cell values are determined based on one or more training sets of input data examples for different classes so that at least part of the cells comprise or point to information based on the number of times the corresponding cell address is sample from one or more sets of training input examples, and weight cell values are determined, corresponding to one or more column vector cells being addressed or sampled by the training examples.