Robert Boschgmbh

Peter Wagner

Jurgen Heitmann

Rolf Loos

Digital video signals are subjected to band limiting by a high pass filter with a pass band terminating somewhat above half of the digital sampling frequency. The filtered signal is in effect subtracted from the original signal in an addition circuit with suitably phased inputs in order to make available a substitute signal of the interpolation low pass filtered kind. A circuit for determining the presence of an erroneous digital word accepts the high pass filtered signal, differentiates it twice and then digitally rectifies it in order to detect a possible maximum at half the sampling rate. When such a maximum is found, a changeover switch is caused to substitute the interpolation-low-pass-filtered signal for the orginal signal. Only a single digital filter is necessary for this operation. The high pass filter is a FIR filter of, for example, eleventh order and the middle coefficient is in effect changed from minus 1 to 0 when the delay equalized original signal is added to the output of the high pass filter.

Recognition and compensation of errors in a digital video signal

Two logic circuits (4A, 4B) are utilized in alternation for controlling the processing of successive pixels that are detected to be erroneous. Each logic circuit has its own counter 5A and 5B for associating the undisturbed pixels with ordinal numbers designating their spacing from the disturbed pixel under processing and likewise has a separate PROM (6A, 6B) and a separate multiplier for weighting the undisturbed pixel values, as well as an accumulator (10A, 10B) for summing the weighted values to produce a substitute value, so that undisturbed pixels lying between two disturbed pixels can all be used, where necessary, both for compensating for a preceding disturbed pixel and a following disturbed pixel. When a following disturbed pixel is detected while a preceding disturbed pixel is being processed by a logic circuit, that logic circuit provides a special reset to its counter and accumulator and selects another PROM program in accordance with the shortening of the 'filter length' by the nearness of the two disturbed pixels. When no defective pixels appear on the counters, multipliers and accumulators do not need to operate.

Signal error concealment circuit and method

A conventional multiple recognition unit provides a multiple error signal whenever digitally coded television signals supplied in successive data blocks to the error recognition unit finds by reference to check-words that the error correction capabilities of the unit are exceeded in any data block. In order to determine when such multiple errors are related to faulty tracking, the multiple error signals are prolonged in duration, overlaps between the multiple error signal and a preceding prolonged signal are detected, and the latter are then integrated. The level of the integrated signal determines the likelihood or presence of tracking error.

Growing community of inventors

Newark, NJ, United States of America

Peter Wagner