Recognition and compensation of errors in a digital video signal

Claim 2

Method of recognizing and compensating for errors in a digital video signal composed of pixel-value digital words by replacement of a pixel value word recognized as erroneous with a substitute word derived by interpolation from values of neighboring pixels, comprising the steps of: high-pass filtering an original digital video signal, which has been produced by digitally sampling an analog video signal in a converter at a predetermined sampling rate, in a digital filter having half said sampling rate within its passband and having its cut-off at a frequency lower than said sampling rate and thereby producing a high-pass filtered signal; detecting energy maxima of said high-pass filtered signal and providing an error recognition signal in response to each of said maxima; at least whenever said error recognition signal is present, subtracting said high-pass filtered video signal from the original digital video signal to produce a low-pass filtered video signal providing an interpolation-derived substitute digital word corresponding to a word of said original digital signal, and in response to each said error recognition signal substituting said substitute digital word for said digital word of said original digital signal which is related to said error recognition signal.

Claim 3

Method according to claim 1, wherein said digital filter is a transversal filter of odd order having a middle filter coefficient and equipped with means for setting the middle coefficient thereof at minus 1 to produce said subtracting step in response to the presence of said error recognition signal, wherein the combination of said high-pass filtering and addition of the input of said digital filter equalized for propagation time to the negatived product of said middle coefficient produces the behavior of an interpolation low-pass filter having a middle coefficient which is effectively set to zero.

Claim 4

Method according to claim 2, wherein the step of detecting said energy maxima is performed by detecting amplitude maxima of said high-pass filtered signal occurring substantially at half of said sampling rate.

Claim 5

Apparatus for recognizing and compensating for errors in a digital video signal composed of pixel value digital words by replacement of a pixel value word recognized as erroneous with a substitute word derived by interpolation from values of neighboring pixels, said digital video signal having a predetermined sampling rate, comprising: high-pass transversal digital filtering means connected to a source of said digital video signal for passing, to an output of said filtering means, components of said digital video signal in the neighboring of half of said predetermined sampling rate and for blocking from said output components of said video signals in the neighborhood of said predetermined sampling rate; means connected to said output of said high-pass filtering means for detecting energy maxima of the output signals of said filtering means and for providing an error recognition signal in response to each of said maxima; means for subtracting the output signal of said high-pass filtering means from a delay-equalized replica of said digital video signal provided as an input to said filtering means, in order to produce a low-pass filtered video signal made up of interpolation-derived substitute digital words, and selector switch means operable in response to each said error recognition signal provided by said detecting means for substituting one of said interpolation-derived substitute words produced by said subtracting means for a digital word of said digital video signal provided as an input for said high-pass filtering means which word is related to the same said error recognition signal.

Claim 6

Apparatus for recognizing and compensating for errors in a digital video signal composed of pixel value digital words by replacement of a pixel value word recognized as erroneous with a substitute word derived by interpolation from values of neighboring pixels, said digital video signal having a predetermined sampling rate, comprising: high-pass transversal digital filtering means connected to a source of said digital video signal for passing, to an output of said filtering means, components of said digital video signal in the neighboring of half of said predetermined sampling rate and for blocking from said output components of said video signals in the neighborhood of said predetermined sampling rate; means connected to said output of said high-pass filtering means for detecting energy maxima of the output signals of said filtering means and for providing an error recognition signal in response to each of said maxima; means for subtracting the output signal of said high-pass filtering means from a delay-equalized replica of said digital video signal provided as an input to said filtering means, in order to produce a low-pass filtered video signal made up of interpolation-derived substitute digital words, and selector switch means operable in response to each said error recognition signal provided by said detecting means for substituting one of said interpolation-derived substitute words produced by said subtracting means for a digital word of said digital video signal provided as an input for said high-pass filtering means which word is related to the same said error recognition signal.

Claim 7

Apparatus according to claim 4, wherein said high-pass transversal digital filter is of uneven order and has a middle filter coefficient of minus 1 at least during the presence of said error recognition signal produced by said detecting means and wherein said subtracting means is constituted by addition means which, by virtue of said minus 1 middle coefficient of said filter, provides a subtraction yielding said interpolation-derived substitute digital word for said digital word of said digital video signal supplied as said input to said filtering means.