Method for detecting image sequences having linewise repeated data
View Patent ↗A method for detecting a video signal that represents an image sequence and exhibits linewise repeated video data comprises selecting a plurality of line groups each having a first, a second and a third line, which originate from two consecutive fields in the case of a field sequence and from one field in the case of a frame sequence. The method determines for each line group a first line-distance measure for a difference between video contents of the first and second lines and a second line-distance measure for a difference between video contents of the second and third lines. The method then interprets the first and second line-distance measures with regard to the presence of a significant difference between these line-distance measures.
1. A method for detecting a video signal that represents a field sequence having linewise structured fields and exhibits linewise repeated video data, the method comprising the following steps for each of a plurality of consecutive fields in a first grid position:
selecting a plurality of line groups each having a first and a third line disposed immediately consecutively in a specified image direction in a first field and each having a second line disposed at a position between the first and third lines in a temporally immediately adjacent field;
determining for each line group a first line-distance measure for a difference between video contents of the first and second lines and a second line-distance measure for a difference between video contents of the second and third lines; and
interpreting the first and second line-distance measures with regard to the presence of a significant difference between these line-distance measures, to provide an output signal indicative of the presence of a field sequence having linewise repeated video data.
2. The method of claim 1 , wherein first and second line-distance measures are determined for a plurality of line groups in two consecutive fields, wherein a first sum of first line-distance measures is formed and a second sum of second line-distance measures is formed and wherein the first sum and the second sum are interpreted with regard to the presence of a significant difference between these sums.
3. The method of claim 2 , wherein first and second sums of line-distance measures are determined for a plurality of consecutive groups each having two fields, wherein the presence of a significant difference is determined for each field group and wherein the presence of an image sequence having linewise repeated video data is inferred if there are alternately a significant difference and no significant difference.
4. The method of claim 1 , wherein a quotient of the first and second line-distance measures or of the sum of the first line-distance measures and the sum of the second line-distance measures is formed for each image and wherein a significant difference is present if this quotient lies above a first specified threshold value or below a second specified threshold value.
5. The method of claim 1 , wherein the determination of the first and second line-distance measures comprises:
selecting a plurality of groups of pixels each exhibiting a first pixel in the first line, a second pixel in the second line and a third pixel in the third line, each at identical positions within the lines;
calculating for each of the groups a measure for a difference between a video information value associated with the first pixel and a video information value associated with the second pixel and a second measure for a difference between the video information value associated with the second pixel and a video information value associated with the third pixel;
summing the measures for the differences between the first and second pixels of the groups, a sum obtained in this way representing the first measure; and
summing the measures for the differences between the second and third pixels of the groups, a sum obtained in this way representing the second measure.
6. The method of claim 5 , wherein calculating the measures for the differences of the video information values comprises forming an absolute value of these differences.
7. The method of claim 5 , wherein calculating the measures for the differences of the video information values comprises forming even-numbered powers of these differences.
8. The method of claim 5 , wherein low-pass filtering is applied to the video information values of the pixels of at least one of the first, second and third lines before the determination of the first and second measures.
9. The method of claim 4 , wherein the pixel differences are mapped onto modified pixel differences in accordance with a coring characteristic before addition.