Television – Format conversion
Patent
1996-02-08
1998-09-01
Lee, Michael
Television
Format conversion
358451, H04N 701, H04N 1393
Patent
active
058017801
DESCRIPTION:
BRIEF SUMMARY
The invention relates to a method for converting a first picture composed of M pixels into a second picture composed of L pixels (M.epsilon., L.epsilon., M.noteq.L) which second picture has a content corresponding with the first picture and which is destined to be displayed with an electronically controlled display unit provided for a pixelwise display, wherein for each j.sup.th (0.ltoreq.j.ltoreq.M-1) pixel from the first picture a first pixel value x(j) is supplied by a picture generator and for each i.sup.th (0.ltoreq.i.ltoreq.L-1) pixel from the second picture a second pixel value y(i) is determined by a summation of k first pixel values which contributes to that i.sup.th second pixel value, and wherein the surface of the first respectively second image is divided into M respectively L picture segments wherein a k.sup.th respectively i.sup.th pixel segment represents the surface of a k.sup.th first respectively i.sup.th second pixel and the surface of the k.sup.th first and i.sup.th second pixel differ from one another; and wherein for determining the i.sup.th second pixel value y(i) the i.sup.th pixel segment is projected in the first picture on a location corresponding for that i.sup.th segment and as k first pixel values those k first picture segments are selected which are situated at least partially within the the surface generated by projection of that i.sub.th picture segment, and wherein said summation each k.sup.th first pixel value is weighed with a weight factor h(k, i) representing that part of the surface of the i.sup.th picture segment that is occupied by the k.sup.th picture segment.
Such a method is known from the U.S. Pat. No. 4,933,775. The picture is built up from pixels which are preferably linewise aligned. The first picture is generated by a picture generator which is for example formed by a memory. For each j.sup.th pixel from the first picture the picture generator supplies a first pixel value x(j). That pixel value is for example formed by a digital r-bits word (r.epsilon.) representing the content of the j.sup.th pixel. In order to determine an i.sup.th pixel from the second picture to be formed by means of a summation of first pixel value, weight factors h(k, i) are each time determined. The weight factors are determined by means of surface projection enabling to assign a dimension to the pixel. The weight factor thus gives an accurate image of the degree according to which the projected second pixel overlaps one or more first pixels. When a change of dimension of the pixel occurs upon conversion the weight factor will comprise that change.
A drawback of the known method is that the conversion does not sufficiently take into account non-linearities in the picture signal and the display unit. A transfer function is added to the picture signal in order to convert a video input voltage into luminance. That transfer function is non-linear, whilst in order to maintain the picture energy upon conversion, a linear behaviour is required. In the known method the conversion is realized on the pixels with a pixel value x(j) wherein the non-linear transfer function is involved. This can lead to intensity shifts in the second pixel value which on their turn will lead to a reduced picture quality of the conversed picture with respect to the original picture.
It is an object of the invention to provide a method and a device for converting pixel values wherein the non-linear behaviour of the transfer function applied to the pixel values is taken into account.
A method according to the invention is therefor characterized in that the surface of each i.sup.th picture segment is multiplied with a sharpness factor (w) before the i.sup.th picture segment is projected on the first picture, and in that before the second pixel value y(i) is determined a first transfer function component (T) is added to each of the k.sup.th first pixel values, and in that after determination of the second pixel value y(i) a second transfer function component (T') is added to the latter.
The multiplication with a sharp
REFERENCES:
patent: 4423520 (1983-12-01), Murayama et al.
patent: 4893319 (1990-01-01), Ziuchkovski
patent: 4933775 (1990-06-01), Shimura
patent: 5400051 (1995-03-01), Krueger, Jr.
Candry Patrick
Schaumont Jan
Barco
Lee Michael
LandOfFree
Method and a device for converting a picture does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method and a device for converting a picture, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method and a device for converting a picture will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-275265