Television – Image signal processing circuitry specific to television – Noise or undesired signal reduction
Patent
1993-03-29
1995-02-28
Groody, James J.
Television
Image signal processing circuitry specific to television
Noise or undesired signal reduction
348620, H04N 521, H04N 5213
Patent
active
053941926
DESCRIPTION:
BRIEF SUMMARY
SUMMARY OF THE INVENTION
1. Field of the Invention
The present invention relates to a method of measuring the noise in an active video image, the said method making it possible automatically to modify the value of the noise-reduction coefficient applied in the noise-reduction methods used in the processing of video images.
2. Discussion of the Background
Currently, in the processing of video images, various algorithms are used to bring about a noise reduction. These algorithms are divided into two main classes, namely the spatial algorithms which mainly include low-pass filtering, and the time domain algorithms which use frame or image storage. In the case where a time domain type noise-reduction algorithm is used for processing of a video signal, the latter can be implemented in a time domain recursive filter such as that represented in FIG. 1. This filter consists mainly of a multiplier 1 which receives at its input the video signal I and which multiplies it by a noise-reduction coefficient K. This noise-reduction coefficient K may take a value varying between 0 and 1. The output of the multiplier 1 is sent to a summer 2 which then adds a fraction of the video signal at the input to a signal coming from a multiplier 3. The multiplier 3 receives at its input the signal I' coming from an image memory 4 which in fact corresponds to the signal coming from the summer 2 delayed by one image. The multiplier 3 then multiplies this signal I' by a coefficient (1-K). The two coefficients applied to the multipliers 1 and 3 have a sum equal to 1 so as to keep a constant direct-current gain through the filter. The output of the summer 2 also makes it possible to obtain the video signal at the output S. When this filter is examined, it is seen that, if the noise-reduction coefficient K is equal to 1, the multiplier coefficient of the multiplier 3 is equal to 0, and, consequently, there is no feedback and no noise reduction. If the noise-reduction coefficient K is equal to 0, there is no input and a "frozen" image recirculates in the circuit, with a delay of one image. In fact, a noise reduction is brought about when K has a value lying between 0and 1, i.e. when a fraction of the input signal is added to the signal recirculating through the circuit with a delay of one image. However, such a system, when it is used to reduce the noise in a television signal, exhibits a certain number of drawbacks. The main drawback resides in the fact that the moving parts of the image appear with a streaking phenomenon. In fact, for a small value of K giving rise to good reduction of the noise, the filter has a long pulse response and, consequently, the information in the filter needs several images to decrease.
In fact, if the only image degradations introduced by such a filter are produced when there is a movement in the signal, it is desirable to introduce a mechanism making it possible to detect the movement and to use it to prevent a detection of movement on the moving parts of the image. Such a system has been proposed and is represented, for example, in FIG. 2. This device therefore mainly comprises a subtracter 10 which receives the video signal I at its input and which substracts it from the video signal I' coming from an image memory 13, this video signal I' being the delayed signal corresponding to the previous processed image. The output of the subtractor 10 is sent to a multiplier 11 in order to be multiplied by the noise-reduction coefficient K coming from a movement detector 14. The output of the multiplier 11 is sent to a summer 12 which receives the video signal I' coming from the image memory 13 on its other input. The output of the summer 12 is sent to the image memory 12 which delays the video signal by one processed image. On the other hand, the video signal S at the output comes from the output of the summer 12. As represented in FIG. 2, the detector 14 supplies the noise reduction coefficient K which it calculates by using both the input video signal I and the video signal I' processed and delayed by one image.
T
REFERENCES:
patent: 4249209 (1981-02-01), Storey
patent: 4296436 (1981-10-01), Achilla
patent: 4646138 (1987-02-01), Willis
patent: 4679086 (1987-07-01), May
patent: 5025316 (1991-06-01), Darby
patent: 5140424 (1992-08-01), Yoshimura et al.
patent: 5161018 (1992-11-01), Matsunaga
patent: 5185664 (1993-02-01), Darby
Groody James J.
Lee Michael H.
Thomson Consumer Electronics
LandOfFree
Method of measuring the noise in an active video image and devic 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 of measuring the noise in an active video image and devic, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of measuring the noise in an active video image and devic will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-851591