Pulse or digital communications – Bandwidth reduction or expansion – Television or motion video signal
Reexamination Certificate
1998-06-04
2001-10-02
Rao, Andy (Department: 2713)
Pulse or digital communications
Bandwidth reduction or expansion
Television or motion video signal
C375S240260
Reexamination Certificate
active
06298090
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention is directed to a system which detects redundant images in a video sequence. In particular, the invention is directed to a system which determines whether a target image in the video sequence is a redundant image by comparing two predetermined threshold values to sums of differences between pixels in the target image and corresponding pixels in at least one other image in the video sequence.
2. Description of the Related Art
Telecines are well known devices that are used to generate video data from film images. These devices are commonly used in the television industry to convert motion pictures into video sequences for display on television. Problems arise from this conversion process, however, due to inherent differences between film and video.
More specifically, film sequences, such as motion pictures, have a frame rate of 24 frames-per-second. In contrast, video sequences have frame rates which are higher than 24 frames-per-second. For example, video sequences comprised of interlaced video images, i.e., video images in which two different fields are interlaced in one frame, have a frame rate of 30 frames-per-second. Video sequences comprised of progressive video images, on the other hand, have an even higher frame rate of 60 frames-per-second. In order to compensate for these different frame rates, and still maintain correspondence between the film sequence and the video sequence, telecine devices introduce additional, redundant images (i.e., fields and/or frames) into the video sequence. These redundant images essentially comprise repeats of earlier images of the video sequence, which are inserted at appropriate points in the video sequence.
Early on, it was recognized that there is no need to code these redundant images, since such images could merely be recreated at a decoder by copying other images of the video sequence. Accordingly, it became standard practice to include, in an encoder, circuitry for detecting redundant images. Conventional circuitry of this type determines redundancy by comparing pixels in an entire field and/or frame of a video sequence to corresponding pixels in another field and/or frame of a video sequence (see, e.g., U.S. Pat. Nos. 5,452,011, 5,406,333, 5,491,516, and 5,426,464). Specifically, conventional circuitry compares sums of differences in the foregoing pixels to a threshold value and, based on this comparison, ascertains whether the content of a target image is the same as the content of another image in the video sequence, usually an immediately preceding image.
To account for noise introduced by a telecine device, redundancy detection circuitry of the type described above sets its threshold high for high-noise telecine devices and low for low-noise telecine devices. However, using the same threshold to account for both variations in image content and telecine noise can lead to errors, particularly in cases where only small localized regions of two images vary in content. That is, in such cases, conventional circuitry often interprets these variations in content as telecine noise, thereby leading to an erroneous redundancy determination.
Thus, there exists a need for a redundant image detection circuit, for use in an encoding system or the like, which is able accurately to detect both telecine noise and variations in image content.
SUMMARY OF THE INVENTION
The present invention addresses the foregoing need by providing an apparatus which detects redundant images in a video sequence. In particular, the invention determines whether a target image in the video sequence is a redundant image by comparing two threshold values to sums of differences between pixels in the target image and corresponding pixels in at least one other image in the video sequence. Preferably, a first of the threshold values is set based on differences in content between the target image and the at least one other image, and a second of the threshold values is set based on noise introduced into the video sequence by a telecine device used to generate the video sequence. As a result, the invention is able to detect both telecine noise and variations in image content more accurately than it; conventional counterparts described above.
Thus, according to one aspect, the present invention is an apparatus which detects redundant images in a video sequence. The apparatus includes a processor and a memory which stores computer executable process steps. The processor executes the process steps stored in the memory so as (i) to divide a target image in the video sequence into a plurality of regions, (ii) to determine, for each of the plurality of regions, a local value comprising a sum of differences between individual pixels of each region and corresponding pixels in at least one other image in the video sequence, (iii) to compare the local value for each of the plurality of regions to a first predetermined threshold, (iv) to determine a global value comprising a sum of differences between substantially all pixels in the target image and corresponding pixels in the at least one other image, (v) to compare the global value to a second predetermined threshold, and (vi) to indicate that the target image is redundant in a case that both (a) the local value does not exceed the first predetermined threshold for all regions, and (b) the global value does riot exceed the second predetermined threshold.
In preferred embodiments of the invention, the processor further determines a “type” of video image comprising the video sequence and selects the at least one other image in accordance with the determined type of the video image. For example, in a case that the processor determines that the type of the video image is 30 frames-per-second interlaced video, the processor selects a field of the video sequence that immediately precedes the target image. On the other hand, in a case that the type of the video image is 60 frames-per-second progressive video, the processor selects two frames of the video sequence that immediately precede the target image. By virtue of this feature, the invention can be used to detect redundant images in different types of video and, in some types of video, more than one redundant image.
According to another aspect, the present invention is a method of detecting redundant images in a video sequence. The method includes dividing a target image in the video sequence into a plurality of regions, and determining, for each of the plurality of regions, a local value comprising a sum of differences between individual pixels of each region and corresponding pixels in at least one other image in the video sequence. The method also includes comparing the local value for each of the plurality of regions to a first predetermined threshold, and determining a global value comprising a sum of differences between substantially all pixels in the target: image and corresponding pixels in the at least one other image. The global value is then compared to a second predetermined threshold, and an indication is output that the target image is redundant in a case that (i) the local value does not exceed the first predetermined threshold for all regions, and (ii) the global value does not exceed the second predetermined threshold.
According to still another aspect, the present invention is an encoding device for encoding video data in a video sequence which includes a target image and at least one other image. The encoding device includes a film mode detection circuit which determines whether the target image is a redundant image by comparing two threshold values to sums of differences between pixels in the target image and corresponding pixels in the at least one other image. In the invention, a first of the threshold values is set based on differences in content between the target image and the at least one other image, and a second of the threshold values is set based on noise introduced into the video sequence by a device used to generate the video sequence. The encoding device also includes coding circu
Challapali Kiran
Chen Yingwei
Gross Russell
Rao Andy
U.S. Philips Corporation
LandOfFree
System for detecting redundant images in a video sequence by... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with System for detecting redundant images in a video sequence by..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and System for detecting redundant images in a video sequence by... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2582353