Image analysis – Color image processing – Compression of color images
Reexamination Certificate
2000-03-29
2002-02-12
Rogers, Scott (Department: 2624)
Image analysis
Color image processing
Compression of color images
C382S243000, C382S239000, C375S240120, C375S240020, C375S240260, C348S391100, C348S396100, C348S404100, C348S425100
Reexamination Certificate
active
06347151
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The present invention relates to an apparatus for encoding and decoding an image signal used for efficient image signal transmission or storage, and to a method therefor.
2. Description of the Prior Art
A method of separating moving pictures into discrete layers each containing one of normally plural objects in a moving picture image at a specific time instance has been proposed to enable the efficient transmission and storage of image signals, and particularly moving picture (video) signals, in which object images are expressed as compositions (collections) of pels. For example, when an image comprising people and background is encoded with this method, the image coding apparatus separates the image into two layers, a people layer and a background layer, and separately codes and transmits each layer.
At the receiving end, the image decoding apparatus decodes the signal encoded for each layer, and then combines the images from each of the decoded layers using a specified method to present a single integrated picture. This method must therefore also provide for each pel in each layer information declaring whether a particular pel hides or does not hide the background image. The information thus used to declare whether a pel hides or does not hide the background image is called “significance information,” and pels that hide the background are called “significant.”
This significance information can also be used to declare the interpel correlation information in images that are recorded as a single layer rather than being segmented into plural layers. However, significant pels are pels that are included in a given object, and non-significant pels are pels that are located outside a given object. A high significance information value therefore means that the ratio of a given pel to the other pels at the same position is high and that pel is visually important. Conversely, a low significance signal value means that the corresponding pel has little influence on the appearance of the output pel, i.e., is nearly transparent.
The signal comprising the significance information for a specific group of pels in the image signal is called a significance signal. When plural image objects included on different layers of the image signal are overlaid to present an image, the significance signal can be used to declare whether a particular pel hides or does not hide the background. A non-zero significance signal value in this case means that that pel is significant and hides the background. There is no significance signal value for non-significant pels, however, and non-significant pels are thus transparent and are not needed for image synthesis. As a result, the significance signal describes the shape of objects in the image, and only significant pels affect the quality of the synthesized image.
In other words, non-significant pels are unrelated to image quality, and the coding efficiency can therefore be improved by encoding only the significant pels.
When encoding image signals containing luminance, color difference, transparency, and other pel information for each pel in the image, however, conventional image signal coding apparatuses perform the same frequency conversion operation during pel subsampling and pel interpolation processing irrespective of the pel significance. As a result, non-significant pel values, i.e., meaningless pel values, affect significant pels, resulting in image quality degradation.
While the resolution of the luminance signal component of the image signal is equal to the significance signal resolution, the resolution of the color difference may be different from the significance signal resolution. In such cases there are no problems encoding the luminance signal based on the significance signal. However, if the color difference signal is coded based on the significance signal, the resolution of both signals must be the same. This means that the resolution of the significance signal must be converted to the color difference signal resolution before the color difference signal is coded. As described above, however, resolution conversion of the significance signal applies the same resolution conversion to every pel regardless of whether or not each pel is significant, and the significance signal is thus degraded in the same way coding degrades the image signal above.
Coding efficiency also drops with the conventional image signal encoding apparatus and the image signal decoding apparatus because it is also necessary to encode the pel value of nearly-transparent pels because even these nearly-transparent pels having no visual effect on the output image are handled as significant pels due to coding errors or minor noise components in the input signal.
SUMMARY OF THE INVENTION
The object of the present invention is therefore to provide an imaging device which solves these problems.
The present invention has been developed with a view towards substantially solving the above described disadvantages and has for its essential object to provide an improved electrophotographic imaging device.
In order to achieve the aforementioned objective, an image signal encoding apparatus is provided for encoding an image signal comprising two or more proximal pels based on a significance signal comprising significance elements expressing the significance value of the pels, the apparatus comprises a resolution conversion characteristics selector for selecting the resolution conversion characteristic based on the significance value of the pels proximal to the pel being processed, and a resolution converter for converting the resolution of the image signal by applying the selected resolution conversion characteristic.
REFERENCES:
patent: 4654484 (1987-03-01), Reiffel et al.
patent: 4672444 (1987-06-01), Bergen et al.
patent: 4816901 (1989-03-01), Music et al.
patent: 4870497 (1989-09-01), Chamzas et al.
patent: 4914508 (1990-04-01), Music et al.
patent: 5274471 (1993-12-01), Kim
patent: 5363213 (1994-11-01), Coward et al.
patent: 5387938 (1995-02-01), Fukuda et al.
patent: 5414527 (1995-05-01), Koshi et al.
patent: 5418626 (1995-05-01), Semasa
patent: 5424854 (1995-06-01), Hashimoto
patent: 5426513 (1995-06-01), Scorse et al.
patent: 5644406 (1997-07-01), Harrington et al.
patent: 5696845 (1997-12-01), Loce et al.
patent: 5760921 (1998-06-01), Miyake
patent: 5802203 (1998-09-01), Black et al.
patent: 6067320 (2000-05-01), Takahashi et al.
patent: 6078694 (2000-06-01), Takahashi et al.
patent: 6292513 (2001-09-01), Takahashi et al.
patent: 0 748 110 (1996-12-01), None
Boon Choong Seng
Kadono Shinya
Takahashi Jun
Matsushita Electric - Industrial Co., Ltd.
Rogers Scott
Wenderoth , Lind & Ponack, L.L.P.
LandOfFree
Image signal converting/encoding method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image signal converting/encoding method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image signal converting/encoding method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2959899