Patent US6493463B1 successfully refreshed from Google Patents.
US6493463B1
Expired / Lapsed
US
⏱️ 0.0 years remaining
Segmentation tag cleanup using neighborhood tags
Assignee: Xerox Corp
•
Filing: 1999-09-09
•
Granted: 2002-12-10
•
Expires: 2019-09-09
Abstract
Abstract A system and method for processing segmentation tags is disclosed. The system comprises a neighborhood analysis module, transition analysis module, and a tag cleaning module. The neighborhood analysis module generates a neighborhood block tag prediction from a block of segmentation tags having a first predefined relationship with a current tag. The transition analysis module generate a transition block tag prediction from a block of segmentation tags having a second predefined relationship with the current tag. The tag cleaning module modifies selected segmentation tags based on the neighborhood block tag prediction and the transition block tag prediction. The method comprises identifying a current segmentation tag to be cleaned; generating a tag prediction from a plurality of segmentation tags having a predefined relationship to the current segmentation tag; and generating a cleaned segmentation tag for the current segmentation tag based upon the current segmentation tag and the tag prediction.
Inventors
- 👤 Ramesh Nagarajan
- 👤 Xing Li
- 👤 John F. Seward, Jr.
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 18:31
CPC Classifications
G; G06; G06V; G06V10/00; G06V10/40; G06V10/44; G06V10/457
Claims Analysis
Total Claims: 25
Claim 1
Independent Claim
135 words
A method for processing segmentation tags, comprising: identifying a current segmentation tag to be cleaned; identifying a neighborhood block comprising a first plurality of segmentation tags neighboring the current segmentation tag; determining a frequency of occurrence for selected segmentation tags within the neighborhood block; selecting a neighborhood block tag prediction based upon the frequency of occurrence; identifying a transition block comprising a second plurality of segmentation tags identifying a region into which the current segmentation tag is transitioning; determining a frequency of occurrence for selected segmentation tags within the transition block; selecting a transition block tag prediction based upon the frequency of occurrence; and generating a cleaned segmentation tag for the current segmentation tag based upon at least one of the current segmentation tag, the neighborhood block tag prediction and the transition block tag prediction.
Claim 2
Independent Claim
135 words
A method for processing segmentation tags, comprising: identifying a current segmentation tag to be cleaned; identifying a neighborhood block comprising a first plurality of segmentation tags neighboring the current segmentation tag; determining a frequency of occurrence for selected segmentation tags within the neighborhood block; selecting a neighborhood block tag prediction based upon the frequency of occurrence; identifying a transition block comprising a second plurality of segmentation tags identifying a region into which the current segmentation tag is transitioning; determining a frequency of occurrence for selected segmentation tags within the transition block; selecting a transition block tag prediction based upon the frequency of occurrence; and generating a cleaned segmentation tag for the current segmentation tag based upon at least one of the current segmentation tag, the neighborhood block tag prediction and the transition block tag prediction.
Claim 3
Dependent Claim
72 words
The method of claim 1 , wherein the step of generating a cleaned segmentation tag comprises: providing the current segmentation tag as the cleaned segmentation tag if the current segmentation tag is the same as one of the neighborhood block tag prediction and the transition block tag prediction; otherwise setting the cleaned segmentation tag as a function of the current tag, the neighborhood block tag prediction and the transition block tag prediction.
Claim 4
Dependent Claim
60 words
The method of claim 2 , wherein the step of setting the cleaned segmentation tag comprises: providing the current segmentation tag as the cleaned segmentation tag if the neighborhood block tag prediction has a value representing an image type of background and the frequency of occurrence within the neighborhood block of the current segmentation tag is greater than a threshold.
Claim 5
Dependent Claim
65 words
The method of claim 3 , wherein the step of setting the cleaned segmentation tag further comprises: setting the cleaned segmentation tag equal to the transition block tag prediction if the current segmentation tag is not the same as one of the neighborhood block tag prediction and the transition block tag prediction and the transition block tag prediction does not represent a predefined image type.
Claim 6
Dependent Claim
32 words
The method of claim 1 , wherein the step of selecting a transition block tag prediction selects the tag having the highest frequency of occurrence as the transition block the tag prediction.
Claim 7
Dependent Claim
36 words
The method of claim 1 , wherein the step of selecting a transition block the tag prediction selects all tags having a frequency of occurrence greater than a threshold as the transition block the tag prediction.
Claim 8
Dependent Claim
29 words
The method of claim 1 , wherein the current segmentation tag, the first plurality of segmentation tags and the second plurality of segmentation tags comprise block level segmentation tags.
Claim 9
Independent Claim
109 words
A method for processing segmentation tags, comprising: identifying a current segmentation tag to be cleaned; generating a tag prediction from a plurality of segmentation tags having a predefined relationship to the current segmentation tag; and generating a cleaned segmentation tag for the current segmentation tag based upon the current segmentation tag and the tag prediction; wherein the step of generating a tag prediction comprises identifying a transition block comprising a plurality of segmentation tags identifying a region into which the current segmentation tag is transitioning, determining a frequency of occurrence for selected segmentation tags within the transition block, and selecting the tag prediction based upon the frequency of occurrence.
Claim 10
Independent Claim
109 words
A method for processing segmentation tags, comprising: identifying a current segmentation tag to be cleaned; generating a tag prediction from a plurality of segmentation tags having a predefined relationship to the current segmentation tag; and generating a cleaned segmentation tag for the current segmentation tag based upon the current segmentation tag and the tag prediction; wherein the step of generating a tag prediction comprises identifying a transition block comprising a plurality of segmentation tags identifying a region into which the current segmentation tag is transitioning, determining a frequency of occurrence for selected segmentation tags within the transition block, and selecting the tag prediction based upon the frequency of occurrence.
Claim 11
Dependent Claim
37 words
The method of claim 8 , wherein the step of generating a cleaned segmentation tag provides the current segmentation tag as the cleaned segmentation tag if the current segmentation tag is the same as the tag prediction.
Claim 12
Dependent Claim
50 words
The method of claim 9 , further comprising: comparing the current segmentation tag to selected segmentation tags adjacent to the current segmentation tag; and providing the current segmentation tag as the cleaned segmentation tag if the current segmentation tag is the same as one of the selected segmentation tags adjacent.
Claim 13
Dependent Claim
54 words
The method of claim 8 , wherein the step of generating a cleaned segmentation tag provides the current segmentation tag as the cleaned segmentation tag if the tag prediction has a predefined value and the frequency of occurrence within the plurality of segmentation tags of the current segmentation tag is greater than a threshold.
Claim 14
Dependent Claim
80 words
The method of claim 8 , wherein the step of generating a tag prediction further comprises; identifying a neighborhood block comprising the plurality of segmentation tags, the neighborhood block having a predefined position with respect to the current segmentation tag, the neighborhood block including a plurality of segmentation tags neighboring the current segmentation tag; determining a frequency of occurrence for selected segmentation tags within the neighborhood block; and selecting a neighborhood block tag prediction based upon the frequency of occurrence.
Claim 15
Dependent Claim
30 words
The method of claim 12 , wherein the neighborhood block is a block of M×N, wherein and M and N are integer values, one of which is greater than 2.
Claim 16
Dependent Claim
17 words
The method of claim 13 , wherein the neighborhood block is centered on the current segmentation tag.
Claim 17
Dependent Claim
58 words
The method of claim 12 , wherein: the segmentation tags within the neighborhood block comprise segmentation tags that have been previously cleaned and segmentation tags that have not been cleaned; and the step of determining a frequency of occurrence weights the segmentation tags that have been previously cleaned higher than the segmentation tags that have not been cleaned.
Claim 18
Dependent Claim
33 words
The method of claim 15 , wherein the step of selecting a neighborhood block the tag prediction selects the tag having the highest frequency of occurrence as the neighborhood block the tag prediction.
Claim 19
Dependent Claim
36 words
The method of claim 15 , wherein the step of selecting a neighborhood block the tag prediction selects all tags having a frequency of occurrence greater than a threshold as the neighborhood block the tag prediction.
Claim 20
Dependent Claim
22 words
The method of claim 8 , wherein the current segmentation tag and the plurality of segmentation tags are block level segmentation tags.
Claim 21
Independent Claim
129 words
A system for processing segmentation tags, comprising: a neighborhood analysis module connected to receive a current segmentation tag and a plurality of neighboring segmentation tags and generate a neighborhood block tag prediction; a transition analysis module connected to receive a transition block comprising a plurality of segmentation tags identifying a region into which the current segmentation tag is transitioning, the transition analysis module determining a frequency of occurrence for selected segmentation tags within the transition block and generating a transition block tag prediction based upon the frequency of occurrence; and a tag cleaning module connected receive the neighborhood block tag prediction and the transition block tag prediction, the tag cleaning module modifying the current segmentation tag based on the neighborhood block tag prediction and the transition block tag prediction.
Claim 22
Independent Claim
129 words
A system for processing segmentation tags, comprising: a neighborhood analysis module connected to receive a current segmentation tag and a plurality of neighboring segmentation tags and generate a neighborhood block tag prediction; a transition analysis module connected to receive a transition block comprising a plurality of segmentation tags identifying a region into which the current segmentation tag is transitioning, the transition analysis module determining a frequency of occurrence for selected segmentation tags within the transition block and generating a transition block tag prediction based upon the frequency of occurrence; and a tag cleaning module connected receive the neighborhood block tag prediction and the transition block tag prediction, the tag cleaning module modifying the current segmentation tag based on the neighborhood block tag prediction and the transition block tag prediction.
Claim 23
Dependent Claim
37 words
The system of claim 19 , wherein the tag cleaning module modifies a segmentation tag if the segmentation tag is not the same as one of the neighborhood block tag prediction and the transition block tag prediction.
Claim 24
Dependent Claim
35 words
The system of claim 20 , wherein the tag cleaning module modifies the segmentation tag to equal the neighborhood block tag prediction if the neighborhood block tag prediction does not represent a predefined image type.
Claim 25
Dependent Claim
41 words
The system of claim 21 , wherein the tag cleaning module modifies the segmentation tag to equal the transition block tag prediction if the neighborhood block tag prediction represents the predefined image type and the transition block tag prediction does not.
References Cited (Backward Citations)
US5327262A
Cited
US5699453A
Cited
US5765029A
Cited
US5778156A
Cited
US5956468A
Cited
US5850474A
Cited
US6240205B1
Cited
US5767978A
Cited
US6337917B1
Cited
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2005-12-09 | 2006-06-10 | 2006-12-10 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2009-12-09 | 2010-06-10 | 2010-12-10 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2013-12-09 | 2014-06-10 | 2014-12-10 | $7700.00 | Window Closed / Lapsed |