US5825909A
Expired / Lapsed
US
⏱️ 0.0 years remaining
Automated method and system for image segmentation in digital radiographic images
Assignee: Carestream Health Inc
•
Filing: 1996-02-29
•
Granted: 1998-10-20
•
Expires: 2016-02-29
Abstract
Abstract An image segmentation method for segmenting various regions of digital radiographic images first uses progressively smoothing techniques to generate smoothed images at multiple scales (resolutions). Next, texture measures are computed at each scale for each pixel in the smoothed image. Then a clustering algorithm is applied to the feature space spanned by the vector of texture measures. A segment spectrum, which is the number of clusters as a function of scale, is subsequently constructed to determine the most stable range of scales and the most stable number of regions into which the image is best segmented. Each pixel in the image is then classified into one of the detected regions according to its belonged cluster in the feature space. Region filling is then performed to eliminate small holes in each segmented region. Finally, a segmented map is generated to function as multivalued templates for any further image processing to be done on various segmented regions of the image.
Inventors
- 👤 Ben Kwei Jang
Bibliographic Details
Application Number
—
Jurisdiction
US
Data Source
google_patents
Last Synchronized
2026-11-09 10:03
CPC Classifications
G; G06; G06T; G06T7/00; G06T7/40; G06T7/41; G06T7/44; G06T7/10; G06T7/11; G06T2207/00; G06T2207/10; G06T2207/10072; G06T2207/10088; G06T2207/10116; G06T2207/20; G06T2207/20081
Claims Analysis
Total Claims: 40
Claim 1
Independent Claim
108 words
A method for segmenting a digital radiographic image into a plurality of regions in a spatial domain, said method comprising the steps of: (a) smoothing the digital radiographic image at a plurality of scales to thereby generate a plurality of smoothed images; (b) computing two or more texture measures for each pixel from each smoothed image at each scale; (c) at each scale, clustering the texture measures in a feature space; (d) determining the most stable number of clusters in the feature space; (e) classifying each pixel in the spatial domain based on the determined number of clusters; and (f) generating a segmented map for further image processing.
Claim 2
Independent Claim
108 words
A method for segmenting a digital radiographic image into a plurality of regions in a spatial domain, said method comprising the steps of: (a) smoothing the digital radiographic image at a plurality of scales to thereby generate a plurality of smoothed images; (b) computing two or more texture measures for each pixel from each smoothed image at each scale; (c) at each scale, clustering the texture measures in a feature space; (d) determining the most stable number of clusters in the feature space; (e) classifying each pixel in the spatial domain based on the determined number of clusters; and (f) generating a segmented map for further image processing.
Claim 3
Dependent Claim
18 words
The method of claim 1 wherein the regions of the radiographic image are physically or anatomically meaningful regions.
Claim 4
Dependent Claim
26 words
The method of claim 1 wherein the step of smoothing a digital image at a plurality of scales includes the use of linear progressively smoothing techniques.
Claim 5
Dependent Claim
21 words
The method of claim 3 wherein the use of a linear smoothing technique includes the use of a Gaussian smoothing kernel.
Claim 6
Dependent Claim
22 words
The method of claim 3 wherein the use of a linear smoothing technique includes the use of a Mexican hat smoothing kernel.
Claim 7
Dependent Claim
20 words
The method of claim 5 wherein the use of a Mexican hat smoothing kernel is related to a wavelet transform.
Claim 8
Dependent Claim
26 words
The method of claim 1 wherein the step of smoothing a digital image at a plurality of scales includes the use of nonlinear progressively smoothing techniques.
Claim 9
Dependent Claim
31 words
The method of claim 1 wherein a texture vector comprises a set of texture measures that are selected to provide information for segmenting the plurality of regions of the radiographic image.
Claim 10
Dependent Claim
24 words
The method of claim 1 wherein the feature space is constituted of an n-dimensional space in which each point represents an n-element texture vector.
Claim 11
Dependent Claim
28 words
The method of claim 9 further including the step of determining the exam type of the input image, wherein the dimensionality of the feature space is thereby reduced.
Claim 12
Dependent Claim
33 words
The method of claim 1 wherein the step of clustering in the feature space includes the use of an unsupervised algorithm which does not require specification of the number of clusters a priori.
Claim 13
Dependent Claim
39 words
The method of claim 1 wherein the step of determining the most stable number of clusters includes the construction of a segment spectrum according to the number of clusters computed in the feature space as a function of scale.
Claim 14
Dependent Claim
26 words
The method of claim 12 wherein the most stable number of clusters in the feature space is determined by the continuum of the broadest constant spectrum.
Claim 15
Dependent Claim
21 words
The method of claim 12 wherein a stable range of scales is determined by the range of the broadest constant spectrum.
Claim 16
Dependent Claim
28 words
The method of claim 12 wherein the step of classifying each pixel in the spatial domain includes the use of the lower bound of the broadest constant spectrum.
Claim 17
Dependent Claim
40 words
The method of claim 1 wherein the step of generating a segmented map includes the use of a multiple-valued image that functions as multiple templates for any further image processing to be done on the segmented regions of the image.
Claim 18
Independent Claim
143 words
A method for segmenting a digital radiographic image into a plurality of regions in a spatial domain, said method comprising the steps of: (a) smoothing the digital radiographic image at a plurality of scales to generate a plurality of smoothed images; (b) at each scale, computing a set of texture measures chacterized by a texture vector for each pixel in each smoothed image; (c) at each scale, clustering the texture vectors in a feature space; (d) generating a segment spectrum relating the number of clusters to the scale; (e) analyzing the segment spectrum to determine the most stable range of scales and the most stable number of clusters in the feature space; (e) classifying each pixel in the spatial domain based on the determined range of scales and the determined number of clusters; and (f) generating a segmented map for further image processing.
Claim 19
Independent Claim
143 words
A method for segmenting a digital radiographic image into a plurality of regions in a spatial domain, said method comprising the steps of: (a) smoothing the digital radiographic image at a plurality of scales to generate a plurality of smoothed images; (b) at each scale, computing a set of texture measures chacterized by a texture vector for each pixel in each smoothed image; (c) at each scale, clustering the texture vectors in a feature space; (d) generating a segment spectrum relating the number of clusters to the scale; (e) analyzing the segment spectrum to determine the most stable range of scales and the most stable number of clusters in the feature space; (e) classifying each pixel in the spatial domain based on the determined range of scales and the determined number of clusters; and (f) generating a segmented map for further image processing.
Claim 20
Dependent Claim
24 words
The method as claimed in claim 17 further including the step of eliminating small holes in each segmented region before generating a segmented map.
Claim 21
Dependent Claim
26 words
The method of claim 17 wherein the step of smoothing a digital image at a plurality of scales includes the use of linear progressively smoothing techniques.
Claim 22
Dependent Claim
21 words
The method of claim 19 wherein the use of a linear smoothing technique includes the use of a Gaussian smoothing kernel.
Claim 23
Dependent Claim
22 words
The method of claim 19 wherein the use of a linear smoothing technique includes the use of a Mexican hat smoothing kernel.
Claim 24
Dependent Claim
20 words
The method of claim 21 wherein the use of a Mexican hat smoothing kernel is related to a wavelet transform.
Claim 25
Dependent Claim
26 words
The method of claim 17 wherein the step of smoothing a digital image at a plurality of scales includes the use of nonlinear progressively smoothing techniques.
Claim 26
Dependent Claim
31 words
The method of claim 17 wherein a texture vector comprises a set of texture measures that are selected to provide information for segmenting the plurality of regions of the radiographic image.
Claim 27
Dependent Claim
24 words
The method of claim 17 wherein the feature space is constituted of an n-dimensional space in which each point represents an n-element texture vector.
Claim 28
Dependent Claim
33 words
The method of claim 17 wherein the step of clustering in the feature space includes the use of an unsupervised algorithm which does not require specification of the number of clusters a priori.
Claim 29
Dependent Claim
26 words
The method of claim 17 wherein the most stable number of clusters in the feature space is determined by the continuum of the broadest constant spectrum.
Claim 30
Dependent Claim
22 words
The method of claim 17 wherein the most stable range of scales is determined by the range of the broadest constant spectrum.
Claim 31
Dependent Claim
28 words
The method of claim 27 wherein the step of classifying each pixel in the spatial domain includes the use of the lower bound of the broadest constant spectrum.
Claim 32
Independent Claim
118 words
A system for segmenting a digital radiographic image into a plurality of regions in a spatial domain, said system comprising: (a) means for smoothing the digital radiographic image at a plurality of scales to thereby generate a plurality of smoothed images; (b) means for computing two or more texture measures for each pixel from each smoothed image at each scale; (c) means, operative at each scale, for clustering the texture measures in a feature space; (d) means for determining the most stable number of clusters in the feature space; (e) means for classifying each pixel in the spatial domain based on the determined number of clusters; and (f) means for generating a segmented map for further image processing.
Claim 33
Independent Claim
118 words
A system for segmenting a digital radiographic image into a plurality of regions in a spatial domain, said system comprising: (a) means for smoothing the digital radiographic image at a plurality of scales to thereby generate a plurality of smoothed images; (b) means for computing two or more texture measures for each pixel from each smoothed image at each scale; (c) means, operative at each scale, for clustering the texture measures in a feature space; (d) means for determining the most stable number of clusters in the feature space; (e) means for classifying each pixel in the spatial domain based on the determined number of clusters; and (f) means for generating a segmented map for further image processing.
Claim 34
Dependent Claim
26 words
The system of claim 30 wherein said means for smoothing a digital image at a plurality of scales includes the use of linear progressively smoothing techniques.
Claim 35
Dependent Claim
26 words
The system of claim 30 wherein said means for smoothing a digital image at a plurality of scales includes the use of nonlinear progressively smoothing techniques.
Claim 36
Dependent Claim
33 words
The system of claim 30 wherein said means for clustering in the feature space includes the use of an unsupervised algorithm which does not require specification of the number of clusters a priori.
Claim 37
Dependent Claim
37 words
The system of claim 30 wherein said means for determining the most stable number of clusters includes means for generating a segment spectrum describing the number of clusters in the feature space as a function of scale.
Claim 38
Dependent Claim
26 words
The system of claim 34 wherein the most stable number of clusters in the feature space is determined by the continuum of the broadest constant spectrum.
Claim 39
Dependent Claim
13 words
The system of claim 30 wherein the radiographic image is a subsampled image.
Claim 40
Dependent Claim
13 words
The system of claim 30 wherein the radiographic image is a non-subsampled image.
References Cited (Backward Citations)
USPTO Annuity & Maintenance Windows
| Annuity Stage | Window Opens | Payment Deadline | Surcharge Period Until | Estimated Fee | Status |
|---|---|---|---|---|---|
| 3.5 Years (Year 4) | 2001-10-19 | 2002-04-20 | 2002-10-20 | $2000.00 | Window Closed / Lapsed |
| 7.5 Years (Year 8) | 2005-10-19 | 2006-04-20 | 2006-10-20 | $3760.00 | Window Closed / Lapsed |
| 11.5 Years (Year 12) | 2009-10-19 | 2010-04-20 | 2010-10-20 | $7700.00 | Window Closed / Lapsed |