FEATURE EXTRACTION USING PIXEL-LEVEL AND OBJECT-LEVEL ANALYSIS

• US 20070116365A1
• Filed: 11/13/2006
• Published: 05/24/2007
• Est. Priority Date: 11/23/2005
• Status: Active Grant

• Current Assignee: Leica Geosystems AG (Hexagon AB)
• Original Assignee: Leica Geosytems Geiospatial Imaging LLC
• Inventors: Kloer, Brian
• Granted Patent: US 7,933,451 B2
• Time in Patent Office: Days
• US Class Current: 382/190
• CPC Class Codes: G06F 18/211 Selection of the most significant subset of features
• CPC Class Codes: G06T 17/05 Geographic models
• CPC Class Codes: G06T 2207/10032 Satellite or aerial image; Remote sensing
• CPC Class Codes: G06T 2207/30184 Infrastructure
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/136 involving thresholding
• CPC Class Codes: G06T 7/143 involving probabilistic approaches, e.g. Markov random field [MRF] modelling
• CPC Class Codes: G06T 7/49 based on structural texture description, e.g. using primitives or placement rules
• CPC Class Codes: G06V 10/255 Detecting or recognising potential candidate objects based on visual cues, e.g. shapes

Abstract

Image processing for extracting features in images. Pixel-level cue algorithms can be performed on raster images. The raster images can be converted to a vector layer. Object-level cue algorithms can be performed on the vector layer. The feature can be detected using a result of the pixel-level cues and using a result of the object-level cue algorithms performed. A computer-readable medium can include a first data field containing data representing pixel-level cues functioning to describe a pixel-level cue of the feature. The computer-readable medium can also include a second data field containing data representing object-level cues functioning to describe the object-level cues of the feature. Relation-level cue algorithms an be performed on the vector layers. The features can be detected using a result of any combination of the pixel-level cue algorithms, object-level cue algorithms, and/or relation-level cue algorithms.

Claims

1. A method for extracting information from an image, the method comprising:

performing a pixel-level cue algorithm on a digital raster image;

converting at least a portion of the raster image to a digital vector layer;

performing an object-level cue algorithm on the digital vector layer; and

identifying a feature using a result of the pixel-level cue algorithm and a result of the object-level cue algorithm performed.

2. The method according to claim 1, further comprising performing a relation-level cue algorithm, wherein the feature is identified using the result from the pixel-level cue algorithm, the result from the object-level cue algorithm, and a result from the relation-level cue algorithm.

3. The method according to claim 1, wherein the object-level cue algorithm is performed only if a result from the pixel-level cue algorithm indicates that an area of the pixel image satisfies a pixel-level probability threshold that the area includes at least a portion of the feature.

4. The method according to claim 2, wherein the relation-level cue algorithm is performed only if a result from the object-level cue algorithm indicates that an area of the vector layer satisfies an object-level probability threshold that the area includes at least a portion of the feature.

5. The method according to claim 2, wherein the object-level cue algorithm is performed only after a result from the relation-level cue algorithm indicates that an area of the vector layer satisfies a vector level probability threshold that the area includes at least a portion of the feature.

6. A method according to claim 1, wherein at least one pixel-level cue employed in connection with the pixel-level cue algorithm includes at least one of:

a pixel cue that is visually detectable by a human; and

a pixel cue that is visually undetectable by a human.

7. The method according to claim 1, wherein:

the pixel-level cue comprises at least one or any combination of the members of a set comprising;

a spectral cue;

a texture cue;

a vegetative cue;

a site cue;

a slope cue;

an aspect cue;

a landcover cue;

a soil cue;

a temperature cue; and

a user defined pixel-level cue; and

the object-level cue comprises at least one or any combination of the members of a set comprising;

a shape cue;

a size cue;

a orientation cue;

a right angle cue;

a template matching cue;

a orthogonality cue;

a perimeter cue;

a length cue;

a width cue;

a boundary complexity cue;

a rectangular fit cue;

an elliptical fit cue;

a compactness cue;

a symmetry cue;

a direction cue;

a background cue;

a spectral information of surrounding pixels cue;

a textural information of surrounding pixels cue;

an extent cue;

an area cue;

a shadow cue;

a background cue; and

a zonal cue.

8. The method according to claim 2, wherein the relation-level cue comprises one of a pattern cue or an association cue.

9. The method according to claim 1, wherein probabilities determined in connection with performance of the pixel-level cue algorithm are calculated using one of:

a Bayesian Network;

Classification and Regression Trees;

or a Neural Network.

10. In a graphical information system (GIS), a method for processing a digital image to locate a feature in the digital image, the method comprising the following:

executing a pixel-level cue algorithm to identify an interesting area of a raster image;

determining a pixel-level probability that the interesting area of the raster image identified is the feature using a result from the pixel-level cue algorithm;

comparing the pixel-level probability to a pixel-level cue threshold; and

if the pixel-level probability satisfies the pixel-level cue threshold;

converting at least a portion of the raster image to a vector layer;

executing an object-level cue algorithm on the vector layer to identify an interesting area of the vector layer;

determining an object-level probability that the interesting area of the vector layer is the feature using a result of the pixel-level cue algorithm; and

comparing the object-level probability to an object-level threshold.

11. The method according to claim 10, further comprising:

if the object-level probability meets the object-level threshold;

identifying the interesting area of the vector layer as the feature.

12. The method according to claim 10, further comprising:

if the object-level probability meets the object-level threshold;

executing a relation-level cue algorithm on the vector layer to identify an interesting area of the vector layer;

determining a relation-level probability for the interesting area of the vector layer using a result of the relation-level cue algorithm;

comparing the relation-level probability to a relation-level threshold; and

if the relation-level probability satisfies the relation-level threshold;

identify the interesting area of the vector layer as the feature.

13. The method according to claim 10, wherein a training cue metric is input by a user.

14. The method according to claim 13, wherein the cue metric is at least one of:

a training pixel; and

a training vector object.

15. The method according to claim 10, wherein:

the pixel-level cue comprises at least one of;

a spectral cue;

a texture cue;

a vegetative cue;

a site cue;

a slope cue;

an aspect cue;

a landcover cue;

a soil cue;

a temperature cue; and

a user defined pixel-level cue; and

the object-level cue comprises at least one of;

a shape cue;

a size cue;

a orientation cue;

a right angle cue;

a template matching cue;

an orthogonality cue;

a perimeter cue;

a length cue;

a width cue;

a boundary complexity cue;

a rectangular fit cue;

an elliptical fit cue;

a compactness cue;

a symmetry cue;

a direction cue;

a background cue;

a spectral information of surrounding pixels cue;

a textural information of surrounding pixels cue;

an extent cue;

an area cue;

a shadow cue;

a background cue; and

a zonal cue.

16. The method according to claim 12, wherein the relation-level cue comprises one of the members of a set comprising:

a pattern cue; and

an association cue.

17. The method according to claim 10, wherein the probabilities are calculated using one of:

a Bayesian Network;

Classification and Regression Trees;

or a Neural Network.

18. The method according to claim 10, further comprising:

extracting the feature; and

generating geometric object data representing the feature.

19. In a computer-readable medium, a feature template data structure including a plurality of data fields, comprising:

a first data field containing data representing a pixel-level cue functioning to describe a pixel-level cue of the feature; and

a second data field containing data representing an object-level cue functioning to describe an object-level cue of the feature.

20. The feature template data structure according to claim 19, further comprising:

a third data field containing data representing a relation-level cue functioning to describe a relation-level cue of the feature.

21. The feature template data structure according to claim 19, wherein:

the first data field contains data describing parameters, weight, result, and probability for a pixel-level cue algorithm; and

the second data field contains data describing parameters, weight, result, and probability inputs for an object-level cue algorithm.

22. The feature template data structure according to claim 19, wherein:

the pixel-level cue comprises at least one of or any combination of the members of a set comprising;

a spectral cue;

a texture cue;

a variance cue;

a skewness cue;

a entropy cue;

a fractal cue;

a vegetative cue;

a site cue;

a slope cue;

an aspect cue;

a landcover cue;

a soil cue;

a temperature cue; and

a user defined pixel-level cue; and

the object-level cue comprises data representing a member of a set comprising;

a shape cue;

a size cue;

a orientation cue;

a right angle cue;

a template matching cue;

a orthogonality cue;

a perimeter cue;

a length cue;

a width cue;

a boundary complexity cue;

a rectangular fit cue;

an elliptical fit cue;

a compactness cue;

a symmetry cue;

a direction cue;

a background cue;

a spectral information of surrounding pixels cue;

a textural information of surrounding pixels cue;

an extent cue;

an area cue;

a shadow cue;

a background cue; and

a zonal cue.

23. The feature template data structure according to claim 20, wherein the relation-level cue comprises data representing at least one of or any combination of the members of a set comprising:

a pattern cue and an association cue.

24. The feature template data structure according to claim 19, further comprising:

executable instructions for performing the following;

executing a pixel-level cue algorithm on a raster image using the first data field;

converting the raster image to a vector layer;

executing an object-level cue algorithm on the vector layer using the second data field; and

detecting the feature using a result of the pixel-level cue and either a result from the object-level cue algorithm or a result from the relation-level cue algorithm performed.

25. In a computing system, one or more computer-readable media including computer-executable instructions for implementing a method of extracting information from an image, wherein the computer executable instructions, when executed by a processor, cause the computing system to perform the following:

performing a pixel-level cue algorithm on a digital raster image;

converting at least a portion of the raster image to a digital vector layer;

performing one of the following;

an object-level cue algorithm on the digital vector layer;

or

a relation-level cue algorithm on the digital vector layer; and

identifying a feature using a result of the pixel-level cue algorithm and either;

a result from the object-level cue algorithm;

or a result from the relation-level cue algorithm performed.