Process for identifying simple and complex objects from fused images and map data

• US 5,631,970 A
• Filed: 05/21/1993
• Issued: 05/20/1997
• Est. Priority Date: 05/21/1993
• Status: Expired due to Term

• Current Assignee: Shin-Yi Hsu
• Original Assignee: Shin-Yi Hsu
• Inventors: Hsu, Shin-yi
• Primary Examiner(s): Mancuso, Joseph
• Application Number:
  US08/066,691
• Time in Patent Office: 1,460 Days
• Field of Search: 382/1, 382/16, 382/19, 382/20, 382/22, 382/48, 382/45, 382/41, 364/274.6, 395/61, 348/116-120, 348/122, 348/123
• US Class Current: 382/113
• CPC Class Codes: G06F 18/251 of input or preprocessed data
• CPC Class Codes: G06V 20/13 Satellite images
• CPC Class Codes: G06V 30/422 Technical drawings; Geographical maps

Abstract

The present invention features a method of identifying and/or extracting an object from multiple fused information sources, such as maps and images. The system allows a user to integrate information freely from multiple sources, such as maps, socio-economic data and various types of images. Data is first retrieved and then processed to transform its pixel representation to a region representation. The region representation is then refined by merging mechanisms, such as segmentation. The identified and/or extracted object can then be visualized by the user. Features or objects can then be extracted using the user'"'"'s expert rules and confidence levels, which confidence levels may be derived by fuzzy logic.

Claims

1. A method of analyzing combined image and map data utilizing a processing system that provides total, seamless integration, both vertical and horizontal, between image and map data, in order to recognize and identify objects disposed within said image and map data, said method comprising:

a) transforming a variety of information sources, including maps and imagery, for a predetermined geographical area, into a mutually compatible, extractable format;

b) defining features and regions of said maps and imagery representative of object information, using a multilevel process, wherein said object information within any level of said multilevel process communicates with each other using an object-based segmentation scheme, said segmentation scheme facilitating or discouraging a merger of neighboring regions of said information sources based upon attributes defined by a set of object-based rules, including spatial relationships;

c) identifying and recognizing objects disposed within said information sources by comparing said object information with said object-based rules, said object-based rules being language-definable;

d) obtaining object-based rules from a database; and

e) recognizing objects by comparing object features of said objects with said obtained object-based rules of known object features and attributes.

2. The method of analyzing combined image and map data in accordance with claim 1, wherein said object-based rules are reconstructible using a non-English human language.

3. A method of analyzing combined image and map data utilizing a reconfigurable processing system that provides integration between image and map data, in order to recognize and identify objects disposed within said image and map data, utilizing a language-defined, object-based set of rules, said method comprising:

a) transforming a variety of information sources, including maps and imagery, for a predetermined geographical area, into a mutually compatible, extractable format;

b) defining features and regions of said maps and imagery representative of object information, using a multi-level process, wherein said object information within any level of said multi-level process communicates with each other using an object-based segmentation scheme, said segmentation scheme facilitating or discouraging a merger of neighboring regions of said information sources, based upon attributes defined by a set of object-based rules, including spatial relationships, said defining operation being facilitated by an interactive technique, whereby a statement can be inserted anywhere therein within said set of object-based rules; and

c) identifying and recognizing objects disposed within said information sources.

4. The method of analyzing combined image and map data in accordance with claim 3, wherein said object-based rules are reconstructible, using a non-English human language.

5. A user-interactive method of analyzing combined image and map data, utilizing a processing system that provides integration between image and map data, in order to recognize and identify objects disposed within said image and map data, utilizing a language-defined, object-based set of rules that is user-modifiable, said method comprising:

a) transforming a variety of information sources, including maps and imagery, for a predetermined geographical area, into a mutually compatible, extractable format;

b) defining features and regions of said maps and imagery representative of object information, using a multi-level process, wherein said object information within any level of said multi-level process communicates with each other, and objects within said maps and imagery are recognizable by comparing them with attributes defined by a user-modifiable set of object-based rules, said object-based rules being part of a database and reconfigurable in a user-friendly human language; and

c) identifying and recognizing objects disposed within said image and map data.