Scaling in a multi-dimensional building model
IPC분류정보
국가/구분 |
United States(US) Patent
등록
|
국제특허분류(IPC7판) |
|
출원번호 |
US-0011126
(2016-01-29)
|
등록번호 |
US-10133830
(2018-11-20)
|
발명자
/ 주소 |
- Bhatawadekar, Vineet
- Sun, Shaohui
- Pavlidis, Ioannis
- Altman, Adam J.
|
출원인 / 주소 |
|
대리인 / 주소 |
|
인용정보 |
피인용 횟수 :
0 인용 특허 :
9 |
초록
▼
A system and method is provided for scaling and constructing a multi-dimensional (e.g., 3D) building model using ground-level imagery. Ground-level imagery is used to identify architectural elements that have known architectural standard dimensional ratios. Dimensional ratios of architectural elemen
A system and method is provided for scaling and constructing a multi-dimensional (e.g., 3D) building model using ground-level imagery. Ground-level imagery is used to identify architectural elements that have known architectural standard dimensional ratios. Dimensional ratios of architectural elements in the multi-dimensional building model (unscaled) are compared with known architectural standard dimensional ratios to scale and construct an accurate multi-dimensional building model.
대표청구항
▼
1. A method of scaling a multi-dimensional building model comprises: retrieving a digital ground-level image for the multi-dimensional building model, wherein the digital ground-level image has been rectified and correlated to vertices of geometric planes or polygons that make up a portion of the mu
1. A method of scaling a multi-dimensional building model comprises: retrieving a digital ground-level image for the multi-dimensional building model, wherein the digital ground-level image has been rectified and correlated to vertices of geometric planes or polygons that make up a portion of the multi-dimensional building model;identifying an architectural element within the digital ground-level image;defining perimeter boundaries for the architectural element;calculating a dimensional ratio of distances between the perimeter boundaries for the architectural element;comparing the calculated dimensional ratio against one or more known dimensional ratios, within a threshold of error, of at least one similar architectural element, wherein the comparing further includes comparing and averaging a plurality of calculated dimensional ratios for differing sizes of the at least one similar architectural element;selecting one of the one or more known dimensional ratios closest to the calculated dimensional ratio;determining an actual scale of the digital ground-level image based on error between the calculated dimensional ratio and the selected one of the one or more known dimensional ratios; andscaling the multi-dimensional building model accordingly with the actual scale. 2. The method of claim 1, wherein the perimeter boundaries are defined by any of: image pixels, vertices, edges or corner points. 3. The method of claim 1, wherein the scaling is performed while constructing the multi-dimensional building model. 4. The method of claim 1, wherein the perimeter boundaries are defined automatically or manually. 5. The method of claim 1, wherein the architectural element include: door(s), brick(s), window(s) or other architectural elements with architectural standard dimensional ratios. 6. The method of claim 1, wherein the retrieved digital ground-level image includes two or more architectural elements. 7. The method of claim 6, wherein the comparing further includes comparing against a plurality of dimensional ratios for the two or more architectural elements. 8. The method of claim 7, wherein the comparing further includes using a weighted decision to select a least error comparison evaluation for the two or more architectural elements. 9. The method of claim 1, wherein the dimensional ratios include any of: width-to-height, or height-to-width or area (width×height). 10. A method of scaling a multi-dimensional building model comprising: retrieving a digital ground-level image for the multi-dimensional building model, wherein the digital ground-level image has been rectified and correlated to vertices of geometric planes or polygons that make up a portion of the multi-dimensional building model;identifying a plurality of known architectural elements within the digital ground-level image;identifying perimeter boundaries for the plurality of known architectural elements identified within the digital ground-level image;calculating dimensional ratios within the identified perimeter boundaries of each of the plurality of known architectural elements;comparing each of the dimensional ratios against one or more known dimensional ratios, within a threshold of error, of at least one known similar architectural element, wherein the comparing further includes comparing and averaging a plurality of computed dimensional ratios for differing sizes of the at least one known similar architectural element;weighting the comparisons based on similarity of calculated dimensional ratios to their respective known architectural element dimensional ratios;selecting a dimensional ratio from the respective known architectural element dimensional ratios, wherein the selected dimensional ratio is closest to one of the calculated dimensional ratios;determining a scale of the digital ground-level image based on relative error of the selected dimensional ratio; andscaling the multi-dimensional building model with the determined scale. 11. The method of claim 10, wherein the scaling is performed while constructing the multi-dimensional building model. 12. The method of claim 10, wherein the at least one known architectural element includes: door(s), brick(s), or window(s). 13. The method of claim 10, wherein the retrieved ground-level image includes two or more known architectural elements. 14. The method of claim 13, wherein the known architectural elements are any of: similar or different. 15. The method of claim 10, wherein the dimensional ratios include any of: width-to-height, or height-to-width or area (width×height). 16. A computer-based system for scaling a multi-dimensional building model comprising: a computer processor with computing instructions to:retrieve a digital ground-level image for the multi-dimensional building model, wherein the digital ground-level image has been rectified and correlated to vertices of geometric planes or polygons that make up a portion of the multi-dimensional building model;identify a plurality of different known architectural elements within the digital ground-level image;identify perimeter boundaries for the plurality of different known architectural elements identified within the digital ground-level image;calculate dimensional ratios within the identified perimeter boundaries of each of the plurality of different known architectural elements;compare each of the dimensional ratios against one or more known dimensional ratios of at least one known similar architectural element;weight the comparisons based on similarity of calculated dimensional ratios to their respective known architectural element dimensional ratios;select a dimensional ratio from the respective known architectural element dimensional ratios, wherein the selected dimensional ratio is closest to one of the calculated dimensional ratios;determine a scale of the digital ground-level image based on relative error of the selected dimensional ratio; andscale the multi-dimensional building model with the determined scale.
이 특허에 인용된 특허 (9)
-
Schrag, John; Watanabe, Steven Takashi, 3D scene orientation indicator system with scene orientation change capability.
-
Lininger, Scott, Automatic method for photo texturing geolocated 3-D models from geolocated imagery.
-
Oldroyd, Lawrence Andrew, Height measurement in a perspective image.
-
Ohashi, Yousuke, Landmark search system for digital camera, map data, and method of sorting image data.
-
Kikuchi, Shin; Sakamoto, Hiroshi, Map display device and map display method.
-
Berry Richard Edmond ; Isensee Scott Harlan ; Roberts David John,GBX ; Bardon Didier Daniel ; Mullaly John Martin, Method and system for providing preferred face views of objects in a three-dimensional (3D) environment in a display in.
-
Shimazu,Hideo, Server and client for improving three-dimensional air excursion and method and programs thereof.
-
Steedly, Drew; Szeliski, Richard; Sinha, Sudipta; Agrawala, Maneesh, Using photo collections for three dimensional modeling.
-
Reinhardt, Tilman, Visualizing oblique images.
※ AI-Helper는 부적절한 답변을 할 수 있습니다.