Importance guided image transformation
Importance guided image transformation. A subject image is accessed, an importance is assigned respective features of the subject image and a scaling scheme is determined for the subject image based on the importance assigned the respective features of the subject image. A transformed image is generated based on the determined scaling scheme and the transformed image is provided to an image presentation system for display.
1. A processor implemented method for importance guided image transformation, comprising:
determining a non-uniform scaling scheme for a subject image based on the importance that is accorded respective features of said subject image;
generating a transformed image based on said non-uniform scaling scheme wherein said respective features are transformed based on respective different scaling types; and
using a processor, providing access to said transformed image to an image presentation system for display,
wherein said importance that is accorded respective features of said subject image is determined by:
segmenting said subject image according to individual buildings;
detecting repeating elements of said individual buildings;
detecting business signs in said individual buildings; and
assigning relative scores to respective features of said individual buildings.
2. The method of claim 1 wherein said subject image is selected from the group consisting of street sides and maps.
3. The method of claim 1 wherein said importance guided image transformation comprises preserving important information from said subject image and minimizing visual distortion in said transformed image.
4. The method of claim 1 wherein said generating a transformed image based on said non-uniform scaling scheme comprises changing a font of a business sign.
5. The method of claim 1 wherein said generating said transformed image comprises detecting pixel sets of said subject image that are to be eliminated.
6. The method of claim 1 wherein said generating said transformed image comprises reducing the size of empty spaces between buildings that are captured in said subject image.
7. The method of claim 1 wherein said generating said transformed image comprises reducing the size of business signs of buildings that are captured in said subject image.
8. A processor implemented method for geo-information system importance guided image transformation, comprising:
accessing a subject image;
placing a value on respective portions of said subject image;
using a processor, determining a non-uniform scaling scheme for said subject image based on said value that is placed on said respective portions of said subject image wherein said non-uniform scaling scheme comprises respective different scaling types for said respective portions of said subject image;
generating a transformed image based on said non-uniform scaling scheme; and
providing said transformed image to an image presentation system of said geo-information system for display,
wherein said placing a value on respective portions of said subject image comprises:
segmenting said subject image according to individual buildings;
detecting repeating elements of said individual buildings;
detecting business signs in said individual buildings; and
assigning relative scores to respective features of said individual buildings.
9. The method of claim 8 wherein said subject image is selected from the group consisting of street sides and maps.
10. The method of claim 8 wherein said importance guided image transformation comprises preserving important information from said subject image and minimizing visual distortion in said transformed image.
11. The method of claim 8 wherein said generating a transformed image based on non-uniform said scaling scheme comprises changing a font of a business sign.
12. The method of claim 8 wherein said generating said transformed image comprises detecting pixel sets of said subject image that are to be eliminated.
13. The method of claim 8 wherein said generating said transformed image comprises reducing the size of empty spaces between buildings that are captured in said subject image.
14. The method of claim 8 wherein said generating said transformed image comprises reducing the size of business signs of buildings that are captured in said subject image.
15. A computer-readable medium having computer-executable instructions for performing steps, comprising:
accessing a subject image;
placing a value on respective portions of said subject image;
determining a non-uniform scaling scheme for said subject image based on said value that is placed on said respective portions of said subject image wherein said non-uniform scaling scheme comprises respective different scaling schemes for said respective portions of said subject image;
generating a transformed image based on said non-uniform scaling scheme; and
providing said transformed image to an image presentation system for display,
wherein said placing a value on respective portions of said subject image comprises:
segmenting said subject image according to individual buildings;
detecting repeating elements of said individual buildings;
detecting business signs in said individual buildings; and
assigning relative scores to respective features of said individual buildings.
16. The medium of claim 15 wherein said subject image is selected from the group consisting of street sides and maps.
17. The medium of claim 15 wherein said importance guided image transformation comprises preserving important information from said subject image and minimizing visual distortion in said transformed image.
18. The medium of claim 15 wherein said generating a transformed image based on said non-uniform scaling scheme comprises changing a font of a business sign.