Object detection and image cropping using a multi-detector approach

1. A computer-implemented method of detecting objects within digital image data, the method comprising:

defining a plurality of analysis profiles, wherein each analysis profile independently corresponds to one of a plurality of detectors, and wherein each analysis profile independently comprises either or both of: a unique set of analysis parameters and a unique detection algorithm;

analyzing the digital image data using some or all of the plurality of detectors in accordance with the analysis profiles, wherein each of the plurality of detectors independently produces a different analysis result;

selecting, based at least in part on a plurality of confidence scores each independently associated with a different one of the analysis results, an optimum analysis result; and

detecting one or more objects based at least in part on the optimum analysis result.

2. The computer-implemented method as recited in claim 1 , wherein the one or more objects each have unique characteristics, and wherein the method comprises detecting at least two of the one or more objects.

3. The computer-implemented method as recited in claim 2 , wherein the detecting is performed using at least two of the detectors.

4. The computer-implemented method as recited in claim 1 , wherein each of the different analysis results comprises:

the associated confidence score;

an image of at least a portion of the one or more detected objects; and/or

coordinates corresponding to a projected location of the one or more detected objects.

5. The computer-implemented method as recited in claim 1 , wherein the plurality of detectors comprises at least one text line detector.

6. The computer-implemented method as recited in claim 1 , wherein the plurality of detectors comprises at least one line segment detector.

7. The computer-implemented method as recited in claim 1 , comprising selecting and/or parameterizing at least one of the plurality of detectors based at least in part on a source of the digital image data.

8. The computer-implemented method as recited in claim 1 , comprising outputting a projected location of the one or more detected objects to either or both of: a memory, and a display.

9. A computer-implemented method of detecting objects within digital image data, the method comprising:

defining a plurality of analysis profiles, wherein each analysis profile independently corresponds to one of a plurality of detectors, wherein each analysis profile independently comprises one or more analysis parameters, and wherein the one or more analysis parameters comprise a subregion of the digital image data to be analyzed by the respective one of the plurality of detectors;

analyzing the digital image data using the plurality of detectors in accordance with the corresponding analysis profiles, wherein each of the plurality of detectors independently produces a different one of a plurality of analysis results;

selecting, based at least in part on some or all of a plurality of confidence scores each independently associated with one of the analysis results, an optimum analysis result; and

detecting one or more objects based at least in part on the optimum analysis result.

10. The computer-implemented method as recited in claim 9 , wherein each subregion of the digital image data corresponds to either:

a different corner of a digital image; or

a projected location of a portion of one of the one or more objects.

11. The computer-implemented method as recited in claim 9 , wherein the plurality of confidence scores each independently relate to one or more characteristics of the one or more objects, wherein the one or more characteristics are selected from:

a location of a portion of the one or more objects;

a color profile of a portion of the one or more objects;

a fitness of a portion of the one or more objects to an expected location or an expected curvature of the portion of the one or more objects;

an edge strength of each of the one or more edges of the one or more objects;

a value of one or more angles formed between adjacent of the one or more edges of the one or more objects;

a sum of two angles formed by opposite corners of a polygon representing the one or more objects;

a color contrast between a foreground of the digital image data and a background of the digital image data; and

combinations thereof.

12. The computer-implemented method as recited in claim 9 , comprising generating a composite analysis result based on at least two of the plurality of analysis results, wherein the optimum analysis result is based on the composite analysis result.

13. The computer-implemented method as recited in claim 12 , wherein the optimum analysis result comprises portions of the digital image data corresponding to at least two of the subregions of the digital image data.

14. The computer-implemented method as recited in claim 9 , comprising generating a composite confidence score based on at least two of the plurality of confidence scores.

15. The computer-implemented method as recited in claim 9 , wherein at least one of the one or more objects is characterized by a polygonal shape, and wherein each subregion of the digital image data corresponds to either an edge or a vertex of the polygonal shape.

16. The computer-implemented method as recited in claim 9 , wherein the plurality of detectors comprises at least one text line detector.

17. The computer-implemented method as recited in claim 9 , wherein the plurality of detectors comprises at least one line segment detector.

18. The computer-implemented method as recited in claim 9 , comprising selecting and/or parameterizing at least one of the plurality of detectors based at least in part on a source of the digital image data.

19. The computer-implemented method as recited in claim 9 , comprising outputting a projected location of the one or more detected objects to either or both of: a memory, and a display.