Automatic ruler detection

1. A system comprising:

a camera; and

one or more computing devices comprising a processor and a memory coupled to said processor, wherein said memory comprises computer-executable instructions that, when executed by said processor, cause the one or more computing device to perform operations comprising:

receiving, from the camera, a sample image that includes a fingerprint and a mensuration reference device, wherein the sample image is associated with a resolution;

processing the sample image to generate a gradient image of the sample image, wherein:

the gradient image comprises a plurality of pixels, and

each pixel within the plurality of pixels is associated with a gradient value representing a respective change in pixel intensity of a corresponding pixel in the sample image with respect to neighboring pixels along an x-direction and neighboring pixels along a y-direction;

identifying a set of edge pixels in the gradient image, the set of edge pixels including pixels in the gradient image that each have a gradient value that satisfies a predetermined threshold;

generating an orientation map for the sample image based on identifying the set of edge pixels in the gradient image, the orientation map specifying a respective orientation for each edge pixel included in the orientation map;

identifying, within a spatial domain of the orientation map, a plurality of edge candidate groups, wherein each of the plurality of edge candidate groups (i) include two or more edge pixels that have respective orientations satisfying a threshold similarity and (ii) identify regions of the sample image that are predicted to be occupied by a mensuration reference device;

determining a set of regularity characteristics for each of the plurality of edge candidate groups;

determining that the set of regularity characteristics for a particular edge candidate group, from among the plurality of edge candidate groups, matches a set of reference regularity characteristics that indicates a mensuration reference device;

extracting a set of predicted ruler marks from the particular edge candidate group based on determining that the set of regularity characteristics for the particular edge candidate group matches the set of reference regularity characteristics that indicates a mensuration reference device;

computing a scale associated with the sample image based at least on extracting the set of predicted ruler marks from the particular edge candidate group; and

generating, based at least on the scale associated with the sample image, a scaled image; and

providing the scaled image for output.

2. The system of claim 1 , wherein the operations comprise:

identifying a quality metric that defines a minimum level of gradient quality for the gradient image;

determining a gradient quality value associated with the gradient image; and

determining that the gradient value satisfies the quality metric.

3. The system of claim 1 , wherein the operations comprise:

computing an orientation histogram based at least on the orientation map;

identifying a plurality of matching orientations within an interval centered at a peak of the orientation histogram; and

generating one or more of the plurality of edge candidate groups using the edge pixels corresponding to the identified plurality of matching orientations.

4. The system of claim 3 , wherein the operations comprise:

defining the interval based at least on the number of orientations represented within the orientation map.

5. The system of claim 1 , wherein the operations comprise:

retrieving a plurality of ruler descriptions; and

determining that the particular edge candidate group includes a valid ruler based at least on the received plurality of ruler descriptions.

6. The system of claim 1 , wherein the operations comprise:

extracting the set of ruler marks based at least on estimating a ruler orientation described in the particular edge candidate group;

performing a plurality of regularity tests on the extracted set of ruler marks; and

removing a plurality of false marks from the set of ruler marks based at least on the plurality of regularity tests.

7. A computer-implemented method for automatic ruler detection comprising:

receiving, from the camera, a sample image that includes a fingerprint and a mensuration reference device, wherein the sample image is associated with a resolution;

processing the sample image to generate a gradient image of the sample image, wherein:

the gradient image comprises a plurality of pixels, and

each pixel within the plurality of pixels is associated with a gradient value representing a respective change in pixel intensity of a corresponding pixel in the sample image with respect to neighboring pixels along an x-direction and neighboring pixels along a y-direction;

identifying a set of edge pixels in the gradient image, the set of edge pixels including pixels in the gradient image that each have a gradient value that satisfies a predetermined threshold;

generating an orientation map for the sample image based on identifying the set of edge pixels in the gradient image, the orientation map specifying a respective orientation for each edge pixel included in the orientation map;

identifying, within a spatial domain of the orientation map, a plurality of edge candidate groups, wherein each of the plurality of edge candidate groups (i) include two or more edge pixels that have respective orientations satisfying a threshold similarity and (ii) identify regions of the sample image that are predicted to be occupied by a mensuration device;

determining a set of regularity characteristics for each of the plurality of edge candidate groups;

determining that the set of regularity characteristics for a particular edge candidate group, from among the plurality of edge candidate groups, matches a set of reference regularity characteristics that indicates a mensuration reference device;

extracting a set of predicted ruler marks from the particular edge candidate group based on determining that the set of regularity characteristics for the particular edge candidate group matches the set of reference regularity characteristics that indicates a mensuration reference device;

computing a scale associated with the sample image based at least on extracting the set of predicted ruler marks from the particular edge candidate group; and

generating, based at least on the scale associated with the sample image, a scaled image; and

providing the scaled image for output.

8. The computer-implemented method of claim 7 , comprising:

identifying a quality metric that defines a minimum level of gradient quality for the gradient image;

determining a gradient quality value associated with the gradient image; and

determining that the gradient value satisfies the quality metric.

9. The computer-implemented method of claim 7 , comprising:

computing an orientation histogram based at least on the orientation map;

identifying a plurality of matching orientations within an interval centered at a peak of the orientation histogram; and

generating one or more of the plurality of edge candidate groups using the edge pixels corresponding to the identified plurality of matching orientations.

10. The computer-implemented method of claim 9 , comprising:

defining the interval based at least on the number of orientations represented within the orientation map.

11. The computer-implemented method of claim 7 , comprising:

retrieving a plurality of ruler descriptions; and

determining that the particular edge candidate group includes a valid ruler based at least on the received plurality of ruler descriptions.

12. The computer-implemented method of claim 7 , comprising:

extracting the set of ruler marks based at least on estimating a ruler orientation described in the particular edge candidate group;

performing a plurality of regularity tests on the extracted set of ruler marks; and

removing a plurality of false marks from the set of ruler marks based at least on the plurality of regularity tests.

13. A computer-readable storage medium encoded with a computer program, the program comprising instructions that when executed by data processing apparatus cause the data processing apparatus to perform operations comprising:

receiving, from the camera, a sample image that includes a fingerprint and a mensuration reference device, wherein the sample image is associated with a resolution;

processing the sample image to generate a gradient image of the sample image, wherein:

the gradient image comprises a plurality of pixels, and

each pixel within the plurality of pixels is associated with a gradient value representing a respective change in pixel intensity of a corresponding pixel in the sample image with respect to neighboring pixels along an x-direction and neighboring pixels along a y-direction;

identifying a set of edge pixels in the gradient image, the set of edge pixels including pixels in the gradient image that each have a gradient value that satisfies a predetermined threshold;

generating an orientation map for the sample image based on identifying the set of edge pixels in the gradient image, the orientation map specifying a respective orientation for each edge pixel included in the orientation map;

identifying, within a spatial domain of the orientation map, a plurality of edge candidate groups, wherein each of the plurality of edge candidate groups (i) include two or more edge pixels that have respective orientations satisfying a threshold similarity and (ii) identify regions of the sample image that are predicted to be occupied by a mensuration device;

determining a set of regularity characteristics for each of the plurality of edge candidate groups;

determining that the set of regularity characteristics for a particular edge candidate group, from among the plurality of edge candidate groups, matches a set of reference regularity characteristics that indicates a mensuration reference device;

extracting a set of predicted ruler marks from the particular edge candidate group based on determining that the set of regularity characteristics for the particular edge candidate group matches the set of reference regularity characteristics that indicates a mensuration reference device;

computing a scale associated with the sample image based at least on extracting the set of predicted ruler marks from the particular edge candidate group; and

generating, based at least on the scale associated with the sample image, a scaled image; and

providing the scaled image for output.

14. The computer-readable storage medium of claim 13 , wherein the operations comprise:

identifying a quality metric that defines a minimum level of gradient quality for the gradient image;

determining a gradient quality value associated with the gradient image; and

determining that the gradient value satisfies the quality metric.

15. The computer-readable storage medium of claim 13 , wherein the operations comprise:

computing an orientation histogram based at least on the orientation map;

identifying a plurality of matching orientations within an interval centered at a peak of the orientation histogram; and

generating one or more of the plurality of edge candidate groups using the edge pixels corresponding to the identified plurality of matching orientations.

16. The computer-readable storage medium of claim 15 , wherein the operations comprise:

defining the interval based at least on the number of orientations represented within the orientation map.

17. The computer-readable storage medium of claim 13 , wherein the operations comprise:

retrieving a plurality of ruler descriptions; and

determining that the ruler candidate group includes a valid ruler based at least on the received plurality of ruler descriptions.

18. The computer-readable storage medium of claim 13 , wherein the operations comprise:

extracting the set of ruler marks based at least on estimating a ruler orientation described in the identified ruler candidate group;

performing a plurality of regularity tests on the extracted set of ruler marks; and

removing a plurality of false marks from the set of ruler marks based at least on the plurality of regularity tests.