Compensating for geometric distortion of images in constrained processing environments
An image processing method determines a geometric transform of a suspect image by efficiently evaluating a large number of geometric transform candidates in environments with limited processing resources. Processing resources are conserved by using complementary methods for determining a geometric transform of an embedded signal. One method excels at higher geometric distortion, and specifically, distortion caused by greater tilt angle of a camera. Another method excels at lower geometric distortion, for weaker signals. Together, the methods provide a more reliable detector of an embedded data signal in image across a larger range of distortion while making efficient use of limited processing resources in mobile devices.
1. A reader device comprising:
a camera operable to capture an image;
memory configured to store the image from the camera;
means for converting the image in the memory into a first domain for evaluating an embedded signal to determine a geometric transform of the embedded signal;
means for evaluating a first set of geometric transform candidates by using the first set of geometric candidates as seeds to a least squares process and producing a first set of refined geometric transform candidates with detection metrics;
means for evaluating a second set of geometric transform candidates by applying a correlation process on a rotation and scale mapping of the image and producing a second set of refined geometric transform candidates with detection metrics;
means for selecting a refined candidate geometric transform from first and second refined geometric transform candidates based on detection metrics; and
means for extracting a digital payload from the embedded signal using a selected refined candidate geometric transform.
2. The reader device of claim 1 wherein the means for evaluating the second set of geometric transform candidates applies a log polar correlation on the image mapped into a log polar coordinate space comprising a range of spatial scales and rotation angles to provide a measure of correlation at candidate pairs of rotation and scale values.
3. The reader device of claim 1 wherein the embedded signal comprises peaks in the first domain.
4. The reader device of claim 1 wherein the first domain comprises a spatial frequency transform domain.
5. The reader device of claim 1 wherein the first set of geometric transform candidates comprise candidates representing a first range of camera tilt angles, and wherein said means for evaluating the second set of geometric transform candidates evaluates a second set of geometric transform candidates defined by pairs of rotation and spatial scale values, the second set of geometric transform candidates having lower camera tilt angle than the first set, wherein the means for evaluating the first set of geometric transform candidates enables digital payload extraction from a suspect image distorted by a greater camera tilt angle than said means for evaluating the second set of geometric transform candidates.
6. A non-transitory computer readable medium on which is stored instructions, which when executed by a processor, perform a method of reading an embedded digital payload in an image, the method comprising:
obtaining an image;
transforming the image into a first domain for evaluating geometric distortion of a two-dimensional (2D) signal;
for a first set of geometric transform candidates, applying a least squares process that produces first refined geometric transform candidates from the first set having detection metrics for the 2D signal that satisfy predetermined criteria;
evaluating geometric transform candidates on a rotation and scale mapping of the image and producing second refined geometric transform candidates having detection metrics for the 2D signal that satisfy predetermined criteria;
selecting a refined candidate geometric transform from the first refined geometric transform candidates and the second refined geometric transform candidates based on detection metrics; and
using the refined candidate geometric transform to extract a plural-bit payload from the 2D signal.
7. The non-transitory computer readable medium of claim 6 wherein the least squares process comprises:
a) obtaining transformed coordinates of reference signal components of the 2D signal, the transformed coordinates having been geometrically transformed by a geometric transform candidate in the first set;
b) for the reference signal components, determining updated coordinates by locating an image feature in a neighborhood in the image around the transformed coordinates of a reference signal component, the image feature corresponding to a potential reference signal component in the image; and
c) determining a new geometric transform that provides a least squares mapping between coordinates of the reference signal components and the updated coordinates.
8. The non-transitory computer readable medium of claim 6 wherein the evaluating geometric transform candidates applies a log polar correlation on the image mapped into a log polar coordinate space comprising a range of spatial scales and rotation angles to provide a measure of correlation at candidate pairs of rotation and scale values.
9. The non-transitory computer readable medium of claim 6 wherein the 2D signal comprises peaks in the first domain.
10. The non-transitory computer readable medium of claim 6 wherein the first domain comprises a spatial frequency transform domain.
11. The non-transitory computer readable medium of claim 6 wherein the first set of geometric transform candidates comprise candidates representing a first range of camera tilt angles, and wherein the least squares process evaluates a second set of geometric transform candidates defined by pairs of rotation and spatial scale values, the second set of geometric transform candidates having lower camera tilt angle than the first set, wherein the evaluating geometric transform candidates enables the plural-bit payload extraction from an image distorted by a greater camera tilt angle than the least squares process.
12. The non-transitory computer readable medium of claim 11 wherein the evaluating geometric transform candidates comprises:
a) obtaining transformed coordinates of reference signal components of the 2D signal, the transformed coordinates having been geometrically transformed by a geometric transform candidate in the first set;
b) for the reference signal components, determining updated coordinates by locating an image feature in a neighborhood in the image around the transformed coordinates of a reference signal component, the image feature corresponding to a potential reference signal component in the image; and
c) determining a new geometric transform that provides a least squares mapping between coordinates of the reference signal components and the updated coordinates.
13. The non-transitory computer readable medium of claim 12 wherein the least squares mapping applies a log polar correlation on the image mapped into a log polar coordinate space comprising a range of spatial scales and rotation angles to provide a measure of correlation at candidate pairs of rotation and scale values.
14. The non-transitory computer readable medium of claim 11 wherein the least squares mapping applies a log polar correlation on the image mapped into a log polar coordinate space comprising a range of spatial scales and rotation angles to provide a measure of correlation at candidate pairs of rotation and scale values.