Digital image generation through an active lighting system
Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for an active lighting system. In one aspect, a method includes receiving a first image of the physical document having a first glare signature and a second image of the physical document having a second glare signature that is different from the first glare signature; determining a first glare map of the first image and a second glare map of the second image; comparing the first glare map to the second glare map; and generating the digital image based on the comparison of the first and second glare maps.
1. A computer-implemented method for providing a digital image of a physical document, the method comprising:
receiving a first image of the physical document having a first glare signature and a second image of the physical document having a second glare signature that is different from the first glare signature;
determining a first glare map of the first image and a second glare map of the second image;
comparing the first glare map to the second glare map; and
generating the digital image based on the comparison of the first and second glare maps.
2. The computer-implemented method of claim 1 , wherein the first image is taken with a flash, and wherein the second image is taken without a flash.
3. The computer-implemented method of claim 1 , wherein the first image and the second image are taken in succession within a threshold temporal distance.
4. The computer-implemented method of claim 1 , compromising:
before determining the first glare map of the first image and the second glare map of the second image, aligning the first image and the second image based on the physical document.
5. The computer-implemented method of claim 4 , wherein aligning the first image and the second image includes:
estimating a homography using features from accelerated segment test (FAST) detector and an oriented FAST and rotated Binary Robust Independent Elementary Features (ORB) detector to provide a description of texture around the physical document as depicted in each image; and
warping each pixel in the second image with the first image through a bi-linear interpolation.
6. The computer-implemented method of claim 5 , wherein the homography is estimated based on a random sample consensus (RANSAC) algorithm.
7. The computer-implemented method of claim 1 , wherein determining the first glare map of the first image includes generating a first greyscale image of the first image, and determining a second glare map of the second image includes generating a second greyscale image of the second image.
8. The computer-implemented method of claim 1 , wherein the first glare map and the second glare map are each binary images where each pixel represents either glare or no glare.
9. The computer-implemented method of claim 1 , further comprising:
before comparing the first glare map to the second glare map, dilating the first glare map and the second glare map to expanded regions of glare represent on each map.
10. The computer-implemented method of claim 9 , wherein the digital image is generated by replacing the regions of glare in the first image with respective mapped regions from the second image, wherein the mapped regions from the second image do not include glare.
11. The computer-implemented method of claim 10 , wherein the mapped regions from the second image are merged into the first image to form the digital image through Poisson imaging blending.
12. The computer-implemented method of claim 11 , wherein gradient information throughout the replaced regions of glare is employed to interpolate a color propagated from a boundary of each replaced glare region in the generated digital image.
13. The computer-implemented method of claim 10 , wherein the mapped regions from the second image are merged into the first image to form the digital image through a Mean Value Coordinates (MVC) for Instant Image Cloning algorithm.
14. The computer-implemented method of claim 1 , wherein the digital image is employed in an analysis of the physical document to identify text or data elements in the physical document.
15. The computer-implemented method of claim 14 , wherein the analysis of the physical document includes at least one of optical character recognition (OCR), optical word recognition (OWR), intelligent character recognition (ICR), intelligent word recognition (IWR), natural language processing (NLP), or machine learning.
16. The computer-implemented method of claim 1 , wherein the digital image is employed in an analysis of the physical document to detect digital tampering or physical tampering.
17. The computer-implemented method claim 1 , wherein the physical document is a professional or government-issued credentials or certifications.
18. One or more non-transitory computer-readable storage media coupled to one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving a first image of a physical document having a first glare signature and a second image of the physical document having a second glare signature that is different from the first glare signature;
determining a first glare map of the first image and a second glare map of the second image;
comparing the first glare map to the second glare map; and
generating a digital image based on the comparison of the first and second glare maps.
19. The one or more non-transitory computer-readable storage media of claim 18 , wherein the digital image is employed in an analysis of the physical document to identify text or data elements in the physical document or to detect digital tampering or physical tampering, wherein the analysis of the physical document includes at least one of optical character recognition (OCR), optical word recognition (OWR), intelligent character recognition (ICR), intelligent word recognition (IWR), natural language processing (NLP), or machine learning, and wherein the physical document is a professional or government-issued credentials or certifications.
20. A system, comprising:
one or more processors; and
a computer-readable storage device coupled to the one or more processors and having instructions stored thereon which, when executed by the one or more processors, cause the one or more processors to perform operations comprising:
receiving a first image of a physical document having a first glare signature and a second image of the physical document having a second glare signature that is different from the first glare signature;
aligning the first image and the second image based on the physical document by:
estimating a homography using features from accelerated segment test (FAST) detector and an oriented FAST and rotated Binary Robust Independent Elementary Features (ORB) detector to provide a description of texture around the physical document as depicted in each image; and
warping each pixel in the second image with the first image through a bi-linear interpolation;
determining a first glare map of the first image by generating a first greyscale image of the first image;
determining a second glare map of the second image by generating a second greyscale image of the second image;
dilating the first glare map and the second glare map to expanded regions of glare represent on each map;
comparing the first glare map to the second glare map; and
generating a digital image by replacing the regions of glare in the first image with respective mapped regions from the second image, wherein the mapped regions from the second image do not include glare.