Systems and methods for mobile image capture and content processing of driver's licenses
Systems and methods are provided for processing and extracting content from an image captured using a mobile device. In one embodiment, an image is captured by a mobile device and corrected to improve the quality of the image. The corrected image is then further processed by adjusting the image, identifying the format and layout of the document, binarizing the image and extracting the content using optical character recognition (OCR). Multiple methods of image adjusting may be implemented to accurately assess features of the document, and a secondary layout identification process may be performed to ensure that the content being extracted is properly classified.
1. A computer-implemented method of processing an image, comprising:
receiving an image captured by a camera included in a mobile device;
performing a shape detection test on the image to produce a first set of coordinates representing a set of edges of the image;
performing an internal feature detection test on the image to produce a second set of coordinates associated with a feature set of the image, wherein the internal feature detection test, comprises:
identifying a photo are within the image,
determining an aspect ratio of the photo, and
determining the type of document associate with the image based at least in part on the aspect ratio;
combining the first set of coordinates and the second set of coordinates to generate a set of adjustment parameters associated with the image;
adjusting the image based at least on the adjustment parameters;
outputting the adjusted image.
2. The computer-implemented method of claim 1 , wherein prior to performing the internal feature detection test on the document image, the method further comprises pre-filtering the document image using a low-pass noise suppression filter.
3. The computer-implemented method of claim 1 , wherein the adjustment parameters includes both coordinates associated with the feature set of the document image and coordinates representing the set of edges of the document image.
4. The computer-implemented method of claim 1 , wherein the feature detection test detects a set of feature points in the document image by detecting a set of local maxima in the document image.
5. The computer-implemented method of claim 4 , wherein the internal feature detection test constructs the feature set by computing local gradients and color distributions in the areas of the set of feature points.
6. The computer-implemented method of claim 1 , wherein the shape detection test includes a rectangular shape detection test.
7. The computer-implemented method of claim 1 , wherein prior to combining the first set of coordinates and the second set of coordinates to generate the adjustment parameters, a rounded corner detection test on the document image to produce a third set of coordinates associated with the feature set of the document image.
8. The computer-implemented method of claim 7 , wherein generating the set of adjustment parameters further includes combining the first set of coordinates and the second set of coordinates with the third set of coordinates.
9. The computer-implemented method of claim 1 , wherein performing the internal feature detection test on the document image includes:
detecting a set of feature points in the document image;
constructing a set of feature descriptors based on the set of feature points;
wherein the feature matching operation between the set of feature descriptors and a set of predetermined features of the document template.
10. The computer-implemented method of claim 9 , wherein detecting the set of feature points in the document image includes detecting a set of local maxima in the document image.
11. The computer-implemented method of claim 9 , wherein constructing the set of feature descriptors based on the set of feature points includes computing local gradients and color distributions in the areas of the set of feature points.
12. The computer-implemented method of claim 1 , wherein the outputs of the feature matching operation include a transformation matrix containing a set of matched feature points.
13. The computer-implemented method of claim 1 , further comprising binarizing the adjusted image to produce a binarized adjusted image.