Method and apparatus for OCR detection of valuable documents by means of a matrix camera
View Patent ↗The invention relates to a method for OCR detection of valuable documents in a cash dispenser in the case of which an image of the valuable document is detected by means of a digital video or matrix camera. A Hough transformation is used to calculate edge lines of the valuable document and a rotation angle is calculated therefrom such that the edges of the valuable document are aligned with the image edges. The detected image is homogenized to compensate an inhomogeneous image background. This is followed by OCR detection of alphanumeric information on the valuable document.
1. Method for optical character recognition (OCR) detection of documents in a self-service machine, in particular an automatic teller machine or cash dispenser, comprising:
a) detecting an image of a document by means of a digital video or matrix camera having a detection region;
b) establishing an intermediate document image by:
i) reducing the detected image,
ii) determining a position of a document region, corresponding to the document and also edge pixels in the detected image,
iii) detecting straight edge lines of the document region with the aid of the determined edge pixels by using a Hough transformation,
iv) determining a rotation angle by which the document region in the detected image must be rotated for alignment at edges of the detection region of the digital video or matrix camera, and
v) rotating the document region by the rotation angle, resulting in the intermediate document image; and
c) forming a background image by:
i) reducing the intermediate document image,
ii) after said reducing the intermediate document image, removing smaller details from the reduced intermediate document image by filtering the document region, resulting in the background image;
d) creating a brightness map of the background image; and
e) binarizing the background image to segment alphanumeric character information.
2. Method according to claim 1 , in which use is made in said detecting straight edge lines of an edge filter which outputs a binary edge image of the document region.
3. Method according to claim 2 , in which the Hough transformation is executed with the aid of the binary edge image.
4. Method according to claim 3 , in which during the Hough transformation it is determined for each pixel, which line runs therethrough, and an assessment of each line is raised when the pixel is an edge pixel, the straight edge lines corresponding to the lines most highly assessed.
5. Method according to claim 1 , in which said forming the background image furthermore comprises:
subtracting the brightness map from the detected intermediate document image.
6. Apparatus for optical character recognition (OCR) detection of documents in a self-service machine, in particular an automatic teller machine or cash dispenser, in particular designed as a document detection module, comprising:
a digital video or matrix camera having a detection region in order to detect an image of a document; and
an image processing section which is designed in order to:
a) establish an intermediate document image by:
i) reducing the detected image,
ii) determining a position of a document region, corresponding to the document and also edge pixels in the detected image,
iii) detecting straight edge lines of the document region with the aid of the determined edge pixels by using a Hough transformation,
iv) determining a rotation angle by which the document region must be rotated for alignment at edges of the detection region of the digital video or matrix camera, and
v) rotating the document region by the rotation angle, resulting in the intermediate document image; and
c) form a background image by:
i) reducing the intermediate document image,
ii) removing smaller details from the reduced intermediate document image after reducing the intermediate document image by filtering the document region, resulting in the background image; and
d) creating a brightness map of the background image and binarizing the background image to detect alphanumeric character information by OCR detection.
7. Apparatus according to claim 6 , in which the image processing section is further designed to apply an edge filter which outputs a binary edge image of the document region.
8. Apparatus according to claim 7 , in which the Hough transformation is executed with the aid of the binary edge image.
9. Apparatus according to claim 8 , in which during the Hough transformation it is determined for each pixel, which line runs therethrough, and an assessment of each line is raised when the pixel is an edge pixel, the straight edge lines corresponding to the lines most highly assessed.
10. Apparatus according to one of claim 6 , in which the image processing section is further designed to subtract said brightness map from the intermediate document image.