Back to results
US 2018/0121718 A1Published application
Document detection
AssigneeNCR Corporation
InventorOlufisayo Adebanke Ekpenyong
Life of this patent
Estimating the term…
Abstract
Video frames of a document are captured. A still image of the document is optimally automatically taken after the video frames have been recorded. The best image from the frames and the still image is determined. The best image is provided for OCR processing to derive document information represented in the best image for the document.
Claims20
A method, comprising:
- retaining frames of a video as document images for a document;
- obtaining a still image of the document;
- selecting a best image from the still image and the document images;
- and
- providing the best image for Optical Character Recognition processing to derive document information represented in the best image for the document.
2 more elements
A method, comprising:
- automatically activating a camera of a device to begin recording frames, each frame representing a document image of a document;
- determining when video is being recorded to activate the camera to take a higher resolution document image;
- selecting a subset of the document images and discarding remaining document images;
- calculating a focus score for each of the subset of document images;
- picking a particular one of the subset of document images based on the calculated focus scores for the subset of document images;
- increasing a current resolution of the particular document image to match a resolution of the higher resolution document image;
- identifying a best document image from the particular document image and the higher resolution document image;
- and
- providing the best document image to perform Optical Character Recognition (OCR) processing and obtaining document information represented in the best document image.
6 more elements
A device, comprising:
- a camera/scanner integrated into the device;
- and
- a document detector operable to (i) execute on one or more hardware processors of the device, (ii) activate the camera/scanner to capture video frames from video taken of a document, each frame including a document image, (ii) activate the camera/scanner while the video frames are being captured to take a higher resolution document image, (iii) select a particular one of the document images from the captured video frames based on calculated focus scores for each of the document images, and (iv) provide either the particular document image or the higher resolution document image as an optimal document image to an Optical Character Recognition (OCR) process for identifying document information represented therein.
Dependencies are inferred from the claim text.