Reading system, reading device, reading method, and storage medium
AssigneeKabushiki Kaisha Toshiba
InventorToshikazu Taki
Life of this patent
Estimating the term…
Abstract
According to one embodiment, a reading system includes an extractor, a determiner, and a reader. The extractor extracts a candidate image from an input image. The candidate image is a candidate of a portion of the input image in which a segment display is imaged. The determiner uses the candidate image and a mask to calculate a match ratio indicating a certainty of a segment display being included in the candidate image, and determines that the candidate image is an image of a segment display when the match ratio is not less than a threshold. The mask and the threshold are preset. The reader reads a numerical value displayed in a segment display from the candidate image determined to be an image of a segment display.
Claims15
A processor, comprising:
- a first processing circuitry,
- using a candidate image and a mask image to calculate a match ratio indicating a certainty of a segment display being included in the candidate image, the candidate image being extracted from an input image, the candidate image being a candidate of a portion of the input image in which the segment display is imaged, the segment display including a plurality of segment groups, each of the segment groups including a plurality of segments and displaying a number, the mask image being a binary image and including a plurality of mask regions for the segment groups,
- determining that the candidate image is an image of a segment display in a case where the match ratio satisfies a first condition;
- a second processing circuitry
- extracting a contour from the input image,
- calculating a surface area of a region surrounded with the contour,
- extracting, based on the contour, an image having a prescribed shape from the input image when the surface area is not less than a threshold, the threshold being preset, and
- outputting, as the candidate image, the extracted image satisfying a second condition;
- and
- a third processing circuitry reading a numerical value displayed by the segments groups from the candidate image determined to be an image of the segment display.
A reading method, comprising:
- extracting a contour from an input image;
- calculating a surface area of a region surrounded with the contour;
- extracting, based on the contour, an image having a prescribed shape from the input image when the surface area is not less than a threshold, the threshold being preset;
- outputting, as a candidate image, the extracted image satisfying a second condition, the candidate image being a candidate of a portion of the input image in which a segment display is imaged, the segment display including a plurality of segment groups, each of the segment groups including a plurality of segments and displaying a number;
- using the candidate image and a mask image to calculate a match ratio indicating a certainty of the segment display being included in the candidate image, the mask image being a binary image and including a plurality of mask regions for the segment groups;
- determining that the candidate image is an image of a segment display in a case where the match ratio satisfies a first condition;
- and
- reading a numerical value displayed by the segments groups from the candidate image determined to be an image of the segment display.
Dependencies are inferred from the claim text.