Image processing system, image processing apparatus, image processing method, and storage medium
An image processing system acquires a scanned image obtained by scanning an original, and extracts a character region that includes characters from within the scanned image. The image processing system performs conversion processing, for converting a font of a character included in the extracted character region from a first font to a second font, on the scanned image using a conversion model for which training has been performed in advance so as to convert characters of the first font in an inputted image into characters of the second font and output a converted image. Then, the image processing system executes OCR on the scanned image after the conversion processing.
1. An image processing system comprising:
at least one memory that stores a program; and
at least one processor that executes the program to perform:
acquiring a scanned image obtained by scanning an original;
extracting a character region that includes first character images from within the scanned image, wherein the first character images are represented by a first font;
obtaining a converted image by converting, on the scanned image, the first character images included in the character region into second character images using a conversion model, wherein the conversion model is generated in advance by performing training based on training data, and wherein the training data includes at least one set of a first image that includes a character represented by the first font and a second image that includes a character that is the same character as the character included in the first image and is represented by a second font, and wherein the first font is a font whose character recognition accuracy in accordance with the OCR is lower than character recognition accuracy in accordance with the OCR of the second font;
obtaining an OCR result by executing OCR on the converted image; and
outputting the obtained OCR result as character information of the scanned image.
2. The image processing system according to claim 1 , wherein
the training data includes the at least one set of the first image and the second image that each includes one character, and
the converted image is obtained by sequentially cutting out each character included in the character region one by one, and inputting an image of the cutout character to the conversion model.
3. The image processing system according to claim 1 , wherein
the training data is generated with the at least one set of the first image and the second image that each include a plurality of characters, and
the converted image is obtained by sequentially cutting out a region of a predetermined size from the character region, and inputting an image of the cutout region to the conversion model.
4. The image processing system according to claim 3 , wherein
the training data is generated with the at least one set of the first image and the second image which is obtained by cutting out a region of the predetermined size from each of a first text image that includes text represented by the first font and a second text image that includes text represented by the second font.
5. The image processing system according to claim 4 , wherein
the training data is generated with a plurality of sets of the first image and the second image which are obtained by cutting out a region of the predetermined size a plurality of times with respect to a different region for each time in the first text image and the second text image.
6. The image processing system according to claim 1 , wherein
the first image is generated by changing the second image to a deteriorated state.
7. The image processing system according to claim 1 , wherein
the image processing system comprises an image processing apparatus and a server apparatus capable of communicating with the image processing apparatus,
the image processing apparatus performs acquisition of the scanned image and OCR on the scanned image, and
the server apparatus performs extraction of the character region and the conversion processing.
8. The image processing system according to claim 7 , wherein
the image processing apparatus
transmits the acquired scanned image to the server apparatus,
receives the scanned image after the conversion processing from the server apparatus, and
executes the OCR on the scanned image received from the server apparatus.
9. An image processing apparatus comprising:
at least one memory that stores a program; and
at least one processor that executes the program to perform:
generating a scanned image by scanning an original;
extracting a character region that includes first character images from within the scanned image, wherein the first character images are represented by a first font;
obtaining a converted image by converting, on the scanned image, the first character images included in the character region into second character images using a conversion model, wherein the conversion model is generated in advance by performing training based on training data, and wherein the training data includes at least one set of a first image that includes a character represented by the first font and a second image that includes a character that is the same character as the character included in the first image and is represented by a second font, and wherein the first font is a font whose character recognition accuracy in accordance with the OCR is lower than character recognition accuracy in accordance with the OCR of the second font;
obtaining an OCR result by executing OCR on the converted image; and
outputting the obtained OCR result as character information of the scanned image.
10. An image processing method including:
acquiring a scanned image obtained by scanning an original;
extracting a character region that includes first character images from within the scanned image, wherein the first character images are represented by a first font;
obtaining a converted image by converting, on the scanned image, the first character images included in the character region into second character images using a conversion model, wherein the conversion model is generated in advance by performing training based on training data, and wherein the training data includes at least one set of a first image that includes a character represented by the first font and a second image that includes a character that is the same character as the character included in the first image and is represented by a second font, and wherein the first font is a font whose character recognition accuracy in accordance with the OCR is lower than character recognition accuracy in accordance with the OCR of the second font;
obtaining an OCR result by executing OCR on the converted image; and
outputting the obtained OCR result as character information of the scanned image.
11. A non-transitory computer-readable storage medium storing a computer program for causing a computer to execute an image processing method including:
acquiring a scanned image obtained by scanning an original;
extracting a character region that includes first character images from within the scanned image, wherein the first character images are represented by a first font;
obtaining a converted image by converting, on the scanned image, the first character images included in the character region into second character images using a conversion model, wherein the conversion model is generated in advance by performing training based on training data, and wherein the training data includes at least one set of a first image that includes a character represented by the first font and a second image that includes a character that is the same character as the character included in the first image and is represented by a second font, and wherein the first font is a font whose character recognition accuracy in accordance with the OCR is lower than character recognition accuracy in accordance with the OCR of the second font;
obtaining an OCR result by executing OCR on the converted image; and
outputting the obtained OCR result as character information of the scanned image.