The present disclosure provides a character detection method and apparatus, a model training method and apparatus, a device and a storage medium. The specific implementation is: acquiring a training sample, where the training sample includes a sample image and a marked image, and the marked image is an image obtained by marking a text instance in the sample image; inputting the sample image into a character detection model, to obtain segmented images and image types of the segmented images output by the character detection model, where the image type indicates that the segmented image includes a text instance, or the segmented image does not include a text instance; and adjusting a parameter of the character detection model according to the segmented images, the image types of the segmented images and the marked image.

In some examples, a system for detecting text in an image includes a memory device to store a text detection model trained using images of up-scaled text, and a processor configured to perform text detection on an image to generate original bounding boxes that identify potential text in the image. The processor is also configured to generate a secondary image that includes up-scaled portions of the image associated with bounding boxes below a threshold size, and perform text detection on the secondary image to generate secondary bounding boxes that identify potential text in the secondary image. The processor is also configured to compare the original bounding boxes with the secondary bounding boxes to identify original bounding boxes that are false positives, and generate an image file that includes the original bounding boxes, wherein those original bounding boxes that are identified as false positives are removed.

A medicine image recognition method applied to an electronic device is provided. The method includes obtaining target images by inputting medicine images into a position detection network. Character feature matrices are generated according to the target images and a character recognition network. Image feature matrices are generated by inputting the target images into a category recognition network. Reference matrices are generated according to the image feature matrices and corresponding character feature matrices. Once a matrix to be tested is generated by processing an image to be tested, and a recognition result of the image to be tested is generated according to a similarity between the matrix to be tested and each of the reference matrices.

The disclosure provides a test result recognizing method and a test result recognizing device. The method includes: controlling an image-capturing device to capture a first image of a display screen according to an image-capturing parameter; in response to determining that a reference image area including a first designated character string exists in the first image, controlling the image-capturing device to capture a first test image of the display screen according to the image-capturing parameter; extracting a first image area corresponding to the reference image area from the first test image, and performing a text dividing operation on the first image area to convert the first image area into a second image area; and performing a text recognition operation on the second image area to obtain a first test result corresponding to the first test image.

The disclosure provides a test result recognizing method and a test result recognizing device. The method includes: controlling an image-capturing device to capture a first image of a display screen according to an image-capturing parameter; in response to determining that a reference image area including a first designated character string exists in the first image, controlling the image-capturing device to capture a first test image of the display screen according to the image-capturing parameter; extracting a first image area corresponding to the reference image area from the first test image, and performing a text dividing operation on the first image area to convert the first image area into a second image area; and performing a text recognition operation on the second image area to obtain a first test result corresponding to the first test image.

Methods, apparatus, systems and articles of manufacture are disclosed for text extraction from a receipt image. An example non-transitory computer readable medium comprises instructions that, when executed, cause a machine to at least improve region of interest detection efficiency by converting pixels of an input receipt image from a first format to a second format, generate a binary representation of the input receipt image based on the converted pixels, the binary representation of the input receipt image corresponding to saturation values for respective ones of the converted pixels, calculate mirror data from the binary representation of the input receipt image, and cluster the binary representation of the input receipt image to identify a first set of candidate regions of interest, the candidate regions of interest characterized by portions of the binary representation of the input receipt image having saturation values that satisfy a threshold value.

Solutions for more efficient and effective optical character recognition with respect to an input text segment are disclosed. In one example, a method includes processing an input text image using a deep character overlap detection machine learning model in order to generate a character map for the input text image, an overlap map for the input text image, and an affinity map for the input text image; generating an overlap-aware word boundary recognition output based at least in part on the character map, the overlap map, and the affinity map, wherein the overlap-aware word boundary recognition output describes one or more inferred word regions of the input text image; and performing one or more prediction-based actions based at least in part on the overlap-aware word boundary recognition output.

A data generation apparatus includes: a separation unit that separates a serial number region and a background region from an original image of a paper currency that includes a serial number; a character image acquisition unit that identifies each of characters included in the separated serial number region, and acquires a character image of each of the identified characters; a background image acquisition unit that acquires a background image by complementing the serial number region in the separated background region; a pre-processing unit that generates a serial number image by combining the character images; an incorporation unit that incorporates the serial number image at a position corresponding to the serial number image in the background image; and an output unit that outputs image data in which the serial number that is combined by the pre-processing unit is associated with an incorporated image generated by the incorporation unit.

A serial number recognition parameter determination apparatus includes: a generation unit, an identification unit, and an evaluation index calculation unit. The generation unit generates a parameter set of a program, the program being used when a paper sheet handing apparatus identifies, from an image of a paper sheet, character present regions in which characters that form a serial number are present. The identification unit identifies, from an image of the paper sheet, the character present regions by using the parameter set that is generated by the generation unit. The evaluation index calculation unit calculates an evaluation index of the parameter set based on the character present regions that are identified by the identification unit.

The present disclosure relates to a technique of generating a disclosable document image based on a document image including confidential information, without using the confidential information. A document input unit obtains a document image scanned with a scanner, separates the document image into character information and background information, and then outputs them to an extraction unit. The extraction unit performs named entity extraction processing on the obtained character information and background information to extract named entities in the document and attributes thereof, and output an extraction result to a generation unit. The generation unit replaces the named entities in the document image with attribute tags and obtains superimposable ranges to generate attribute tag document data. A management unit registers the received extraction result of the named entities and the attribute tag document data in a database.