An image forming apparatus includes: a document reading section; a character detection section detecting, based on image data of a document obtained through reading by the document reading section, characters included in an image formed on the document; a character concentration detection section detecting concentration of the characters detected by the character detection section; a character direction detection section detecting a direction of the characters whose concentration detected by the character concentration detection section is in a preset specified concentration range; and a document direction determination section determining, based on the direction of the characters detected by the character direction detection section, a direction of the image formed on the document as a document direction, wherein the character direction detection section, upon determination that the image on the document is a monochromatic image, defines, as the specified concentration range, a concentration range higher than predefined first concentration.

A method for information compression/decompression, apparatuses and a non-transitory computer-readable storage medium are disclosed. The method for information compression may include: clustering text blocks to be processed into respective-text areas according to pixel distribution information of the text blocks to be processed; acquiring text row distribution information of each of the text areas according to foreground pixels of each text row in each of the text areas; scanning each text row in each of the text areas according to the acquired text row distribution information to acquire original pixel information of each text row; and performing lossless compression on the text row distribution information of a plurality of the text areas and the original pixel information of each text row of the plurality of the text areas.

A method for a device to perform region recognition is provided. The method includes: obtaining a position of a face region in an identification image; determining at least one information region based on the position of the face region; and segmenting the information region to obtain at least one character region.

An image processing device may acquire original image data representing an original image including M lines of original character strings, and combine M items of original character string image data representing the M lines of original character strings so as to create target character string image data representing one line of target character string. The image processing device may divide the target character string image data so as to create N items of divided character string image data representing N lines of divided character strings into which the one line of target character string is divided, and create output image data representing the N lines of divided character strings by using the N items of divided character string image data. A number of characters of a leading line of divided character string may be different from a number of characters of a leading line of original character string.

A supplemental device (2) for attachment to a pen-type drug injection device, the supplemental device comprising an optical sensor array or camera (25) configured to capture an image of one or more numbers representing dose values on a dose dial sleeve of the injection device, and a processor (24) and memory (240,241) arrangement configured to run an algorithm to receive image data from the sensor/camera and extract the dialed dose value from said image via optical character recognition (OCR). The algorithm is able to determine the set dose value which is at the centre of the image (902), also if the dose scale only comprises integer values in steps of two (see FIG. 9).

An example method of processing images of deformed indicia-bearing surfaces includes: detecting, within a document image, a plurality of image fragments, wherein each image fragment of the plurality of image fragment contains a respective sequence of alphabet symbols; grouping the plurality of image fragments by lines of text to be reconstructed in the document image; generating a map of isolines associated with the document image, wherein an isoline identifies a set of points that lie on a straight line of an undistorted image corresponding to the document image; generating a reverse transformation matrix that defines a set of transformations to be applied to the document image in order to remove image distortions caused by deformations of an indicia bearing surface; and generating an undistorted document image by applying the reverse transformation matrix to the document image.

Machine learning based computer devices, systems and methods are proposed for automating the evaluation and visual testing of graphical user interface (GUI) designs using a combination of image transformations for scoring the GUI designs and machine learning data architectures with a set of logical and conditional rules. The approach describes an automated process that transforms the GUI designs into clusters of pixels before using a chained series of image transformations to obtain similarity scores and underlying distributions for the GUI designs and then uses a machine learning data architecture in combination with a set of logical and conditional rules to computationally generate a prediction of error estimates based on the underlying distributions of the GUI designs.