In a character string reading device, an imaging part obtains an image of a read object, and an output format setting part sets an output format of a character string to be read from the image and to be output. A character recognition condition setting part sets a character recognition condition, that only a group of characters including all characters defined by the output format are to be identified, among characters which the character string reading device can identify. A character string recognizing part recognizes a character string in the image according to the character recognition condition. A character extracting part obtains a character string for output, at a portion matching the output format among the recognized character string.

A handwriting geometry recognition and calibration system by using neural network and mathematical feature includes: a pre-processor for pre-processing coordinate points of geometric figures from user's handwriting so as to get a plurality of sample points which expresses the geometric figures to be recognized; a neural network connected to the pre-processor for receiving the sample points of the geometric figure and recognizing the geometric figure so as to acquire a coarse class of the geometric figure; and an mathematical logic unit connected to the neural network for receiving recognition results from the neural network, including coarse classifications which are used in a secondary classification by using conventional mathematical recognition logics so as to determine an exact geometry shape of the geometric figure; then the geometric figure being calibrated so as to get a geometry with a regular shape.

A document processing system processes a document image to identify document image regions including floating images, structured data units, and unstructured floating text. A first masked image is generated by deleting any floating images from the document image and a second masked image is generated by deleting any structured data units from the first masked image. The structured data units and the unstructured floating text are thus identified serially one after another. Textual data is extracted from the structured data units and the unstructured floating text by processing the corresponding document image regions via optical character recognition (OCR). Entities are extracted from the textual data using natural language processing (NLP) techniques.

Methods, systems, and apparatuses for image segmentation are provided. For example, a computing device may obtain an image, and may apply a process to the image to generate input image feature data and input image segmentation data. Further, the computing device may obtain reference image feature data and reference image classification data for a plurality of reference images. The computing device may generate reference image segmentation data based on the reference image feature data, the reference image classification data, and the input image feature data. The computing device may further blend the input image segmentation data and the reference image segmentation data to generate blended image segmentation data. The computing device may store the blended image segmentation data within a data repository. In some examples, the computing device provides the blended image segmentation data for display.

Computer-readable media, methods, and systems are disclosed for applying machine learning mechanisms to classify and validate documents based on expense rule sets and external data validation services. Document images associated with expenses are received in connection with a reimbursable event. For each received document image data associated with the received document image is transmitted to an optical character recognition image processor that can recognize contents and associated coordinates. OCR data is received and transmitted to a text tokenizer. Tokenized text is received corresponding to expense details, and the tokenized text and coordinates are sent to a text feature generator. Text feature vectors are received and transmitted to a document classifier and a document classification received. Document fields are extracted and based thereon a document is validates and a corresponding reimbursement instruction generated.

A method for recognizing handwritten text is disclosed. The method comprises receiving data comprising a sequence of ink points; applying the received data to a neural network-based sequence classifier trained with a Connectionist Temporal Classification (CTC) output layer using forced alignment to generate an output; generating a character hypothesis as a portion of the sequence of ink points; applying the character hypothesis to a character classifier to obtain a first probability corresponding to the probability that the character hypothesis includes the given character; processing the output of the CTC output layer to determine a second probability corresponding to the probability that the given character is observed within the character hypothesis; and combining the first probability and the second probability to obtain a combined probability corresponding to the probability that the character hypothesis includes the given character.

A method for producing android device test reproducible on any android device, comprising: receiving a previously recorded android test video file and processing the video file to extract video frames; searching for touch coordinates in the video frames; identifying the touch coordinates in the video frames and generating touch coordinate groups. The method includes translating touch coordinate groups into android actions using heuristic rules; recognizing and classifying widgets in the android actions video frames; generating a description for each of the recognized and classified widgets; associating a widget with each android action; generating a user-readable test step text file and a test step file with detailed information for each step, and iteratively founding the most similar widget on the device under test screen when compared with the human-readable described step at each timestamp at execution time.

A method for reading an optical prescription on an optical prescription image. The method includes detecting a region comprising the optical prescription on the optical prescription image; extracting the optical prescription and converting the optical prescription into machine-encoded optical prescription data; classifying a portion of the optical prescription data into one or more predetermined categories, to generate an optical prescription value associated with a respective one of the one or more predetermined categories; and determining whether the optical prescription value associated with the respective one of the one or more predetermined categories contains an error, and, if the optical prescription value contains the error, correcting the error within the optical prescription value, to generate a corrected optical prescription value associated with the respective one of the one or more predetermined categories. A system for reading an optical prescription on an optical prescription is also disclosed.

An example computing platform is configured to: obtain a two-dimensional drawing of a portion of a construction project; perform an image processing analysis of the two-dimensional drawing to identify one or more location entities within the two-dimensional drawing; derive embeddings for each location entity in the two-dimensional drawing; based on the derived embeddings, determine relationships between the one or more location entities; and based on the determined relationships between the one or more location entities, generate a location entity data taxonomy that includes each identified location entity as a respective node that is related to at least one other location entity.