Wipro Limited

Centre for Development of Advanced Computing (c-dac)

Swapnil Dnyaneshwar Belhe

Aniket Anand Gurav

Rupesh Wadibhasme

Pravin Pawar

Zaheer Juzer Javi

Vinodkumar Tejlal Pache

Shishir Vivek Bhaskarwar

Umesh Hareshwar Vithalkar

Mahesh Dattatray Kulkarni

Ajay Sharadchandra Lohokare

Vishal Prakash Chavan

Disclosed herein is a method and system for determining quality of an input document during risk and compliance assessment. The method includes receiving input document for risk and compliance assessment, identifying a document type, and at least one sub-type of the input document using a Natural Language Processing (NLP) technique and a trained neural network model. Layout of content present in the input document is detected based on each of a plurality of segments extracted from content and structural parameters associated with respective segments. A document review model is identified from a plurality of document review models based on type and at least one sub-type of input document. Thereafter, the quality of the input document and a compliance score is determined by identifying one or more deviations of content of the input document from content of a predefined template for the input document.

Method and system for determining quality of a document during risk and compliance assessment

Disclosed herein is method and a system for extracting information from an input document comprising multi-format information. In an embodiment, a Hypertext Markup Language (HTML) document corresponding to the input document is created by analyzing the input document comprising documents of multiple data formats. Further, the HTML document is realigned based on a number of columns in each page of the HTML document. Furthermore, a document Identifier (ID) associated with each of the documents is determined in realigned HTML document by classifying information in each of the document pages using a pretrained Machine Learning (ML) model. Subsequently, a hierarchy configuration file, corresponding to the realigned HTML document, is generated based on the document ID. Finally, information from the hierarchy configuration file associated with each of the document ID is extracted by orchestrating one or more data extractors for extracting data attributes from the hierarchy configuration file.

Method and system for extracting information from input document comprising multi-format information

The present disclosure is related to field of machine learning and image processing, disclosing method and device for identifying cell region of table including cell borders from an image document. Table detecting system rescales a primary image document into plurality of secondary image documents of different size and resolution to detect plurality of candidate regions comprising predefined table features in each secondary image document. Further, for each candidate region, set of connected components are determined and the connected components corresponding to the IDs that are present in more than one set of the connected components are clustered. Subsequently, areas corresponding to the clusters that are determined to form a table are cropped from the primary image document and each cell region of the table is identified by modifying pixel values of the clusters of the connected components in the cropped area.

Method and system for identifying cell region of table comprising cell borders from image document

Examples of present disclosure relate to method and system to extract text from engineering drawing for performing accurate OCR. Initially, for the extraction, image of engineering drawing is received with a plurality of components. Each of the plurality of components in the image is classified to be one of a textual component and a non-textual component. At least one word element for textual components from the plurality of components is identified based on segmentation of the plurality of components. The segmentation is performed by drawing a plurality of horizontal edge projections of a predefined length for each of the textual components. Further, the textual components is identified to be associated with the at least one word element when horizontal edge projection of each of the textual components overlaps with adjacent textual component. The at least one word element is provided as extracted text for performing OCR on the engineering drawing.

Method and system for extracting text from an engineering drawing

The present disclosure relates to system(s) and method(s) for compression and decompression of Unicode characters. The system is configured to maintain a set of character tables and a cluster table in a memory. Each character table is configured to store a set of Unicode characters corresponding to a character class of a set of characters classes, wherein each Unicode character from the character table is assigned with a shortened bit representation. Furthermore, the cluster table may be configured to maintain a set of cluster types and a cluster identifier corresponding to each of the cluster type. The system is configured to compress text data in Unicode format using the set of character tables and the cluster table by identifying the different clusters in each word and replacing the clusters with cluster identifier followed by the shorten bit representation of characters in each cluster.

Growing community of inventors

Pune, India

Swapnil Dnyaneshwar Belhe