No description available.

Background: To preserve valuable sources and cultural heritage, digitization of handwritten characters is crucial. For this, Optical Character Recognition (OCR) systems were introduced and most widely used to recognize digital characters. Incase of ancient or historical characters, automatic transcription is more challenging due to lack of data, high complexity and low quality of the resource. To solve these problems, multiple image based handwritten dataset were collected from historicaland modern document images. But these dataset also have some limitations. To overcome the limitations, we were inspired to create a new image-based historical handwritten character and word dataset and evaluate it’s performance using machine learning algorithms. Objectives: The main objective of this thesis is to create a first ever Swedish historical handwritten character and word dataset named CArDIS (Character Arkiv Digital Sweden) which will be publicly available for further research. In addition,verify the correctness of the dataset and perform a quantitative analysis using different machine learning methods. Methods: Initially we searched for existing character dataset to know how modern character dataset differs from the historical handwritten dataset. We have performed literature review to learn about most commonly used dataset for OCR. On the other hand, we have also studied different machine learning algorithms and their applica-tions. Finally, we have trained six different machine learning methods namely Support Vector Machine, k-Nearest Neighbor, Convolutional Neural Network, Recurrent Neural Network, Random Forest, SVM-HOG with existing dataset and newly created dataset to evaluate the performance and efficiency of recognizing ancient handwritten characters. Results: The performance/evaluation results show that the machine learning classifiers struggle to recognise the ancient handwritten characters with less recognition accuracy. Out of which CNN outperforms with highest recognition accuracy. Conclusions: The current thesis introduces first ever newly created historical hand-written character and word dataset in Swedish named CArDIS. The character dataset contains 1,01,500 Latin and Swedish character images belonging to 29 classes while the word dataset contains 10,000 word images containing ten popular Swedish names belonging to 10 classes in RGB color space. Also, the performance of six machine learning classifiers on CArDIS and existing datasets have been reported. The thesis concludes that classifiers when trained on existing dataset and tested on CArDIS dataset show low recognition accuracy proving that, the CArDIS dataset have unique characteristics and features over the existing handwritten datasets. Finally, this re-search provided a first Swedish character and word dataset, which is robust with a proven accuracy; also it is publicly available for further research.

No description available.

Hantering av ostrukturerade och fysiska dokument inom vissa områden, såsom finansiell rapportering, medför betydande ineffektivitet i dagsläget. Detta examensarbete fokuserar på utmaningen att extrahera data från ostrukturerade finansiella dokument, specifikt balansräkningar i svenska årsredovisningar, genom att använda en AI-driven pipeline. Syftet är att utveckla en metod för att automatisera datautvinning och möjliggöra förbättrad dataanalys. Projektet fokuserade på att automatisera utvinning av finansiella poster från balansräkningar genom en kombination av Optical Character Recognition (OCR) och en modell för Named Entity Recognition (NER). TesseractOCR användes för att konvertera skannade dokument till digital text, medan en BERT-baserad NER-modell tränades för att identifiera och klassificera relevanta finansiella poster. Ett Python-skript användes för att extrahera de numeriska värdena som är associerade med dessa poster. Projektet fann att NER-modellen uppnådde hög prestanda, med ett F1-score på 0,95, vilket visar dess effektivitet i att identifiera finansiella poster. Den fullständiga pipelinen lyckades extrahera över 99% av posterna från balansräkningar med en träffsäkerhet på cirka 90% för numerisk data. Projektet drar slutsatsen att kombinationen av OCR och NER är en lovande lösning för att automatisera datautvinning från ostrukturerade dokument med liknande attribut som årsredovisningar. Framtida arbeten kan utforska att förbättra träffsäkerheten i OCR och utvidga utvinningen till andra sektioner av olika typer av ostrukturerade dokument. / The persistence of unstructured and physical document management in fields such as financial reporting presents notable inefficiencies. This thesis addresses the challenge of extracting valuable data from unstructured financial documents, specifically balance sheets in Swedish annual reports, using an AI-driven pipeline. The objective is to develop a method to automate data extraction, enabling enhanced data analysis capabilities. The project focused on automating the extraction of financial posts from balance sheets using a combination of Optical Character Recognition (OCR) and a Named Entity Recognition (NER) model. TesseractOCR was used to convert scanned documents into digital text, while a fine-tuned BERT-based NER model was trained to identify and classify relevant financial features. A Python script was employed to extract the numerical values associated with these features. The study found that the NER model achieved high performance metrics, with an F1-score of 0.95, demonstrating its effectiveness in identifying financial entities. The full pipeline successfully extracted over 99% of features from balance sheets with an accuracy of about 90% for numerical data. The project concludes that combining OCR and NER technologies could be a promising solution for automating data extraction from unstructured documents with similar attributes to annual reports. Future work could explore enhancing OCR accuracy and extending the methodology to other sections of different types of unstructured documents.

Thesis (PhD (Mathematical Sciences))--University of Stellenbosch, 2005. / A great deal of work has been done in the area of off-line signature verification over the past two decades. Off-line systems are of interest in scenarios where only hard copies of signatures are available, especially where a large number of documents need to be authenticated. This dissertation is inspired by, amongst other things, the potential financial benefits that the automatic clearing of cheques will have for the banking industry.

In Chinese orthography, phonetic compounds comprise about 80% of the most frequent characters. They contain separate phonological and semantic elements, referred to as phonetic and semantic radicals respectively. A dominant type exists in which the se-mantic radical appears on the left and the phonetic radical on the right (SP characters); an opposite, minority structure also exists in which the semantic radical appears on the right and the phonetic radical on the left (PS characters). Through statistical analyses, connectionist modelling, behavioural experiments, and neuroimaging studies, this dis-sertation demonstrates that the distinct structures of these two types of characters allow us crucial insights into the relationship between brain structure and reading processes. The statistical analyses of a Chinese lexical database show that, because of the different information profiles of SP and PS characters and the imbalanced distribution between them in the lexicon, the overall information is skewed to the right. This information skew provides important opportunities to examine the interaction between foveal split-ting and the information structure of the characters. The foveal splitting hypothesis as-sumes a vertical meridian split in the foveal representation and the consequent contra-lateral projection to the two cerebral hemispheres; it has been shown to have important implications for visual word recognition. The square shape and the condensed structure of Chinese characters make them a severe test case for the split fovea claim. Through a lateralized cueing examination and a TMS study of the semantic radical combinability effect with foveally presented characters in character semantic judgements, a flexible division of labour between the hemispheres in character recognition is demonstrated, with each hemisphere responding optimally to the information in the contralateral visual hemifield. The interaction between stimulation site and radical combinability in the TMS study also provides further support for the split fovea claim, suggesting functional foveal splitting as a universal processing constraint in reading. Even if foveal splitting is true, it is still unclear about how far the effects of foveal split-ting can extend from the retina into the process of character recognition. We show that, in naming isolated, foveally presented SP and PS characters, adult male and female readers process them differently, with opposite patterns of ease and difficulty: males responded significantly faster to SP than PS characters; females showed a non-significant tendency in the opposite direction. This result is also supported by a corre-sponding ERP study showing larger N350 amplitude elicited by PS character than SP characters in the male brain, and an opposite pattern in the female brain. The split fovea claim suggests that the two halves of a centrally fixated character are initially processed in different hemispheres. The male brain typically relies more on the left hemisphere for phonological processing compared with the female brain, causing this gender difference to emerge. This interaction is also predicted by an implemented computational model, contrasting a split cognitive architecture, in which the mapping between orthography to phonology is mediated by two partially encapsulated, interconnected processing do-mains, and a non-split cognitive architecture, in which the mapping is mediated by a single, undifferentiated processing domain. Thus, the effects of foveal splitting in read-ing extend far enough to interact with the gender of the reader in a naturalistic reading task. In short, this dissertation demonstrates that foveal splitting is a universal language proc-essing phenomenon, precise enough to project the two radicals of a centrally-fixated Chinese character to different hemispheres to allow a flexible division of labour be-tween the two hemispheres to emerge, and its effects in reading extend far enough into word recognition to interact with the gender of the reader in a naturalistic reading task. The results can also be extrapolated to Chinese word and sentence processing as well as to other languages. This dissertation thus has contributed to a better understanding of the relationship between brain structure and language processes.

Optical character recognition (OCR) software has advanced greatly in recent years. Machine-printed text can be scanned and converted to searchable text with word accuracy rates around 98%. Reasonably neat hand-printed text can be recognized with about 85% word accuracy. However, cursive handwriting still remains a challenge, with state-of-the-art performance still around 75%. Algorithms based on hidden Markov models have been only moderately successful, while recurrent neural networks have delivered the best results to date. This thesis explored the feasibility of using a special type of feedforward neural network to convert freeform cursive handwriting to searchable text. The hidden nodes in this network were grouped into clusters, with each cluster being trained to recognize a unique character bigram. The network was trained on writing samples that were pre-segmented and annotated. Post-processing was facilitated in part by using the network to identify overlapping bigrams that were then linked together to form words and sentences. With dictionary assisted post-processing, the network achieved word accuracy of 66.5% on a small, proprietary corpus. The contributions in this thesis are threefold: 1) the novel clustered architecture of the feed-forward neural network, 2) the development of an expanded set of observers combining image masks, modifiers, and feature characterizations, and 3) the use of overlapping bigrams as the textual working unit to assist in context analysis and reconstruction.

We present solution of steps necessary for binarization and text lines detection contained in printed documents digitized by the camera. We introduce a normalization of non-uniform illumination method for text photographs. We propose input bitmap binarization algorithm based on two-dimensional probability pixel model which also considers its surrounding. We continue with description of robust text lines orientation detector based on optimization of risk function using first order derivatives of image function. In the end we present text lines detection and segmentation algorithm. Final shape of segmented lines is optimized with usage of graph algorithm. Powered by TCPDF (www.tcpdf.org)

M.Com. (Informatics) / Please refer to full text to view abstract

Thesis (Ph.D.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 1979. / MICROFICHE COPY AVAILABLE IN ARCHIVES AND ENGINEERING. / Vita. / Bibliography: leaves 365-385. / by Theodore Thomas Kuklinski. / Ph.D.