Wyniki wyszukiwania - BazTech

1

Efektywność sztucznych sieci neuronowych w rozpoznawaniu znaków pisma odręcznego

Miłosz Marek, Gazda Janusz

Journal of Computer Sciences Institute

|

2018

|

Vol. 7

210--214

PL

Sztuczne sieci neuronowe są jednym z narzędzi współczesnych systemów odtwarzania z obrazów tekstów, w tym pisanych odręcznie. W artykule przedstawiono rezultaty eksperymentu obliczeniowego mającego na celu analizę jakości rozpoznawania cyfr pisanych odręcznie przez dwie sztuczne sieci neuronowe (SSN) o różnej architekturze i parametrach. Jako podstawowe kryterium jakości rozpoznawania znaków użyto wskaźnika poprawności. Poza tym analizie poddano liczbę neuronów i ich warstw oraz czas uczenia SSN. Do stworzenia SSN, oprogramowania algorytmów ich uczenia i testowania wykorzystano język Python i bibliotekę TensorFlow. Obydwie SSN uczono i testowano przy pomocy tych samych dużych zbiorów obrazów znaków pisanych odręcznie.

EN

Artificial neural networks are one of the tools of modern text recognising systems from images, including handwritten ones. The article presents the results of a computational experiment aimed at analyzing the quality of recognition of handwritten digits by two artificial neural networks (ANNs) with different architecture and parameters. The correctness indicator was used as the basic criterion for the quality of character recognition. In addition, the number of neurons and their layers and the ANNs learning time were analyzed. The Python language and the TensorFlow library were used to create the ANNs, and software for their learning and testing. Both ANNs were learned and tested using the same big sets of images of handwritten characters.

2

The Impact of Filters on the Quality of Binarization for Handwriting Images

Adamski M., Pryzmont K.

Advances in Computer Science Research

|

2018

|

Nr 14

17--30

EN

Filtration and binarization techniques are often used in handwriting recognition systems. These operations are performed as part of a stage called preprocessing, the result of which is passed to feature extraction and classification processes. Operations performed as part of the preprocessing are important because their result affects the outcome of the entire system. This paper focuses on the assessment of filtration techniques influence on the binarization of handwriting images. In the experiments, four filtration methods were tested with seven thresholding algorithms for various combinations of filtration and binarization parameters. The experiments were conducted on handwriting images selected from Document Binarization Competitions (DIBCO) datasets with ground truth images for the assessment of binarization correctness. The final evaluation was conducted based on the average of quality measures: F-measure, Accuracy, Relative Foreground Area Error and Region Nonuniformity.

PL

Filtracja i binaryzacja są często stosowanymi technikami w systemach rozpoznawania pisma odręcznego. Operacje te są wykonywane w ramach etapu zwanego wstępnym przetwarzaniem, którego rezultat jest przekazywany do kolejnych etapów: ekstrakcji cech i klasyfikacji. Operacje wykonywane w ramach wstępnego przetwarzania są istotne ponieważ ich wyniki wpływają na poprawność pracy całego systemu. W niniejszej pracy skupiono się nad oceną wpływu wyboru metody filtracji obrazu na efekt procesu binaryzacji dla obrazów z pismem odręcznym. W eksperymentach zbadano 4 metody filtracji w połączeniu z 7 metodami progowania dla różnych kombinacji parametrów tych metod. Do eksperymentów użyto wybrane obrazy z pismem odręcznym z baz konkursów binaryzacji dokumentów DIBCO oraz obrazy referencyjne do oceny poprawności binaryzacji. Ocenę wykonano na bazie średniej z miar F-measure, Accuracy, Relative Foreground Area Error, Region nonuniformity.

3

Analysis of the Arabic using neural networks: an overview

Soori H., Platos J., Snasel V.

Przegląd Elektrotechniczny

|

2013

|

R. 89, nr 11

47-50

EN

This paper is a quick review of some of the scholarly work aiming at solving various problems of the Arabic language using neural networks. It includes some research work concerning online recognition of handwritten Arabic characters, speech recognition, offline character text recognition, text categorization and recognition of printed text. This paper concludes that more research should be conducted in this area considering the importance of the Arabic language, the rapid growth of internet users in the Arab world, and the widespread usage of Arabic characters by many languages other than Arabic.

PL

W artykule przedstawiono metody analizy języka arabskiego z wykorzystaniem sieci neuronowych. Analizowano możliwości rozpoznawania pisma odręcznego, drukowanego jak i mowy.

4

Efficient online handwritten Chinese character recognition system using a two-dimensional functional relationship model

Chang Y. F., Lee J. C., Mohd Rijal O., Syed Abu Bakar S. A. R.

International Journal of Applied Mathematics and Computer Science

|

2010

|

Vol. 20, no 4

727-738

EN

This paper presents novel feature extraction and classification methods for online handwritten Chinese character recognition (HCCR). The X-graph and Y-graph transformation is proposed for deriving a feature, which shows useful properties such as invariance to different writing styles. Central to the proposed method is the idea of capturing the geometrical and topological information from the trajectory of the handwritten character using the X-graph and the Y-graph. For feature size reduction, the Haar wavelet transformation was applied on the graphs. For classification, the coefficient of determination [...] from the two-dimensional unreplicated linear functional relationship model is proposed as a similarity measure. The proposed methods show strong discrimination power when handling problems related to size, position and slant variation, stroke shape deformation, close resemblance of characters, and non-normalization. The proposed recognition system is applied to a database with 3000 frequently used Chinese characters, yielding a high recognition rate of 97.4% with reduced processing time of 75.31%, 73.05%, 58.27% and 40.69% when compared with recognition systems using the city block distance with deviation (CBDD), the minimum distance (MD), the compound Mahalanobis function (CMF) and the modified quadratic discriminant function (MQDF), respectively. High precision rates were also achieved.

5

O Ukrytych Modelach Markowa i ich zastosowaniu do rozpoznawania ciągów znaków pisma odręcznego

Klęsk P.

Metody Informatyki Stosowanej

|

2008

|

nr 1 (Tom 13)

175-189

EN

In the paper, selected informations on Hidden Markov Models (also called Hidden Markov Chains) are reminded. Basic riotions are defined and algorithms related to these models are shortly presented. The research part of the papers shows results of three conducted experiments entitled: "pork cutlet", "form sheet" and ''poetaster". The most important experiment "form sheet" gives a good starting point to a practical application of HMMs to the. handwriting recognition. The "poetaster" experiment shows possible application of HMMs in so called "artifial creation".