Annotated comparisons of proposed preprocessing techniques for script recognition

Saba, Tanzila; Rehman, Amjad; Altameem, Ayman; Uddin, Mueen

doi:10.1007/s00521-014-1618-9

Cited by 49 publications

(26 citation statements)

References 51 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To remove unwanted information present in acquired images, different preprocessing methods are used (Saba, Almazyad, & Rehman, ; Saba, Rehman, Al‐Dhelaan, & Al‐Rodhaan, ; Saba, Rehman, Altameem, & Uddin, ; Saba, Rehman, & Sulong, ). Tahir et al () conducted a comprehensive study of preprocessing techniques and their effect on feature extraction from medical images.…”

Section: Proposed Methodologymentioning

confidence: 99%

Deep learning model integrating features and novel classifiers fusion for brain tumor segmentation

Iqbal

Khan

Saba

et al. 2019

Microscopy Res & Technique

Self Cite

View full text Add to dashboard Cite

Automatic and precise segmentation and classification of tumor area in medical images is still a challenging task in medical research. Most of the conventional neural network based models usefully connected or convolutional neural networks to perform segmentation and classification. In this research, we present deep learning models using long short term memory (LSTM) and convolutional neural networks (ConvNet) for accurate brain tumor delineation from benchmark medical images. The two different models, that is, ConvNet and LSTM networks are trained using the same data set and combined to form an ensemble to improve the results. We used publicly available MICCAI BRATS 2015 brain cancer data set consisting of MRI images of four modalities T1, T2, T1c, and FLAIR. To enhance the quality of input images, multiple combinations of preprocessing methods such as noise removal, histogram equalization, and edge enhancement are formulated and best performer combination is applied. To cope with the class imbalance problem, class weighting is used in proposed models. The trained models are tested on validation data set taken from the same image set and results obtained from each model are reported. The individual score (accuracy) of ConvNet is found 75% whereas for LSTM based network produced 80% and ensemble fusion produced 82.29% accuracy.

show abstract

Section: Proposed Methodologymentioning

confidence: 99%

Deep learning model integrating features and novel classifiers fusion for brain tumor segmentation

Iqbal

Khan

Saba

et al. 2019

Microscopy Res & Technique

Self Cite

View full text Add to dashboard Cite

show abstract

“…This image will be having a noise that makes the OCR to erroneously recognize the characters [15]. This is rectified by applying some preprocessing techniques to smoothen or sharpen the image and to remove the noise in it [16]. At first the image that is taken is converted into black and white using preprocessing methods and then only segmentation is carried out [17].…”

Section: Proposed Methodsmentioning

confidence: 99%

Identification of Broken Characters in Degraded Documents

Ramalingam¹,

Bhojan²

2018

IJIES

View full text Add to dashboard Cite

Optical Character Recognition (OCR) deals with the recognition of characters in a text document. Steps like Preprocessing, Segmentation and Recognition are embedded in the OCR machine. When a document is scanned it will be taken into OCR and will recognize the characters. But noisy scanning of documents, low-quality printed documents and thresholding error leads to the generation of broken characters. When these documents are given as inputs into OCR, the recognition becomes a tedious process since the broken characters are misunderstood by the OCR machine. So the broken characters have to be identified and segmented separately. This work aims to enhance the degraded documents with broken characters using image processing techniques. For identifying or recognizing the broken character from the image various techniques like vertical projection profile, horizontal projection profile, chain code, mean based thresholding are used. The lines from the document are separated using line segmentation. Separate characters are extracted using Vertical Projection Profile and Horizontal Projection Profile. The character is identified using chain coding. The broken characters are found from them using Mean-based Thresholding and is merged using Heuristic information. The proposed method achieves an accuracy of 92.88% and also performs well for color image documents as well as black and white image documents also because of the effective preprocessing.

show abstract

“…The statistical features include texture measurements, grapheme distributions, grey-level statistics and crosscorrelation distributions. These features are normally extracted from handwriting fragments and are effectively employed for classification (Hussain et al, 2017;Saba et al, 2014c;.…”

Section: Research Backgroundmentioning

confidence: 99%

Fuzzy ARTMAP Approach for Arabic Writer Identification using Novel Features Fusion

Saba

2018

Journal of Computer Science

Self Cite

View full text Add to dashboard Cite

Arabic writer identification and associated tasks are still fresh due to huge variety of Arabic writer's styles. Current research presents a fusion of statistical features, extracted from fragments of Arabic handwriting samples to identify the writer using fuzzy ARTMAP classifier. Fuzzy ARTMP is supervised neural model, especially suited to classification problems. It is faster to train and need less number of training epochs to "learn" from input data for generalization. The extracted features are fed to Fuzzy ARTMP for training and testing. Fuzzy ARTMAP is employed for the first time along with a novel fusion of statistical features for Arabic writer identification. The entire IFN/ENIT database is used in experiments such that 75% handwritten Arabic words from 411 writers are employed in training and 25% for testing the system at random. Several combinations of extracted features are tested using fuzzy ARTMAP classifier and finally one combination exhibited promising accuracy of 94.724% for Arabic writer identification on IFN/ENIT benchmark database.

show abstract

Annotated comparisons of proposed preprocessing techniques for script recognition

Cited by 49 publications

References 51 publications

Deep learning model integrating features and novel classifiers fusion for brain tumor segmentation

Deep learning model integrating features and novel classifiers fusion for brain tumor segmentation

Identification of Broken Characters in Degraded Documents

Fuzzy ARTMAP Approach for Arabic Writer Identification using Novel Features Fusion

Contact Info

Product

Resources

About