Hanaa Fathi Mahmood - Profile on Academia.edu (original) (raw)

Uploads

Papers by Hanaa Fathi Mahmood

Text-detection and -recognition from natural images

Text detection and recognition from images could have numerous functional applications for docume... more Text detection and recognition from images could have numerous functional applications for document analysis, such as assistance for visually impaired people; recognition of vehicle license plates; evaluation of articles containing tables, street signs, maps, and diagrams; keyword-based image exploration; document retrieval; recognition of parts within industrial automation; content-based extraction; object recognition; address block location; and text-based video indexing. This research exploited the advantages of artificial intelligence (AI) to detect and recognise text from natural images. Machine learning and deep learning were used to accomplish this task.In this research, we conducted an in-depth literature review on the current detection and recognition methods used by researchers to identify the existing challenges, wherein the differences in text resulting from disparity in alignment, style, size, and orientation combined with low image contrast and a complex background make automatic text extraction a considerably challenging and problematic task. Therefore, the state-of-the-art suggested approaches obtain low detection rates (often less than 80%) and recognition rates (often less than 60%). This has led to the development of new approaches. The aim of the study was to develop a robust text detection and recognition method from natural images with high accuracy and recall, which would be used as the target of the experiments. This method could detect all the text in the scene images, despite certain specific features associated with the text pattern. Furthermore, we aimed to find a solution to the two main problems concerning arbitrarily shaped text (horizontal, multi-oriented, and curved text) detection and recognition in a low-resolution scene and with various scales and of different sizes.In this research, we propose a methodology to handle the problem of text detection by using novel combination and selection features to deal with the classification algorithms of the text/non-text regions. The text-region candidates were extracted from the grey-scale images by using the MSER technique. A machine learning-based method was then applied to refine and validate the initial detection. The effectiveness of the features based on the aspect ratio, GLCM, LBP, and HOG descriptors was investigated. The text-region classifiers of MLP, SVM, and RF were trained using selections of these features and their combinations. The publicly available datasets ICDAR 2003 and ICDAR 2011 were used to evaluate the proposed method. This method achieved the state-of-the-art performance by using machine learning methodologies on both databases, and the improvements were significant in terms of Precision, Recall, and F-measure. The F-measure for ICDAR 2003 and ICDAR 2011 was 81% and 84%, respectively. The results showed that the use of a suitable feature combination and selection approach could significantly increase the accuracy of the algorithms.A new dataset has been proposed to fill the gap of character-level annotation and the availability of text in different orientations and of curved text. The proposed dataset was created particularly for deep learning methods which require a massive completed and varying range of training data. The proposed dataset includes 2,100 images annotated at the character and word levels to obtain 38,500 samples of English characters and 12,500 words. Furthermore, an augmentation tool has been proposed to support the proposed dataset. The missing of object detection augmentation tool encroach to proposed tool which has the ability to update the position of bounding boxes after applying transformations on images. This technique helps to increase the number of samples in the dataset and reduce the time of annotations where no annotation is required. The final part of the thesis presents a novel approach for text spotting, which is a new framework for an end-to-end character detection and recognition system designed using an improved SSD convolutional neural network, wherein layers are added to the SSD networks and the aspect ratio of the characters is considered because it is different from that of the other objects. Compared with the other methods considered, the proposed method could detect and recognise characters by training the end-to-end model completely. The performance of the proposed method was better on the proposed dataset; it was 90.34. Furthermore, the F-measure of the method’s accuracy on ICDAR 2015, ICDAR 2013, and SVT was 84.5, 91.9, and 54.8, respectively. On ICDAR13, the method achieved the second-best accuracy. The proposed method could spot text in arbitrarily shaped (horizontal, oriented, and curved) scene text.

مجلة التربية والعلم, 1970

Text detection and recognition from images have numerous applications for document analysis and i... more Text detection and recognition from images have numerous applications for document analysis and information retrieval tasks. An accurate and robust method for detecting texts in natural scene images is proposed in this paper. Text-region candidates are detected using maximally stable extremal regions (MSER) and a machine learning based method is then applied to refine and validate the initial detection. The effectiveness of features based on aspect ratio, GLSM, LBP, HOG descriptors are investigated. Text-region classifiers of MLP, SVM and RF are trained using selections of these features and their combination. A publicly available multilingual dataset ICDAR 2003,2011 has been used to evaluate the method. The proposed method achieved excellent performance on both databases and the improvements are significant in terms of Precision, Recall, and F-measure. The results show that using a suitable feature combination and selection approach can can significantly increase the accuracy of the algorithms.

مجلة التربية والعلم, Jun 1, 2009

This paper proposes a new algorithm for recognizing text object images by using fractal geometry.... more This paper proposes a new algorithm for recognizing text object images by using fractal geometry. The fractal dimension was used as a main feature for recognizing text objects within images. Box-counting method was used to estimate the fractal dimension for image contents. In order to determine a threshold value for the textual objects within image, the fractal dimension was computed for a number of gray scale textual images. The fractal number of each pixel was calculated, then the mean value of all these fractal values were computed. The threshold value was used in recognizing and retrieving the textual objects within image. This algorithm was applied on 75 image samples, 25 image samples were used in training phase, the threshold value was determined throughout this phase; whereas 50 image samples were used in testing the algorithm. The proposed algorithm has performed extremely well with recognition rates 91.5% which is considered good performance. It is a promising technique for optical character recognition system.

Proceedings of the 1st International Conference on Internet of Things and Machine Learning

AL-Rafidain Journal of Computer Sciences and Mathematics, 2013

This paper proposes a new intelligent off-line Arabic handwritten signature identification and ve... more This paper proposes a new intelligent off-line Arabic handwritten signature identification and verification system based on texture analysis. The system uses the texture as feature and back propagation neural network as classifier. The signature image is preprocessed by several operations (Noise removal, Conversion of the signature image to binary image, Finding outer rectangle, Thinning and Size normalization) then the fractal number and co-occurrence matrix are computed to estimate texture features. In this work, two off-line Arabic handwritten signature identification systems are constructed. The first one uses the nearest Euclidean distance, while the other uses back propagation neural network. The paper analyzes and compares the results obtained from the two proposed systems to show the robustness level of the proposed intelligence system. Furthermore, the proposed system was tested by using Genuine signatures and has achieved a CCR (Correct Classification Rate) of 100% in best cases, while it was tested by using Forged signatures it has achieved a CRR approximated to 96.3% in best cases. The experimental results showed that the proposed system is efficient and competent with other state-of-the-art texture-based offline signature identification systems.

JOURNAL OF EDUCATION AND SCIENCE, 1970

Computational and mathematical methods in medicine, 2017

Free-breathing cardiac magnetic resonance (CMR) imaging has short examination time with high repr... more Free-breathing cardiac magnetic resonance (CMR) imaging has short examination time with high reproducibility. Detection of the end-diastole and the end-systole frames of the free-breathing cardiac magnetic resonance, supplemented by visual identification, is time consuming and laborious. We propose a novel method for automatic identification of both the end-diastole and the end-systole frames, in the free-breathing CMR imaging. The proposed technique utilizes the convolutional neural network to locate the left ventricle and to obtain the end-diastole and the end-systole frames from the respiratory motion signal. The proposed procedure works successfully on our free-breathing CMR data, and the results demonstrate a high degree of accuracy and stability. Convolutional neural network improves the postprocessing efficiency greatly and facilitates the clinical application of the free-breathing CMR imaging.

JOURNAL OF EDUCATION AND SCIENCE, 2013

Automatic character recognition has been the subject of intensive research for almost last decade... more Automatic character recognition has been the subject of intensive research for almost last decades. Because of the complexity of printed and handwritten Arabic text a little research has been conducted on the automatic recognition of Arabic characters. This research proposed a new technique for recognizing printed Arabic character. After acquisition Arabic character image a number of preprocessing steps are performed for the digitized image. These steps generally include smoothing by using median filter, the horizontal and vertical histogram profile are used for segmentation and a standard Guo thinning algorithm for thinning, Etc.. Variation Method and Discrete Cosine Transform Method are used for feature extraction. For classification radial basis function (RBF) network are used. This method performs extremely well. This new technique is able to handle printed Arabic character task efficiently.