Stroke Width Research Articles

Deep Learning and Particle Swarm Optimisation-Based Techniques for Visually Impaired Humans' Text Recognition and Identification

Blind people can benefit greatly from a system capable of localising and reading comprehension text embedded in natural scenes and providing useful information that boosts their self-esteem and autonomy in everyday situations. Regardless of the fact that existing optical character recognition programmes seem to be quick and effective, the majority of them are not able to correctly recognise text embedded in usual panorama images. The methodology described in this paper is to localise textual image regions and pre-process them using the naïve Bayesian algorithm. A weighted reading technique is used to generate the correct text data from the complicated image regions. Usually, images hold some disturbance as a result of the fact that filtration is proposed during the early pre-processing step. To restore the image's quality, the input image is processed employing gradient and contrast image methods. Following that, the contrast of the source images would be enhanced using an adaptive image map. The stroke width transform, Gabor’s transform, and weighted naïve Bayesian classifier methodologies have been used in complicated degraded images to segment, feature extraction, and detect textual and non-textual elements. Finally, to identify categorised textual data, the confluence of deep neural networks and particle swarm optimisation is being used. The text in the image is transformed into an acoustic output after identification. The dataset IIIT5K is used for the development portion, and the performance of the suggested come up is evaluated using parameters such as accuracy, recall, precision, and F1-score.

Analysis of Text Identification Techniques Using Scene Text and Optical Character Recognition

In today's era, data in digitalized form is needed for faster processing and performing of all tasks. The best way to digitalize the documents is by extracting the text from them. This work of text extraction can be performed by various text identification tasks such as scene text recognition, optical character recognition, handwriting recognition, and much more. This paper presents, reviews, and analyses recent research expansion in the area of optical character recognition and scene text recognition based on various existing models such as convolutional neural network, long short-term memory, cognitive reading for image processing, maximally stable extreme regions, stroke width transformation, and achieved remarkable results up to 90.34% of F-score with benchmark datasets such as ICDAR 2013, ICDAR 2019, IIIT5k. The researchers have done outstanding work in the text recognition field. Yet, improvement in text detection in low-quality image performance is required, as text identification should not be limited to the input quality of the image.

Text Region Identification in Indian Street Scene Images Using Stroke Width Transform and Support Vector Machine

Text Region Identification in Indian Street Scene Images Using Stroke Width Transform and Support Vector Machine

An enhanced binarization framework for degraded historical document images

Binarization plays an important role in document analysis and recognition (DAR) systems. In this paper, we present our winning algorithm in ICFHR 2018 competition on handwritten document image binarization (H-DIBCO 2018), which is based on background estimation and energy minimization. First, we adopt mathematical morphological operations to estimate and compensate the document background. It uses a disk-shaped structuring element, whose radius is computed by the minimum entropy-based stroke width transform (SWT). Second, we perform Laplacian energy-based segmentation on the compensated document images. Finally, we implement post-processing to preserve text stroke connectivity and eliminate isolated noise. Experimental results indicate that the proposed method outperforms other state-of-the-art techniques on several public available benchmark datasets.

The effect of character stroke width on legibility: The relationship between duty ratio and contrast threshold

Stroke width is an important factor affecting letter legibility. Although there seems to be a critical or optimal value for stroke width, it has been difficult to estimate quantitatively because shape deterioration and the increase of stroke width are covariates. This study attempted to predict the optimal value of stroke width considering it as a duty ratio. The duty ratio is related to the amplitude of the fundamental frequency, with the maximum amplitude of the fundamental frequency occurring at a duty ratio of 0.5. The fundamental frequency decreases when the duty ratio is set either lower or higher than 0.5. Based on previous research indicating that letters also have a critical band (CB), or the frequency band which contributes most to letter recognition, which corresponds to their fundamental frequency, we hypothesized that the amplitude of CB is related to the legibility of letters. We measured the contrast thresholds using square-wave gratings and letters of differing duty ratios. In both the grating and letter stimuli, the amplitude of CB reached its maximum when the duty ratio was around 0.5. Contrast thresholds for stimuli with various duty ratios were accurately predicted by the amplitudes of CB and the contrast thresholds in the individual’s sine-wave grating. Our results suggest that the amplitude of a single critical component contributes significantly to the legibility of letters, even though they contain diverse frequency components.

An Improved Deep Learning Network Structure for Multitask Text Implication Translation Character Recognition

With the rapid development of artificial intelligence technology, multitasking textual translation has attracted more and more attention. Especially after the application of deep learning technology, the performance of multitask translation text detection and recognition has been greatly improved. However, because multitasking contains the interference problem faced by the translated text, there is a big gap between recognition performance and actual application requirements. Aiming at multitasking and translation text detection, this paper proposes a text localization method based on multichannel multiscale detection of the largest stable extreme value region and cascade filtering. This paper selects the appropriate color channel and scale to extract the maximum stable extreme value area as the character candidate area and designs a cascaded filter from coarse to fine to remove false detections. The coarse filter is based on some simple morphological features and stroke width features, and the fine filter is trained by a two-recognition convolutional neural network. The remaining character candidate regions are merged into horizontal or multidirectional character strings through the graph model. The experimental results on the text data set prove the effectiveness of the improved deep learning network character model and the feasibility of the textual implication translation analysis method based on this model. Among them, the text contains translation character recognition results prove that the model has good description ability. The characteristics of the model determine that this method is not sensitive to the scale of the sliding window, so it performs better than the existing typical methods in retrieval tasks.

The effect of induced blur on the Beery-Buktenica developmental test of visual-motor integration and its supplemental tests.

BackgroundThe Beery-Buktenica Test of Visual-Motor Integration (Beery VMI) is a commonly used standardized test of visual-motor integration. Performance on the test is related to academic achievement, but the effect of reduced visual acuity on test results is unknown. This study determined the visual acuity demand and the spacing of the test forms for the Beery VMI and its supplemental tests and investigated the effect of induced optical blur on test results in both adults and children.MethodsThe overall size and critical detail size of each form and the distance between the center of each form and its adjacent crowding source were measured. The visual acuity demand and angular separation of forms were calculated. Two groups of participants (adults aged ≥18 years [n = 19] and children aged 7–12 years [n = 20]) completed four sessions in which they performed the Beery VMI and the supplemental tests under different blur conditions (habitual near correction, monocular spherical blur, binocular spherical blur and binocular astigmatic blur) in a randomized order.ResultsStroke width, overall form size and box size remained constant throughout the Beery VMI, whereas these reduced with increasing difficulty for the supplemental tests. Reduced near visual acuity from simulated blur resulted in reduced mean scores for the Beery VMI and its supplemental tests, compared with habitual near vision in both adults and children. Binocular spherical blur had the most detrimental effect (p<0.001), followed by binocular astigmatic blur (p<0.001) then monocular spherical blur (p = 0.022).ConclusionsIn individuals with uncorrected spherical or astigmatic ametropia, reduced scores on the Beery VMI and its supplemental tests may be due to impaired near visual acuity and not reflect reduced visual-motor abilities. This highlights the importance of excluding reduced near visual acuity as a cause of reduced performance before diagnosing impairment and initiating treatment strategies for visual-motor integration.

Hybrid deep neural network with adaptive galactic swarm optimization for text extraction from scene images

Text obtained in natural scenes contains various information; therefore, it is extensively used in various applications to understand the image scenarios and also to retrieve the visual information. The semantic information provided by this scene image is very much valuable for human beings to realize the whole environment. But the text in such natural images depicts a flexible appearance in an unconstrained environment which makes the text identification and character recognition process a more challenging one. Therefore, a weighted naive Bayes classifier (WNBC)-based deep learning process is used in this framework to effectively detect the text and to recognize the character from the scene images. Normally, the natural scene images may carry some kind of noise in it, and to remove that, the guided image filter is introduced at the pre-processing stage. The features that are useful for the classification process are extracted using the Gabor transform and stroke width transform techniques. Finally, with these extracted features, the text detection and character recognition is successfully achieved by WNBC and deep neural network-based adaptive galactic swarm optimization. Then, the performance metrics such as accuracy, F1-score, precision, mean absolute error, mean square error and recall metrics are evaluated to estimate the adeptness of the proposed method.

Multi-lingual scene text detection and language identification

Scene text analysis is a field of research that poses challenges to researchers owing to the background complexities, image quality, text orientation, text size, etc. The problem gets more complex when the image contains multi-lingual texts. Most scene text detection techniques approach the problem as either a feature-based or deep learning-based problem. In this work, an end-to-end system is proposed for scene text detection, localization and language identification to combine feature-based and deep learning-based approaches. The model uses Maximally Stable Extremal Regions and Stroke Width Transform for generating text proposals, followed by proposal refinement using Generative Adversarial Network. Finally, a Convolution Neural Network based model is used for language identification of the detected scene texts. Experiments have been conducted on standard datasets like KAIST, COCO, CTW1500, CVSI and ICDAR along with an in-house multi-lingual Indic scene text dataset for which the proposed model achieves satisfactory results.

River extraction from GF-1 satellite images combining stroke width transform and a geometric feature set

River extraction from GF-1 satellite images combining stroke width transform and a geometric feature set

Grouping Effects on Foveal Spatial Interactions in Children.

PurposeGrouping of flankers from the target can modulate crowding in adults. Visual acuity in children is measured clinically using charts with targets and different flankers to enhance spatial interactions. We investigated grouping effects on interactions using visual acuity letters, flanked by contours and letters, in children.MethodsVisual acuity for isolated and flanked letters was measured in 155 three- to 11-year old children and 32 adults. Flankers were one stroke width from the target and were a box or four bars and black or red letters. Magnitudes of interaction were flanked minus isolated logMAR acuities. Psychometric function slopes were also examined.ResultsMagnitudes of interaction by contours did not change significantly with age. They were 0.047 ± 0.014 logMAR more with bars than a box. Interaction from flanking letters reduced with age, adults being not different from 9- to 11-year-olds for black and red letter surrounds. It was weaker by 0.033 ± 0.013 logMAR when a black letter was surrounded by red rather than black letters. Psychometric function slopes for visual acuity were steepest for the youngest children (3–5 years).ConclusionsFor contour and letter flankers, grouping effects on interaction magnitude are age independent. Grouping bars into a box forming a single object reduces magnitude of effect. Grouping letter flankers by color and ungrouping them from the target reduce interaction magnitude by ∼8%, suggesting that luminance-defined form dominates. Differently colored letter flankers of high-luminance contrast on acuity charts could draw attention to the target but retain significant interaction strength.

Modified Sauvola binarization for degraded document images

The binarization of historical documents is a difficult job due to the presence of many degradations. Many existing local binarization techniques use certain manually adjusted parameters. The output of these techniques is much dependent on the value of these parameters. One of such parameters is window size which is kept fixed for the whole text image. The fixed window size will not be able to perform well for images having variable stroke widths and text sizes. The proposed binarization technique (Modified Sauvola) is the modification of state of art Sauvola’s binarization technique. It automatically computes window size dynamically across the image pixel to pixel using the stroke width transform (SWT). This led to reduction in number of manually adjusted parameters. The results are compared with the nine existing techniques using the quantitative measures: FM, PSNR, NRM, MPM, and DRD. The results show that the proposed method outperforms existing methods for images having variable stroke widths and text sizes.

Legibility of prints on paper made from Japanese knotweed

The spread of invasive alien plant species (IAPS) is a leading reason for worldwide environmental change due to their effects on biodiversity and humans. Some valued goods from IAPS have been produced, e.g. paper that consists of cellulose fibres from Japanese knotweed. Therefore, the aim of this study was to establish the usability of this paper grade as a printing substrate, since it does not have ideal optical properties as it is expected from commercial office paper. Because it is widely used, inkjet printing technology was employed. Print permanence is essential, especially when printing documents. However, typographic characteristics must be considered to make a text more legible. Two widely used typefaces (Arial and Times) were tested in three commonly used type sizes (8 pt, 10 pt, and 12 pt). The results showed that the paper made from Japanese knotweed could have valuable properties and suitable legibility, especially when using typefaces with a moderate counter size, high x-height, and minimal differences in the letter stroke width to obtain an appropriate typographic tonal density with an adequate type size. Even after exposure to light, the texts printed in a proper type size and stroke width remained visible.

Association between red cell distribution width level and risk of stroke: A systematic review and meta-analysis of prospective studies.

Background:Red cell distribution width level may have relations with the incidence and prognosis of cerebrovascular diseases. Recent researches have reported that red cell distribution width level was linked to the occurrence of stroke. However, the predicted effect of red cell distribution width in stroke is still disputed. We sought to assess the relationship between red cell distribution width and risk of stroke in this meta-analysis.Methods:Relevant studies were picked out from the databases of Embase, PubMed, and Cochrane Library. Hazard ratio with 95% confidence interval was chosen to analyze each trial, which was extracted from results of the highest and lowest red cell distribution width group. Funnel plots, Begg and Egger test were used to assess publication bias in the meta-analysis. Stata(12.0) was utilized to perform statistic analysis in the process.Results:A total of 6 studies with 5783 patients were included in this meta-analysis. The results showed that red cell distribution width level in patients with stroke was significantly higher than it in those without stroke (HR = 1.34, 95%CI:1.23–1.47, P < .001), in particular ischemic stroke(HR = 1.34,95% confidence interval:1.1–1.54, P < .001). There was no evidence of heterogeneity across the studies (P = .355, I2 = 5.53%).Conclusions:The higher red cell distribution width level was associated with an increased risk of stroke, especially ischemic infarction.

Text Region Extraction From Scene Images Using AGF and MSER

The natural scene images contain text as an integral part of that image that supplies paramount knowledge about it. This information and knowledge can be used in the variety of purposes like image-based searching, automatic number plate recognition, robot navigation, etc. but text region extraction and detection in scenery images could be quite a challenging job due to image blur, distortion, noise, etc. In this paper, we discuss a method for extraction of text regions by generating prospective components by applying maximally stable extremal regions (MSER) and boundary smoothing by Alternating guided image filter, which is one of the newest filters to deal with noise and halo effect elimination. The separation of non-text &amp; text components is achieved by AdaBoost classifier that separates text and non-text on the basis of the three text specific features namely maximum stroke width ratio, compactness, color divergence. The proposed method assist in extracting text regions from the blurred and low contrast natural scene images effectively. The ICDAR 2013 training and testing dataset is applied for the experiments and evaluation of the method. The evaluation is carried out using deteval software for calculating precision, f-measure, recall for the detected, and extracted text regions.

SOLVING THE CLASSIFICATION PROBLEM OF CIRCULAR METAL OBJECTS WITH ENGRAVED CHARACTERS BY IMAGE PROCESSING METHODS

In this study, two different solution ways have been developed for the problem of classification of industrial small circular metal objects on the surfaces of engraved metal. It is the first proposed solution to perform the pattern matching with XOR operator by extract the character region of the circular metal objects as a pre-process, making the model of the Daugman’s Rubber Sheet Model (DRSM) and performing feature extraction. As a result, obtained that average processing time is 69,72 milliseconds and 0,9398 accuracy rate in the first proposed solution. The second solution is the optical character recognition (OCR) on the circular metal objects that to be realized character region detection and character segmentation as a result of the Maximal Stabil Extremal Region (MSER) and Stroke Width Transform (SWT) algorithms. Character recognition realized by using the model of Convolutional Neural Network (CNN) class which is a deep machine learning approach of artificial intelligence. The character recognition problem of the circular metal objects provided at the same time solved the problem of object classification. As a result, obtained that average processing time is 1,596 second and 0,9719 accuracy rate in the second proposed solution.

A ROAD EXTRACTION METHOD BASED ON HIGH RESOLUTION REMOTE SENSING IMAGE

Abstract. Aiming at the road extraction in high-resolution remote sensing images, the stroke width transformation algorithm is greatly affected by surrounding objects, and it is impossible to directly obtain high-precision road information. A new road extraction method combining stroke width transformation and mean drift is proposed. In order to reduce road holes and discontinuities, and preserve better edge information, the algorithm first performs denoising preprocessing by means of median filtering to the pre-processed image. Then, the mean shift algorithm is used for image segmentation. The adjacent parts of the image with similar texture and spectrum are treated as the same class, and then the fine areas less than the maximum stroke width are reduced. On the basis , the road information is extracted by the stroke width transformation algorithm, and the information also contains a small amount of interference information such as spots (non-road). In order to further improve road extraction accuracy and reduce speckle and non-road area interference, the basic operations and combinations in mathematical morphology are used to optimize it. The experimental results show that the proposed algorithm can accurately extract the roads on high-resolution remote sensing images, and the better the road features, the better the extraction effect. However, the applicability of the algorithm is greatly affected by the surrounding objects.

Application on character recognition system on road sign for visually impaired: case study approach and future

Many visually impaired people worldwide are unable to travel safely and autonomously because they are physically unable to perceive effective visual information during their daily lives. In this research, we study how to extract the character information of the road sign and transmit it to the visually impaired effectively, so they can understand easier. Experimental method is to apply the Maximally Stable External Region and Stroke Width Transform method in Phase I so that the visually impaired person can recognize the letters on the road signs. It is to convey text information to the disabled. The result of Phase I using samples of simple road signs was to extract the sign information after dividing the exact character area, but the accuracy was not good for the Hangul (Korean characters) information. The initial experimental results in the Phase II succeeded in transmitting the text information on Phase I to the visually impaired. In the future, it will be required to develop a wearable character recognition system that can be attached to the visually impaired. In order to perform this task, we need to develop and verify a miniaturized and wearable character recognition system. In this paper, we examined the method of recognizing road sign characters on the road and presented a possibility that may be applicable to our final development.

Detecting Gurmukhi and Hindi text object present in images

This paper resolves the trouble detecting Gurmukhi and Hindi text objects present in images. Therefore, we propose a new region-based bottom-up method to extract text from images, where we first identify elementary substructures using connected components and edges and then merge them successively. As each character counter has high contrast as compared to its neighbours, all character pixels and several non-character pixels which exhibit excessive neighbourhood. Although a lot of noise is removed by this step a second pass filter is applied based on stroke width. The stroke width of Hindi and Gurmukhi characters are approximately uniform. The candidates with similar properties are aggregated into chains, according to the observation that the figures of a book. For performance evaluation, we created our own dataset containing 60 images of varied size, resolution and type. The performance is evaluated on this dataset containing caption, document, and scene text images using different performance metrics.

Detecting Gurmukhi and Hindi text object present in images

This paper resolves the trouble detecting Gurmukhi and Hindi text objects present in images. Therefore, we propose a new region-based bottom-up method to extract text from images, where we first identify elementary substructures using connected components and edges and then merge them successively. As each character counter has high contrast as compared to its neighbours, all character pixels and several non-character pixels which exhibit excessive neighbourhood. Although a lot of noise is removed by this step a second pass filter is applied based on stroke width. The stroke width of Hindi and Gurmukhi characters are approximately uniform. The candidates with similar properties are aggregated into chains, according to the observation that the figures of a book. For performance evaluation, we created our own dataset containing 60 images of varied size, resolution and type. The performance is evaluated on this dataset containing caption, document, and scene text images using different performance metrics.