Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702584
Navdeep Bhatnagar, Suchi Johari
With internet availability in almost all types of devices, downloading has become a widespread activity. Data is consumed when any content is downloaded from the web servers. The effectiveness of any internet service provider depends upon the internet connection speed and the time consumed to download a file. The internet connection speed is not always the same. The speed can sometimes be so slow that downloads are not possible. In a typical organizational setup, different types of files and software applications are downloaded regularly. Downloading the same file multiple times by different people is a wastage of resources. The proposed approach aims to decrease the search and download time. The primary purpose of this study is to ensure the availability of the file for download even if there is poor or no internet connectivity. The paper proposes a Web Crawler-based caching technique that helps resolve the issues mentioned above and makes downloading more efficient.
{"title":"Web Crawler based Caching Technique for efficient Downloading","authors":"Navdeep Bhatnagar, Suchi Johari","doi":"10.1109/ICIIP53038.2021.9702584","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702584","url":null,"abstract":"With internet availability in almost all types of devices, downloading has become a widespread activity. Data is consumed when any content is downloaded from the web servers. The effectiveness of any internet service provider depends upon the internet connection speed and the time consumed to download a file. The internet connection speed is not always the same. The speed can sometimes be so slow that downloads are not possible. In a typical organizational setup, different types of files and software applications are downloaded regularly. Downloading the same file multiple times by different people is a wastage of resources. The proposed approach aims to decrease the search and download time. The primary purpose of this study is to ensure the availability of the file for download even if there is poor or no internet connectivity. The paper proposes a Web Crawler-based caching technique that helps resolve the issues mentioned above and makes downloading more efficient.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129886210","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702624
R. Muzammil, M. Wajid, Mohammed Usman
The equalizer essentially acts as the inverse of the channel so that the channel and equalizer together introduce no distortion to the signal at the input to the detector. The channel here is modeled as consisting of two paths with 60% and 40% power distribution among them. This introduces inter-symbol interference (ISI), and that is removed using the equalizer. The modulation scheme used in this work is 4-QAM. The equalizer is modeled using a linear regression model. This employs a Maximum Aposteriori (MAP) estimation of the parameter vector. Simulation results are subsequently shown in the end which demonstrate complete removal of ISI.
{"title":"Design of a Linear Equalizer using Regression Model and Maximum Aposteriori Estimation","authors":"R. Muzammil, M. Wajid, Mohammed Usman","doi":"10.1109/ICIIP53038.2021.9702624","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702624","url":null,"abstract":"The equalizer essentially acts as the inverse of the channel so that the channel and equalizer together introduce no distortion to the signal at the input to the detector. The channel here is modeled as consisting of two paths with 60% and 40% power distribution among them. This introduces inter-symbol interference (ISI), and that is removed using the equalizer. The modulation scheme used in this work is 4-QAM. The equalizer is modeled using a linear regression model. This employs a Maximum Aposteriori (MAP) estimation of the parameter vector. Simulation results are subsequently shown in the end which demonstrate complete removal of ISI.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121760468","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702671
Digvijay Puri, H. Saini, Suman Saha
Linear Time clustering is a critical topic in data analysis, and it has received a lot of attention in the earlier days. There are a collection of data points in this problem, as well as a "similarity (or dissimilarity)" metric that determines how identical two data points are. A linear-time clustering algorithm's goal is to group data points together so that data points in the same cluster are similar and data points in separate clusters are distinct. Linear Time Clustering is employed in data mining to uncover groupings and find interesting distributions in the underlying data. This paper presents a review on linear time clustering models. The simulation platforms, analysis on algorithms, utilized datasets, and the evaluated performance metrics are performed in this review. The planned comparison research will use publicly available large data to show how each approach affects linear-time clustering along with the future research direction.
{"title":"Linear Time Clustering Models: A Review","authors":"Digvijay Puri, H. Saini, Suman Saha","doi":"10.1109/ICIIP53038.2021.9702671","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702671","url":null,"abstract":"Linear Time clustering is a critical topic in data analysis, and it has received a lot of attention in the earlier days. There are a collection of data points in this problem, as well as a \"similarity (or dissimilarity)\" metric that determines how identical two data points are. A linear-time clustering algorithm's goal is to group data points together so that data points in the same cluster are similar and data points in separate clusters are distinct. Linear Time Clustering is employed in data mining to uncover groupings and find interesting distributions in the underlying data. This paper presents a review on linear time clustering models. The simulation platforms, analysis on algorithms, utilized datasets, and the evaluated performance metrics are performed in this review. The planned comparison research will use publicly available large data to show how each approach affects linear-time clustering along with the future research direction.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"50 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127060980","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702641
Gagandeep Kaur, Shubhi Kansal
A new tri-histogram equalisation algorithm based on weighted clipping is suggested for digital image enhancement using histogram equalisation. The power law transformation function is used to perform the clipping operation which avoids the over-enhancement and under-enhancement of the image. The clipped histogram is further divided into three sub-histograms on the basis of the mean and standard deviation of the input image. Sub-histograms are independently equalised and the combined to produce the final enhanced image. Experimental results clearly shows the weighted clipping tri-histogram equalisation method enhances the image quality. Comparisons made with other methods also demonstrate the superiority of the proposed algorithm. Subjective and objective quality assessment of proposed method outperformed the other state-of-the-art methods. Objective metrics like Entropy, Feature Similarity Index Matrix (FSIM), Visual Saliency Induced Matrix (VSI) and Gradient Magnitude Similarity Matrix(GMSD) have been evaluated for the comparison purpose. Proposed method work very well both for gray and color images.
{"title":"Power Law Transformation based Weighted Clipping Tri-Histogram Equalisation","authors":"Gagandeep Kaur, Shubhi Kansal","doi":"10.1109/ICIIP53038.2021.9702641","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702641","url":null,"abstract":"A new tri-histogram equalisation algorithm based on weighted clipping is suggested for digital image enhancement using histogram equalisation. The power law transformation function is used to perform the clipping operation which avoids the over-enhancement and under-enhancement of the image. The clipped histogram is further divided into three sub-histograms on the basis of the mean and standard deviation of the input image. Sub-histograms are independently equalised and the combined to produce the final enhanced image. Experimental results clearly shows the weighted clipping tri-histogram equalisation method enhances the image quality. Comparisons made with other methods also demonstrate the superiority of the proposed algorithm. Subjective and objective quality assessment of proposed method outperformed the other state-of-the-art methods. Objective metrics like Entropy, Feature Similarity Index Matrix (FSIM), Visual Saliency Induced Matrix (VSI) and Gradient Magnitude Similarity Matrix(GMSD) have been evaluated for the comparison purpose. Proposed method work very well both for gray and color images.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128246544","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702630
Sharad Salunke, M. Venkatadri, Md. Farukh Hashmi
The cloud-based data storage scenario is continually demanding more storage space with the increasing use of images from various sources, in this way the information should be compacted likewise with the security. Representing an image by using a lesser number of bits and too without much compromise of the image quality is known as faithful image compression. Compressing files will conserve disk space, speed up file transmission, and reduce expenses for computer equipment and communication bandwidth. The current manuscript presents the idea of combining the phenomenon of Beta and Logistic based chaos for encryption with a powerful compression scheme DWT-SPIHT. DWT is being used to reduce the image file size without reducing the quality, resulting in higher resolution. SPIHT is a technique that incorporates the implicit similarities in a wavelet decomposition of an image across the sub-bands for compression. It is among the most effective technique for the compression of images dependent on wavelets. The fundamental advantage of the DWT-SPIHT with Beta and chaos strategy is decent picture quality with less mean square error. The validity of the method is demonstrated and analyzed with the help of some common metrics such as compression rate, PSNR, MSE, and correlation coefficient.
{"title":"Beta and Logistic Function Driven Encryption for Compressed Images","authors":"Sharad Salunke, M. Venkatadri, Md. Farukh Hashmi","doi":"10.1109/ICIIP53038.2021.9702630","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702630","url":null,"abstract":"The cloud-based data storage scenario is continually demanding more storage space with the increasing use of images from various sources, in this way the information should be compacted likewise with the security. Representing an image by using a lesser number of bits and too without much compromise of the image quality is known as faithful image compression. Compressing files will conserve disk space, speed up file transmission, and reduce expenses for computer equipment and communication bandwidth. The current manuscript presents the idea of combining the phenomenon of Beta and Logistic based chaos for encryption with a powerful compression scheme DWT-SPIHT. DWT is being used to reduce the image file size without reducing the quality, resulting in higher resolution. SPIHT is a technique that incorporates the implicit similarities in a wavelet decomposition of an image across the sub-bands for compression. It is among the most effective technique for the compression of images dependent on wavelets. The fundamental advantage of the DWT-SPIHT with Beta and chaos strategy is decent picture quality with less mean square error. The validity of the method is demonstrated and analyzed with the help of some common metrics such as compression rate, PSNR, MSE, and correlation coefficient.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"49 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127967967","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Efficient locating of a vegetable/fruit on a tree is one of the major requirements for automated agricultural harvesting process. In a tree with green fruits, the vegetable and the leaves look quite similar and are sometimes not detected. This work presents a way to detect and differentiate between fruits and leaves using a novel algorithm. The process uses methodology of 2-D and 3-D image differentiation and image processing. The algorithm is designed with the aim of analysing and studying the input image on features like colour, intensity, and shape and detect the location of fruit on tree. The proposed approach can be applied for targeting vegetables and fruits for robotic harvesting. The proposed algorithm is useful for farming and home purposes where with the involvement of this fruits can be detected before they are spoiled. Thus, helping in proper and timely utilisation of fruits for mankind saving time and money.
{"title":"2D and 3D Object Differentiation in Trees with Green Fruits","authors":"Sumit Bhardwaj, Sukeerat Singh Goindi, Radhika Raghav, Punit Gupta, D. Saini","doi":"10.1109/ICIIP53038.2021.9702693","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702693","url":null,"abstract":"Efficient locating of a vegetable/fruit on a tree is one of the major requirements for automated agricultural harvesting process. In a tree with green fruits, the vegetable and the leaves look quite similar and are sometimes not detected. This work presents a way to detect and differentiate between fruits and leaves using a novel algorithm. The process uses methodology of 2-D and 3-D image differentiation and image processing. The algorithm is designed with the aim of analysing and studying the input image on features like colour, intensity, and shape and detect the location of fruit on tree. The proposed approach can be applied for targeting vegetables and fruits for robotic harvesting. The proposed algorithm is useful for farming and home purposes where with the involvement of this fruits can be detected before they are spoiled. Thus, helping in proper and timely utilisation of fruits for mankind saving time and money.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127986508","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702629
Preet Navdeep, Neeraj Sharma, Manish Arora
Facial Emotion Recognition (FER) is an essential research topic in the science of human-computer interaction. When it comes to successful communication, facial expressions are a powerful way of revealing one's intentions and feelings to others. In the perception of human articulations, facial emotions such as sadness, happiness, anger, disgust, and fear play an important role. Hence, the study of such facial expressions is a fascinating and intriguing topic to investigate. In this review article, we will describe the methods proposed by many researchers in terms of facial expression recognition utilising machine and deep learning models, as well as their contributions and the problems that still exist in this field. In order to draw some conclusions, we have provided our assessment of the various techniques that were given.
{"title":"Computational Learning Based Facial Emotions Recognition: A Review","authors":"Preet Navdeep, Neeraj Sharma, Manish Arora","doi":"10.1109/ICIIP53038.2021.9702629","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702629","url":null,"abstract":"Facial Emotion Recognition (FER) is an essential research topic in the science of human-computer interaction. When it comes to successful communication, facial expressions are a powerful way of revealing one's intentions and feelings to others. In the perception of human articulations, facial emotions such as sadness, happiness, anger, disgust, and fear play an important role. Hence, the study of such facial expressions is a fascinating and intriguing topic to investigate. In this review article, we will describe the methods proposed by many researchers in terms of facial expression recognition utilising machine and deep learning models, as well as their contributions and the problems that still exist in this field. In order to draw some conclusions, we have provided our assessment of the various techniques that were given.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"99 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134029703","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702619
Manu Gangyan, Pravir Kumar
Alzheimer’sdisease (AD) and Diabetes Mellitus are two of most common diseases worldwide, increasing at an alarming rate. AD patients are diagnosed with gradual decline in cognitive abilities and memory function, and characterized by formation of extracellular amyloid-beta plaques andhyperphosphorylated tau proteinintracellular neurofibrillary tangles and loss of neurons. T2DM is characterized with hyperglycemia, hyperinsulinemia, insulin-resistance, and islet amyloid-polypeptide formation. T2DM patients has reduced graymatter in AD vulnerable regions, reduced hippocampus andwhole brain volume, indicating declined cognitive functions later on, and further risk for developing AD. The alliance between AD and T2DM is more complex and stronger as both interlinked with several pathological features(insulin resistance, hyperglycemia, hyperinsulinemia and Aβ-plaque formation)and hence referred as Type3 diabetes mellitus or diabetes of Brain. Vascular endothelial growth factor is a potent cytokine and its alteredsignalling leads to Diabetes and AD pathology as reduced pancreatic and islet vascularisation results in declined β-cell proliferation which ultimately impairs glucose clearance and might results in development of insulin resistance which leads to chronic diabetes while in AD serum, its correlation with Amyloid-beta pathology suggests further AD progression through neurodegeneration and vascular dysfunction. In this article, we have introduceddifferentdisease biomarkers and their interactions with the VEGFfortargeting glucose clearance, insulin resistance and Aβ clearance. The objective of this report is to highlight pathological intimation of VEGF to establish a possible therapeutic window for treatment of various abnormalities related to impaired VEGF to slow down the progression of AD and diabetes.
{"title":"VEGF and its role in the treatment of Diabetes and Alzhiemer’s Disease","authors":"Manu Gangyan, Pravir Kumar","doi":"10.1109/ICIIP53038.2021.9702619","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702619","url":null,"abstract":"Alzheimer’sdisease (AD) and Diabetes Mellitus are two of most common diseases worldwide, increasing at an alarming rate. AD patients are diagnosed with gradual decline in cognitive abilities and memory function, and characterized by formation of extracellular amyloid-beta plaques andhyperphosphorylated tau proteinintracellular neurofibrillary tangles and loss of neurons. T2DM is characterized with hyperglycemia, hyperinsulinemia, insulin-resistance, and islet amyloid-polypeptide formation. T2DM patients has reduced graymatter in AD vulnerable regions, reduced hippocampus andwhole brain volume, indicating declined cognitive functions later on, and further risk for developing AD. The alliance between AD and T2DM is more complex and stronger as both interlinked with several pathological features(insulin resistance, hyperglycemia, hyperinsulinemia and Aβ-plaque formation)and hence referred as Type3 diabetes mellitus or diabetes of Brain. Vascular endothelial growth factor is a potent cytokine and its alteredsignalling leads to Diabetes and AD pathology as reduced pancreatic and islet vascularisation results in declined β-cell proliferation which ultimately impairs glucose clearance and might results in development of insulin resistance which leads to chronic diabetes while in AD serum, its correlation with Amyloid-beta pathology suggests further AD progression through neurodegeneration and vascular dysfunction. In this article, we have introduceddifferentdisease biomarkers and their interactions with the VEGFfortargeting glucose clearance, insulin resistance and Aβ clearance. The objective of this report is to highlight pathological intimation of VEGF to establish a possible therapeutic window for treatment of various abnormalities related to impaired VEGF to slow down the progression of AD and diabetes.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"402 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131754190","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702561
Rahul Pradhan
YouTube is the most used social media platform, and it has been the most popular website where users can post the video. The public generally does comment, like or dislike, video-sharing on a YouTube video. Comment plays a vital role in expressing opinions and mindset, and it is used as an expression of public opinion. The massive amount of comment comes mainly on famous channels where challenges arise to analyze public opinion or behavior regarding that particular video. This article proposes sentiment analysis on YouTube video by Natural Language Processing (NLP) technique. Sentiment Analysis is when comprehension, citation, and processing of text-based data is done, and it directly converts it into sentiment information. This analysis help users to get the report of their YouTube Video. The output of this analysis gives the classification of sentiment analysis, i.e., positive, negative, or neutral.
{"title":"Extracting Sentiments from YouTube Comments","authors":"Rahul Pradhan","doi":"10.1109/ICIIP53038.2021.9702561","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702561","url":null,"abstract":"YouTube is the most used social media platform, and it has been the most popular website where users can post the video. The public generally does comment, like or dislike, video-sharing on a YouTube video. Comment plays a vital role in expressing opinions and mindset, and it is used as an expression of public opinion. The massive amount of comment comes mainly on famous channels where challenges arise to analyze public opinion or behavior regarding that particular video. This article proposes sentiment analysis on YouTube video by Natural Language Processing (NLP) technique. Sentiment Analysis is when comprehension, citation, and processing of text-based data is done, and it directly converts it into sentiment information. This analysis help users to get the report of their YouTube Video. The output of this analysis gives the classification of sentiment analysis, i.e., positive, negative, or neutral.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132673362","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-11-26DOI: 10.1109/ICIIP53038.2021.9702659
Jaspreet Kaur, Vishal Goyal, Manish Kumar
The development of Optical Character Recognition (OCR) in Indian text is an active area of research today. The presence of a large number of characters in a set of alphabets, their complex combinations is a major challenge for the OCR designer. This OCR is giving very good results even for Hindi documents that are machine printed. But, it is not giving good results for typewriter typed Hindi documents. Even there is no OCR that is available to date which is capable of recognizing text from typewriter typed Hindi documents. This paper introduces an automated training data framework, provided only with labelled text images, thus removing the need for manually generated text. In contrast to the training images and images text as ground-truth, this approach is based on the random, rule-based generation of meaningless text in an image file and their ground-truth text file. In this paper we describe a dataset of typewriter type Hindi documents and Ground truth (GT) typewritten Hindi text images paired with their transcription. Typewriter typed documents can be incorporated into the repository of Plagiarism detection tools because text cannot be recognized by any OCR. Thus, the functionality of the existing OCR needs to be extended for recognizing typewriter typed documents.
印度文本光学字符识别(OCR)的发展是当今研究的一个活跃领域。在一组字母中存在大量字符，它们的复杂组合是OCR设计者面临的主要挑战。这个OCR甚至对机器打印的印地语文档也给出了非常好的结果。但是，对于用打字机打出的印地语文档，它并没有给出好的结果。甚至到目前为止还没有可用的OCR能够识别来自打字机键入的印地语文档的文本。本文介绍了一个自动化的训练数据框架，只提供标记的文本图像，从而消除了手动生成文本的需要。与训练图像和图像文本作为基础真值相比，这种方法是基于图像文件及其基础真值文本文件中无意义文本的随机、基于规则的生成。在本文中，我们描述了一个打字机式印地语文档和Ground truth (GT)打字的印地语文本图像与其转录配对的数据集。打字机打字的文档可以被合并到抄袭检测工具库中，因为文本不能被任何OCR识别。因此，需要扩展现有OCR的功能，以识别打字机键入的文档。
{"title":"Tesseract OCR for Hindi Typewritten Documents","authors":"Jaspreet Kaur, Vishal Goyal, Manish Kumar","doi":"10.1109/ICIIP53038.2021.9702659","DOIUrl":"https://doi.org/10.1109/ICIIP53038.2021.9702659","url":null,"abstract":"The development of Optical Character Recognition (OCR) in Indian text is an active area of research today. The presence of a large number of characters in a set of alphabets, their complex combinations is a major challenge for the OCR designer. This OCR is giving very good results even for Hindi documents that are machine printed. But, it is not giving good results for typewriter typed Hindi documents. Even there is no OCR that is available to date which is capable of recognizing text from typewriter typed Hindi documents. This paper introduces an automated training data framework, provided only with labelled text images, thus removing the need for manually generated text. In contrast to the training images and images text as ground-truth, this approach is based on the random, rule-based generation of meaningless text in an image file and their ground-truth text file. In this paper we describe a dataset of typewriter type Hindi documents and Ground truth (GT) typewritten Hindi text images paired with their transcription. Typewriter typed documents can be incorporated into the repository of Plagiarism detection tools because text cannot be recognized by any OCR. Thus, the functionality of the existing OCR needs to be extended for recognizing typewriter typed documents.","PeriodicalId":431272,"journal":{"name":"2021 Sixth International Conference on Image Information Processing (ICIIP)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-11-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131044066","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：