Pub Date : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776471
Aditya Rajput, Dinesh B. Kumbhar, Hrushikesh J. Joshi, Nandkishor M. Dhawale, Rewa H. Joshi, S. S. Patil
Flood is a extremely dangerous and harmful natural catastrophe. Various actions have been taken to reduce severity caused by flooding and flood damages. However, the problem of drainage remains one of the main challenge. Therefore, the main objective is to avoid or minimize the environmental impacts of flood hazards & to reduce the risk of flood and channel migration hazards. In order to address the problem, and to provide flood protection and decrease flood damage different types of works have been undertaken by engineers from all disciplines across the globe. This paper discusses about an interactive model illustrating the flood like situation and its solution to control and manage. The model is designed and created using instrumentation and control based hardware and software solutions to help manage the situation as well as to work like an emergency flood monitoring and warning system. This model is implemented with level monitoring and measurement system, alarm system, and use of LED's using which people can be warned and therefore help to reduce the after effects due to floods.
{"title":"Flood Level Control and Management Using Instrumentation and Control","authors":"Aditya Rajput, Dinesh B. Kumbhar, Hrushikesh J. Joshi, Nandkishor M. Dhawale, Rewa H. Joshi, S. S. Patil","doi":"10.1109/CCGE50943.2021.9776471","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776471","url":null,"abstract":"Flood is a extremely dangerous and harmful natural catastrophe. Various actions have been taken to reduce severity caused by flooding and flood damages. However, the problem of drainage remains one of the main challenge. Therefore, the main objective is to avoid or minimize the environmental impacts of flood hazards & to reduce the risk of flood and channel migration hazards. In order to address the problem, and to provide flood protection and decrease flood damage different types of works have been undertaken by engineers from all disciplines across the globe. This paper discusses about an interactive model illustrating the flood like situation and its solution to control and manage. The model is designed and created using instrumentation and control based hardware and software solutions to help manage the situation as well as to work like an emergency flood monitoring and warning system. This model is implemented with level monitoring and measurement system, alarm system, and use of LED's using which people can be warned and therefore help to reduce the after effects due to floods.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115248733","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776384
Sagar Mohite, Aditya Sarda, S. Joshi
Methodology of aspects is a combination of multiple concerns. Multiple types of concerns are facts, logic, area of interest, security, properties of systems. In the Development phase of the aspect model Aspect considers different systems concerns, aspects divide these system concerns into software modules, Different aspects modules use as a way for analysis of systems requirements. In some aspects, methodology use UML design modeling for understanding system requirements, in aspects Methodology UML class is for knowing the system necessities in modeling phases. a UML class structure is used in the Aspects process model for requirement analysis of the system. Class design in UML design consists of various types of attributes, classes, objects, methods, Join of points & Point of cut, various approaches to defining various concerns of the system. Few tools for crafting graph grammar rules for analysis system concerns, crafting G-graph grammar rules start from pre-condition G-grammar rules after that crafting post condition G-graph grammar rules, after crafting G-graph pre and post condition grammar rules G-graph transformation process done on rules in tools, next step is a method of creating a matrix, a matrix is basically cross applying rules to each other and find output, Two types of the matrix created first is a matrix of dependency and matrix of conflicts, this matrix for analysis conflict and dependency in crafted G-graph grammar rules, these G-rules apply as input to aspect methodology Tool. Next step transformation G-graph grammar, G-graph rules shows pre and post transformations of G-graph grammar rules when applying matrix of dependencies with a matrix of conflict, the conflict shows clash in G-graph rules, dependencies show requirement among the G-graph rules.
{"title":"Analysis of System Requirements by Aspects-J Methodology","authors":"Sagar Mohite, Aditya Sarda, S. Joshi","doi":"10.1109/CCGE50943.2021.9776384","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776384","url":null,"abstract":"Methodology of aspects is a combination of multiple concerns. Multiple types of concerns are facts, logic, area of interest, security, properties of systems. In the Development phase of the aspect model Aspect considers different systems concerns, aspects divide these system concerns into software modules, Different aspects modules use as a way for analysis of systems requirements. In some aspects, methodology use UML design modeling for understanding system requirements, in aspects Methodology UML class is for knowing the system necessities in modeling phases. a UML class structure is used in the Aspects process model for requirement analysis of the system. Class design in UML design consists of various types of attributes, classes, objects, methods, Join of points & Point of cut, various approaches to defining various concerns of the system. Few tools for crafting graph grammar rules for analysis system concerns, crafting G-graph grammar rules start from pre-condition G-grammar rules after that crafting post condition G-graph grammar rules, after crafting G-graph pre and post condition grammar rules G-graph transformation process done on rules in tools, next step is a method of creating a matrix, a matrix is basically cross applying rules to each other and find output, Two types of the matrix created first is a matrix of dependency and matrix of conflicts, this matrix for analysis conflict and dependency in crafted G-graph grammar rules, these G-rules apply as input to aspect methodology Tool. Next step transformation G-graph grammar, G-graph rules shows pre and post transformations of G-graph grammar rules when applying matrix of dependencies with a matrix of conflict, the conflict shows clash in G-graph rules, dependencies show requirement among the G-graph rules.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"366 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122814343","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776399
Vinit Khetani, Yatin Gandhi, Rashmi Patil
Language is means of communicating and expressing by using sound wherein, Sign language is communicating through hand gesture, face expression, fingers movement. This sign language is for deaf and dump people. It is visual method to communicate. Very less work has been done on identification of sign language. This paper mainly focuses on comparative study of Sign Language techniques. It reviews different techniques for feature extraction and classification of Sign Language from video or images. Recognition of sign language from video is one of the challenging tasks as video has to be converted to frames and that frames can be used for identification. In future, we will propose a system which will identify gesture from Indian sign language.
{"title":"A Study on Different Sign Language Recognition Techniques","authors":"Vinit Khetani, Yatin Gandhi, Rashmi Patil","doi":"10.1109/CCGE50943.2021.9776399","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776399","url":null,"abstract":"Language is means of communicating and expressing by using sound wherein, Sign language is communicating through hand gesture, face expression, fingers movement. This sign language is for deaf and dump people. It is visual method to communicate. Very less work has been done on identification of sign language. This paper mainly focuses on comparative study of Sign Language techniques. It reviews different techniques for feature extraction and classification of Sign Language from video or images. Recognition of sign language from video is one of the challenging tasks as video has to be converted to frames and that frames can be used for identification. In future, we will propose a system which will identify gesture from Indian sign language.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129822202","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776422
J. Wireko, Boahemaa Brenya, Ruchi Doshi
The Internet technology has had profound impact on our lives as well as changed the ways in which businesses interact with customers and has immense potential to improve the quality of education, a sustainable development goal set by the United Nations. Access to the Internet helps both educational administrators and students to reduce the costs of learning and improve the quality of education in schools and colleges. In developing countries, especially in Africa, the cost of accessing the Internet by students continues to be high and out of reach of many as one considers other key elements such as affordable broadband, computers, smartphone devices and the knowledge, skills, and competency to use the Internet technology. This paper examined the financial burden placed on students in accessing the Internet for their education at the tertiary level during the COVID-19 pandemic period. The results showed that tertiary students were financially impaired during the period as the cost of accessing the Internet (Internet data/subscription, Internet devices etc.) generally increased during the period.
{"title":"Financial Impact of Internet Access Infrastructure of Online Learning Mode on Tertiary Students in Covid-19 Era in Ghana","authors":"J. Wireko, Boahemaa Brenya, Ruchi Doshi","doi":"10.1109/CCGE50943.2021.9776422","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776422","url":null,"abstract":"The Internet technology has had profound impact on our lives as well as changed the ways in which businesses interact with customers and has immense potential to improve the quality of education, a sustainable development goal set by the United Nations. Access to the Internet helps both educational administrators and students to reduce the costs of learning and improve the quality of education in schools and colleges. In developing countries, especially in Africa, the cost of accessing the Internet by students continues to be high and out of reach of many as one considers other key elements such as affordable broadband, computers, smartphone devices and the knowledge, skills, and competency to use the Internet technology. This paper examined the financial burden placed on students in accessing the Internet for their education at the tertiary level during the COVID-19 pandemic period. The results showed that tertiary students were financially impaired during the period as the cost of accessing the Internet (Internet data/subscription, Internet devices etc.) generally increased during the period.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"171 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131417167","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776355
B. Mahajan, S. Shastri, S. Londhe
Present research work was done for 23.2 Km stretch of river Pavana from Kiwale to Dapodi in District Pune, Maharashtra. In this research, discharge measurement was done for 19 waste water streams in Pimpri Chinchwad city area. These waste water streams are continuously disposing untreated waste water in River Pavana. Total pollutional load was calculated by using characteristic values such as Biochemical Oxygen Demand, Chemical Oxygen Demand, Total Suspended Solids along with the discharge measurement of waste water stream. Area velocity method was adopted for measurement of waste water stream discharge. BODmix and Degree of Treatment required were calculated for River Pavana. BODmix of river water of River Pavana (stretch from Ravet to Dapodi) was compared with raw water quality standards.
{"title":"Impact Assessment of Waste Water Streams on Water Quality of River Pavana in Pimpri Chinchwad City Area","authors":"B. Mahajan, S. Shastri, S. Londhe","doi":"10.1109/CCGE50943.2021.9776355","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776355","url":null,"abstract":"Present research work was done for 23.2 Km stretch of river Pavana from Kiwale to Dapodi in District Pune, Maharashtra. In this research, discharge measurement was done for 19 waste water streams in Pimpri Chinchwad city area. These waste water streams are continuously disposing untreated waste water in River Pavana. Total pollutional load was calculated by using characteristic values such as Biochemical Oxygen Demand, Chemical Oxygen Demand, Total Suspended Solids along with the discharge measurement of waste water stream. Area velocity method was adopted for measurement of waste water stream discharge. BODmix and Degree of Treatment required were calculated for River Pavana. BODmix of river water of River Pavana (stretch from Ravet to Dapodi) was compared with raw water quality standards.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"13 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132796773","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776442
Neha Salunke, Sayali Mane, Samruddhi Kulkarni, S. Rao, Mahendra Deore
Healthcare sector is one of the fastest growing sector, which is making extensive use of technologies like IoT, Blockchain etc. Since there is rise in the use of e-healthcare system, huge amount of data is being generated so it has become necessary to protect the data from various kinds of cyber-attacks. Due to different cyber-attacks the CIA (Confidentiality, integrity, Availability) of the system can be compromised hence efficient mechanisms are required to ensure security. In this paper we have demonstrated a system which examines trust evaluation algorithm against various attacks. Simulation of the sensor cloud system is done and its results are analyzed. Trusts of individual nodes were calculated and trusted path calculation for sensors is done. Proposed system detects malicious nodes in order to protect network from insider attacks viz; black hole attack, on-off attack and replicated mobile sink attack, etc.
{"title":"Privacy Preserving Computation on Health Care Data","authors":"Neha Salunke, Sayali Mane, Samruddhi Kulkarni, S. Rao, Mahendra Deore","doi":"10.1109/CCGE50943.2021.9776442","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776442","url":null,"abstract":"Healthcare sector is one of the fastest growing sector, which is making extensive use of technologies like IoT, Blockchain etc. Since there is rise in the use of e-healthcare system, huge amount of data is being generated so it has become necessary to protect the data from various kinds of cyber-attacks. Due to different cyber-attacks the CIA (Confidentiality, integrity, Availability) of the system can be compromised hence efficient mechanisms are required to ensure security. In this paper we have demonstrated a system which examines trust evaluation algorithm against various attacks. Simulation of the sensor cloud system is done and its results are analyzed. Trusts of individual nodes were calculated and trusted path calculation for sensors is done. Proposed system detects malicious nodes in order to protect network from insider attacks viz; black hole attack, on-off attack and replicated mobile sink attack, etc.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132089129","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776403
Anjali Saner, Ashish Sharma, Atharva Patil, Avinash Soni, Pranav More
In this paper, we have introduce a real-time color recognition framework using openCv and PyObjC API. Hand Gestures with color detection are a natural form of communication between humans as well as between humans and machines. They are also useful in fields like sign language recognition, human-machine interface, and behavior understanding. The focus of creating colors recognition to create better communication between humans and computers for conveying information. The proposed project aims to create a video game controller wherein we can use our color detection and play the video Game on the computer. Several combinations of hand motions can be detected using deep learning and neural networks, rather than simple image processing and machine learning.
{"title":"Dynamic Color Recognition for Video Game Controller","authors":"Anjali Saner, Ashish Sharma, Atharva Patil, Avinash Soni, Pranav More","doi":"10.1109/CCGE50943.2021.9776403","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776403","url":null,"abstract":"In this paper, we have introduce a real-time color recognition framework using openCv and PyObjC API. Hand Gestures with color detection are a natural form of communication between humans as well as between humans and machines. They are also useful in fields like sign language recognition, human-machine interface, and behavior understanding. The focus of creating colors recognition to create better communication between humans and computers for conveying information. The proposed project aims to create a video game controller wherein we can use our color detection and play the video Game on the computer. Several combinations of hand motions can be detected using deep learning and neural networks, rather than simple image processing and machine learning.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132720705","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776430
Nikita Deshmukh, Anamika Ahire, S. Bhandari, Apurva Mali, Kalyani Warkari
Lip reading is an approach for understanding speech by visually interpreting lip movements. Vision based lip leading system takes a video (without audio) as an input of a person speaking some word or phrase and provides the predicted word or phrase the person is speaking as output. This paper presents the method for Vision based Lip Reading system that uses convolutional neural network (CNN) with attention-based Long Short-Term Memory (LSTM). The dataset includes video clips pronouncing single digits. The pretrained CNN is used for extracting features from pre-processed video frames which then are processed for learning temporal characteristics by LSTM. The SoftMax layer of architecture provides the result of lip reading. In the present work experiments are performed with two pre-trained models namely VGG19 and ResNet50 and the results are compared. To further improve the performance of the system ensembled learning is also used. The system provides 85% accuracy using ResNet50 and ensemble learning.
{"title":"Vision based Lip Reading System using Deep Learning","authors":"Nikita Deshmukh, Anamika Ahire, S. Bhandari, Apurva Mali, Kalyani Warkari","doi":"10.1109/CCGE50943.2021.9776430","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776430","url":null,"abstract":"Lip reading is an approach for understanding speech by visually interpreting lip movements. Vision based lip leading system takes a video (without audio) as an input of a person speaking some word or phrase and provides the predicted word or phrase the person is speaking as output. This paper presents the method for Vision based Lip Reading system that uses convolutional neural network (CNN) with attention-based Long Short-Term Memory (LSTM). The dataset includes video clips pronouncing single digits. The pretrained CNN is used for extracting features from pre-processed video frames which then are processed for learning temporal characteristics by LSTM. The SoftMax layer of architecture provides the result of lip reading. In the present work experiments are performed with two pre-trained models namely VGG19 and ResNet50 and the results are compared. To further improve the performance of the system ensembled learning is also used. The system provides 85% accuracy using ResNet50 and ensemble learning.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"45 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124852420","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776476
Amit Hire, J. Katti
The most prevalent form of authentication or computer system intrusion protection is the use of alphanumeric usernames and passwords. Strong password schemes have traditionally provided a certain level of protection, but they are complicated and difficult to remember, prompting their owners to write them down on paper. As a consequence, protection is jeopardized.Increasing the sharing of information, internet use, data transfer, protection and validation of passwords have become significant challenges.As a result, many programmes use different hashing and salting algorithms to store password files in databases.Attackers, on the other hand, do not consider password hashing to be safe since they use brute force attacks, password guessing attacks, and other methods to obtain the user's password from the database's password log.Some systems store passwords in databases using honeywords or bogus passwords generated using the generation technique for honeywords to prevent hacker attacks.In our recommended approach, passwords are saved using a distinct hashing mechanism with a less complexity of time in which majority of the processes need basic operation based on binarys.
{"title":"Security of User Credentials on Web Portals","authors":"Amit Hire, J. Katti","doi":"10.1109/CCGE50943.2021.9776476","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776476","url":null,"abstract":"The most prevalent form of authentication or computer system intrusion protection is the use of alphanumeric usernames and passwords. Strong password schemes have traditionally provided a certain level of protection, but they are complicated and difficult to remember, prompting their owners to write them down on paper. As a consequence, protection is jeopardized.Increasing the sharing of information, internet use, data transfer, protection and validation of passwords have become significant challenges.As a result, many programmes use different hashing and salting algorithms to store password files in databases.Attackers, on the other hand, do not consider password hashing to be safe since they use brute force attacks, password guessing attacks, and other methods to obtain the user's password from the database's password log.Some systems store passwords in databases using honeywords or bogus passwords generated using the generation technique for honeywords to prevent hacker attacks.In our recommended approach, passwords are saved using a distinct hashing mechanism with a less complexity of time in which majority of the processes need basic operation based on binarys.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124088583","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 : 2021-09-23DOI: 10.1109/CCGE50943.2021.9776351
D. Salunke, Pooja Sabne, Hitesh Saini, Vivekanand Shivanagi, Pradnya Jadhav
Devanagari language comprises 47 primary characters that include 14 vowels and 33 consonants. Nearly 120 scripts have been developed from it like Marathi, Hindi, Bengali, etc. This paper deals with a Marathi handwritten word recognition system. Marathi is an Indian language that is spoken primarily in the state of Maharashtra, most of the government paperwork is done in this language only. This system will mainly be used in recognizing various handwritten words present in the titles of different municipal documents. The main concern regarding these documents was the task of storing them digitally. The existing works have used the method of segmentation and have proposed a hidden Markov model to recognize pseudo characters, but the accuracy can be increased. We have devised a system that will overcome this limitation. In our system, a dataset is created for 2000 samples and by applying different augmentation techniques the dataset size is further increased to 12000 images. The proposed system is implemented using deep learning based Customized Convolution Neural Network which is pre-trained on a manually collected dataset containing various Marathi handwritten words. Out of which 75% of the dataset is used for training and 25% for testing purposes. This system has proved better results on augmented data with 94% accuracy.
{"title":"Handwritten Devanagari Word Recognition using Customized Convolution Neural Network","authors":"D. Salunke, Pooja Sabne, Hitesh Saini, Vivekanand Shivanagi, Pradnya Jadhav","doi":"10.1109/CCGE50943.2021.9776351","DOIUrl":"https://doi.org/10.1109/CCGE50943.2021.9776351","url":null,"abstract":"Devanagari language comprises 47 primary characters that include 14 vowels and 33 consonants. Nearly 120 scripts have been developed from it like Marathi, Hindi, Bengali, etc. This paper deals with a Marathi handwritten word recognition system. Marathi is an Indian language that is spoken primarily in the state of Maharashtra, most of the government paperwork is done in this language only. This system will mainly be used in recognizing various handwritten words present in the titles of different municipal documents. The main concern regarding these documents was the task of storing them digitally. The existing works have used the method of segmentation and have proposed a hidden Markov model to recognize pseudo characters, but the accuracy can be increased. We have devised a system that will overcome this limitation. In our system, a dataset is created for 2000 samples and by applying different augmentation techniques the dataset size is further increased to 12000 images. The proposed system is implemented using deep learning based Customized Convolution Neural Network which is pre-trained on a manually collected dataset containing various Marathi handwritten words. Out of which 75% of the dataset is used for training and 25% for testing purposes. This system has proved better results on augmented data with 94% accuracy.","PeriodicalId":130452,"journal":{"name":"2021 International Conference on Computing, Communication and Green Engineering (CCGE)","volume":"23 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127983583","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：