Skip to main content

Mehul Parikh

GTU, Information Technology, Faculty Member

Followers

2

Following

1

Public Views

InterestsView All (6)

Uploads

Papers by Mehul Parikh

Estimation of uncertainty in Brain Tumor segmentation using modified multistage 3D-UNet on multimodal MRI images

Journal of integrated science and technology, May 7, 2024

Automated brain tumor segmentation is challenging due to the tumor tissues' shape, size, and appe... more Automated brain tumor segmentation is challenging due to the tumor tissues' shape, size, and appearance. Various methods used multi-mode MRI scans to segment sub-regions of brain tumors. 3D CNN methods improved performance in recent years, but most methods do not use uncertainty information in segmentation. For reliability and understanding, model prediction is vital for clinical decisions. This work studies three models namely 3D-UNet, Modified 3D-UNet, and Modified Multistage-3D-UNet for brain tumor segmentation. MRI volume bias correction and normalization were carried out using z-score normalization. Two patch generation strategies reduce memory use and class imbalance. Voxel-wise uncertainty evaluation was made for aleatoric and epistemic uncertainties using test time augmentation and dropout, respectively. Variance and entropy are used to measure the uncertainty of the modified multistage-3D-Unet segmentation model from ground truth. Variance creates separate uncertainty maps for each tumor sub-regions, whereas, entropy provides only global information. Uncertainty is used to filter miss-segmented predictions and improve accuracy. Uncertainty awareness increases model accuracy with dice scores of 0.93, 0.91, and 0.83 for tumor sub-regions WT, TC, and ET respectively.

Twitter Popularity v/s Yelp Ratings: Predictive and Sentiment Analysis of Food Chains

International Journal of Advance Research and Innovative Ideas in Education, 2016

Yelp connects people to great local businesses. Twitter is another important social media portal ... more Yelp connects people to great local businesses. Twitter is another important social media portal where twitter users express their experiences in a few sentences called tweets. This paper, focus on the reviews for restaurants. This model predict the star ratings for the restaurant from the Tweets extracted from Twitter. Through current work here, this paper perform a predictive analysis between the Yelp ratings of among popular food chains vs their equivalent twitter recommendations. Three machine learning algorithms are used with the sentiment analysis. After analyzed the performance of each models, the best model for predicting the ratings from reviews is the Naive Bayes algorithm. Also, found that sentiment features are very useful for rating prediction.

Blur Estimation and Identification of Degraded Images

During the process of image acquisition, sometimes images are degraded by various reasons. Image ... more During the process of image acquisition, sometimes images are degraded by various reasons. Image restoration is a challenging task in the field of Image processing. The process of recovering such degraded or corrupted image is called Image Restoration. Restoration process improves the appearance of the image. The degraded image is the convolution of the original image, degraded function, and additive noise. Process of restoration is deconvolved this degraded image to obtain noiseless and deblurred original image. This paper proposed the easiest method for blur estimation and identification. Usual types of blur such as motion, defocus, Gaussian have been taken as Point Spread Function (PSF).

Multiple Image Encryption Using Chaotic Map And DNA Computing

Security of image and video data has become increasingly important for various applications such ... more Security of image and video data has become increasingly important for various applications such as internetbased communications, video conferencing, medical and military application etc. Encryption is usually recommended to solve this issue. Repeatedly, Encrypt the single image is very undesirable. So Multiple image encryption has received attention. Multiple image encryption based on mixed image element and chaos proposed by Zhang,Wang is less secure due to only the order of image blocks scrambled in algorithm and not a content of image.So to resolve that problem novel algorithm proposed in which image content is also scrambled using DNA computing. which will more dependent on plaintext to resist plaintext attack and reduce correlation between pixels to resist stastical attack.DNA technology has been used with chaotic cryptosystem to double assurance the security of image cryptosystem by chaotic system and DNA biological manipulation. So proposed algorithm will be more secure and ...

A Review of Recent Advancements in Infant Brain MRI Segmentation Using Deep Learning Approaches

Lecture notes in networks and systems, 2023

Brain Tumor Segmentation and Classification of Mri Images Using Gabor Filter and Machine Learning Techniques

Journal of critical reviews, 2020

A Survey on Multiple Image Encryption Using Chaos Based algorithms And DNA Computing

International journal of scientific research in science, engineering and technology, Jan 20, 2018

Due to the development in the field of network technology and multimedia applications, every minu... more Due to the development in the field of network technology and multimedia applications, every minute thousands of messages which can be text, images, audios, videos are created and transmitted over wireless network. So encryption is used to provide security .To ensure the security of image transmission, people have proposed many single-image encryption (SIE) algorithms. In the age of big data, although multiple images can be repeatedly encrypted by the SIE algorithm in theory, the encryption efficiency is always undesirable. In this paper different multiple image encryption technique based on chaotic map and DNA computing have been studied.

Analyses of Collaborative Filtering Using Item Clustering and Hybrid Clustering

International journal of engineering research and technology, Nov 25, 2013

Personalized recommendation systems are producing recommendation and widely used in today's world... more Personalized recommendation systems are producing recommendation and widely used in today's world. Collaborative filtering technique is most successful technique for recommendation.. Collaborative filtering is a method of making prediction about interest of user by collecting preferences from many users. The growth of users and products are increase very quickly and its challenge for nearest-neighbor filtering algorithm. Many algorithms proposed so far, where the principal concern is collaborative filtering challenges. This paper analyses the collaborative filtering challenges using clustering technology. This approach can be implemented based on user clustering, item clustering and another method is hybrid method which use user and item clustering.

Twitter Popularity v/s Yelp Ratings: Predictive and Sentiment Analysis of Food Chains

International Journal of Advance Research and Innovative Ideas in Education, 2016

Yelp connects people to great local businesses. Twitter is another import ant social media portal... more Yelp connects people to great local businesses. Twitter is another import ant social media portal where twitter users express their experiences in a few sentences called tweets. This paper, focus on the reviews for restaurants. This model predict the star ratings for the restaurant from the Tweets extracted from Twitter. Through current work here, this paper perform a predictive analysis between the Yelp ratings of among popular food chains vs their equivalent twitter recommendations. Three machine learning algorithms are used with the sentiment analysis. After analyzed the performance of each models, the best model for predicting the ratings from reviews is the Naïve Bayes algorithm. Also, found that sentiment features are very useful for rating prediction .

A Hybrid approach for Optical Character Verification

International Journal for Scientific Research and Development, May 6, 2017

At present scenario, there is growing demand for the software system to recognize characters in a... more At present scenario, there is growing demand for the software system to recognize characters in a computer system when information is scanned through paper documents. This paper presents detailed review in the field of Optical Character Recognition. Various techniques are determined that have been proposed to realize the center of character recognition in an optical character recognition system. OCR (Optical Character Recognition) translates images of typewritten or handwritten characters into the electronically editable format and it preserves font properties. Where OCV (Optical Character Verification) is a hybrid approach of OCR and pattern matching. Different techniques for pre-processing and segmentation have been surveyed and discussed. Proposed methodology and the dataset are presented here. Intermediate results have shown in this paper.

Optimized index construction for large text collections using blocked sort-based indexing

Indexes are the key technology corroborate efficient text search. This paper, present an optimize... more Indexes are the key technology corroborate efficient text search. This paper, present an optimized index construction strategies for file system search engine. Index construction can be done in two phase: apply analyzer rule and index compression respectively. For fast index construction in-memory blocked sort-based Index partitions approach is used. Inverted index data structure used to provide efficient query evaluation. Special attention is given to a particular case of desktop search system such as Windows Desktop Search and Apple spotlight. As one of the main result of this paper demonstrate the time complexity and space complexity of indexed file.

Comparative Analysis of Motion base Image Segmentation using Machine Learning Techniques

Indian journal of science and technology, Mar 14, 2016

A Review on Optical Character Recognition Techniques

International journal of computer applications, Feb 15, 2017

At present scenario, there is growing demand for the software system to recognize characters in a... more At present scenario, there is growing demand for the software system to recognize characters in a computer system when information is scanned through paper documents. This paper presents detailed review in the field of Optical Character Recognition. Various techniques are determined that have been proposed to realize the center of character recognition in an optical character recognition system. OCR (Optical Character Recognition) translates images of typewritten or handwritten characters into the electronically editable format and it preserves font properties. Different techniques for preprocessing and segmentation have been surveyed and discussed in this paper.

Various License Plate Detection and Recognition Methods using Computer Vision and Machine Learning

ITM Web of Conferences

With the increasing advancements in the technology, our lives have become significantly more conv... more With the increasing advancements in the technology, our lives have become significantly more convenient. We now have automated many things. One example of such things is the automated number plate recognition system. There are many ways to perform the ANPR (Automatic Number Plate Recognition). Performing ANPR in wild still remains a big challenge. This review focuses on some techniques that have tried to overcome this challenge.

Comparative Analysis for Moving Object Detection in Vibrant Background Using SVM, LS SVM and Random Forest Classifier

CRC Press eBooks, Jun 28, 2022

A Hybrid approach for Optical Character Verification

International Journal for Scientific Research and Development, 2017

At present scenario, there is growing demand for the software system to recognize characters in a... more At present scenario, there is growing demand for the software system to recognize characters in a computer system when information is scanned through paper documents. This paper presents detailed review in the field of Optical Character Recognition. Various techniques are determined that have been proposed to realize the center of character recognition in an optical character recognition system. OCR (Optical Character Recognition) translates images of typewritten or handwritten characters into the electronically editable format and it preserves font properties. Where OCV (Optical Character Verification) is a hybrid approach of OCR and pattern matching. Different techniques for pre-processing and segmentation have been surveyed and discussed. Proposed methodology and the dataset are presented here. Intermediate results have shown in this paper.

Optimized index construction for large text collections using blocked sort-based indexing

2014 IEEE International Conference on Advanced Communications, Control and Computing Technologies, 2014

Indexes are the key technology corroborate efficient text search. This paper, present an optimize... more Indexes are the key technology corroborate efficient text search. This paper, present an optimized index construction strategies for file system search engine. Index construction can be done in two phase: apply analyzer rule and index compression respectively. For fast index construction in-memory blocked sort-based Index partitions approach is used. Inverted index data structure used to provide efficient query evaluation. Special attention is given to a particular case of desktop search system such as Windows Desktop Search and Apple spotlight. As one of the main result of this paper demonstrate the time complexity and space complexity of indexed file.

Comparative Analysis of Motion base Image Segmentation using Machine Learning Techniques

Indian Journal of Science and Technology, 2016

The region-based segmentation method classifies the image in number of classes or regions. These ... more

A Parallel Hybrid Approach With MPI And OpenMP

International journal of scientific research in science, engineering and technology, Oct 31, 2017

Parallel Hybrid Computing has been an emerging research field in Parallel Computing and Multi Cor... more Parallel Hybrid Computing has been an emerging research field in Parallel Computing and Multi Core and Many Core Programming in recent years. It is the hybrid approach to make the combination of MPI (Message Passing Interface) and OpenMP(Open Multi-Processor) programming structure to enhance the speed and running time of the program. In the last few years, hybrid approach became the eye catching approach for the large and iterative programs to enhance the performance and reduce the delay. With PHC we can get the advantage of shared memory paradigm with the distributed memory system. with this hybrid approach we can enhance the performance of the sorting algorithm, iterative programs like Pi and Prime Numbers, and all computing algorithms. Paper will mainly focus on the comparison of the Pure MPI, OpenMP and the Hybrid Version of both approaches.

A Survey On Efficient Frequent Pattern Mining Techniques

International journal of scientific research in science, engineering and technology, 2018

Data Mining is the technique to abstract the useful data from the large dataset for different per... more Data Mining is the technique to abstract the useful data from the large dataset for different perspectives. Frequent pattern mining has become an important data mining technique to find the frequent patterns from the data set that appears frequently. Frequent Pattern Technique is widely used in financial, retail, telecommunication and many more. The major concern of these industries is faster processing of a very large amount of data. Various techniques and algorithms have been proposed for this purpose. Apriori, FP-tree are the pioneer techniques among them. In this paper, we have analysed algorithms for finding frequent patterns with the purpose of discovering how these algorithms can be used to obtain frequent patterns over large transactional databases with most efficient way in various aspects. This has been presented in the form of a comparative study of the following algorithms: Apriori, Frequent Pattern (FP) Growth, dNC-ECPM Algorithm, OCFP–growth, IA-TJ-FGTT(Important Attri...

Estimation of uncertainty in Brain Tumor segmentation using modified multistage 3D-UNet on multimodal MRI images

Journal of integrated science and technology, May 7, 2024

Automated brain tumor segmentation is challenging due to the tumor tissues' shape, size, and appe... more Automated brain tumor segmentation is challenging due to the tumor tissues' shape, size, and appearance. Various methods used multi-mode MRI scans to segment sub-regions of brain tumors. 3D CNN methods improved performance in recent years, but most methods do not use uncertainty information in segmentation. For reliability and understanding, model prediction is vital for clinical decisions. This work studies three models namely 3D-UNet, Modified 3D-UNet, and Modified Multistage-3D-UNet for brain tumor segmentation. MRI volume bias correction and normalization were carried out using z-score normalization. Two patch generation strategies reduce memory use and class imbalance. Voxel-wise uncertainty evaluation was made for aleatoric and epistemic uncertainties using test time augmentation and dropout, respectively. Variance and entropy are used to measure the uncertainty of the modified multistage-3D-Unet segmentation model from ground truth. Variance creates separate uncertainty maps for each tumor sub-regions, whereas, entropy provides only global information. Uncertainty is used to filter miss-segmented predictions and improve accuracy. Uncertainty awareness increases model accuracy with dice scores of 0.93, 0.91, and 0.83 for tumor sub-regions WT, TC, and ET respectively.

Twitter Popularity v/s Yelp Ratings: Predictive and Sentiment Analysis of Food Chains

International Journal of Advance Research and Innovative Ideas in Education, 2016

Yelp connects people to great local businesses. Twitter is another important social media portal ... more Yelp connects people to great local businesses. Twitter is another important social media portal where twitter users express their experiences in a few sentences called tweets. This paper, focus on the reviews for restaurants. This model predict the star ratings for the restaurant from the Tweets extracted from Twitter. Through current work here, this paper perform a predictive analysis between the Yelp ratings of among popular food chains vs their equivalent twitter recommendations. Three machine learning algorithms are used with the sentiment analysis. After analyzed the performance of each models, the best model for predicting the ratings from reviews is the Naive Bayes algorithm. Also, found that sentiment features are very useful for rating prediction.

Blur Estimation and Identification of Degraded Images

During the process of image acquisition, sometimes images are degraded by various reasons. Image ... more During the process of image acquisition, sometimes images are degraded by various reasons. Image restoration is a challenging task in the field of Image processing. The process of recovering such degraded or corrupted image is called Image Restoration. Restoration process improves the appearance of the image. The degraded image is the convolution of the original image, degraded function, and additive noise. Process of restoration is deconvolved this degraded image to obtain noiseless and deblurred original image. This paper proposed the easiest method for blur estimation and identification. Usual types of blur such as motion, defocus, Gaussian have been taken as Point Spread Function (PSF).

Multiple Image Encryption Using Chaotic Map And DNA Computing

Security of image and video data has become increasingly important for various applications such ... more Security of image and video data has become increasingly important for various applications such as internetbased communications, video conferencing, medical and military application etc. Encryption is usually recommended to solve this issue. Repeatedly, Encrypt the single image is very undesirable. So Multiple image encryption has received attention. Multiple image encryption based on mixed image element and chaos proposed by Zhang,Wang is less secure due to only the order of image blocks scrambled in algorithm and not a content of image.So to resolve that problem novel algorithm proposed in which image content is also scrambled using DNA computing. which will more dependent on plaintext to resist plaintext attack and reduce correlation between pixels to resist stastical attack.DNA technology has been used with chaotic cryptosystem to double assurance the security of image cryptosystem by chaotic system and DNA biological manipulation. So proposed algorithm will be more secure and ...

A Review of Recent Advancements in Infant Brain MRI Segmentation Using Deep Learning Approaches

Lecture notes in networks and systems, 2023

Brain Tumor Segmentation and Classification of Mri Images Using Gabor Filter and Machine Learning Techniques

Journal of critical reviews, 2020

A Survey on Multiple Image Encryption Using Chaos Based algorithms And DNA Computing

International journal of scientific research in science, engineering and technology, Jan 20, 2018

Due to the development in the field of network technology and multimedia applications, every minu... more Due to the development in the field of network technology and multimedia applications, every minute thousands of messages which can be text, images, audios, videos are created and transmitted over wireless network. So encryption is used to provide security .To ensure the security of image transmission, people have proposed many single-image encryption (SIE) algorithms. In the age of big data, although multiple images can be repeatedly encrypted by the SIE algorithm in theory, the encryption efficiency is always undesirable. In this paper different multiple image encryption technique based on chaotic map and DNA computing have been studied.

Analyses of Collaborative Filtering Using Item Clustering and Hybrid Clustering

International journal of engineering research and technology, Nov 25, 2013

Personalized recommendation systems are producing recommendation and widely used in today's world... more Personalized recommendation systems are producing recommendation and widely used in today's world. Collaborative filtering technique is most successful technique for recommendation.. Collaborative filtering is a method of making prediction about interest of user by collecting preferences from many users. The growth of users and products are increase very quickly and its challenge for nearest-neighbor filtering algorithm. Many algorithms proposed so far, where the principal concern is collaborative filtering challenges. This paper analyses the collaborative filtering challenges using clustering technology. This approach can be implemented based on user clustering, item clustering and another method is hybrid method which use user and item clustering.

Twitter Popularity v/s Yelp Ratings: Predictive and Sentiment Analysis of Food Chains

International Journal of Advance Research and Innovative Ideas in Education, 2016

Yelp connects people to great local businesses. Twitter is another import ant social media portal... more Yelp connects people to great local businesses. Twitter is another import ant social media portal where twitter users express their experiences in a few sentences called tweets. This paper, focus on the reviews for restaurants. This model predict the star ratings for the restaurant from the Tweets extracted from Twitter. Through current work here, this paper perform a predictive analysis between the Yelp ratings of among popular food chains vs their equivalent twitter recommendations. Three machine learning algorithms are used with the sentiment analysis. After analyzed the performance of each models, the best model for predicting the ratings from reviews is the Naïve Bayes algorithm. Also, found that sentiment features are very useful for rating prediction .

A Hybrid approach for Optical Character Verification

International Journal for Scientific Research and Development, May 6, 2017

At present scenario, there is growing demand for the software system to recognize characters in a... more At present scenario, there is growing demand for the software system to recognize characters in a computer system when information is scanned through paper documents. This paper presents detailed review in the field of Optical Character Recognition. Various techniques are determined that have been proposed to realize the center of character recognition in an optical character recognition system. OCR (Optical Character Recognition) translates images of typewritten or handwritten characters into the electronically editable format and it preserves font properties. Where OCV (Optical Character Verification) is a hybrid approach of OCR and pattern matching. Different techniques for pre-processing and segmentation have been surveyed and discussed. Proposed methodology and the dataset are presented here. Intermediate results have shown in this paper.

Optimized index construction for large text collections using blocked sort-based indexing

Indexes are the key technology corroborate efficient text search. This paper, present an optimize... more Indexes are the key technology corroborate efficient text search. This paper, present an optimized index construction strategies for file system search engine. Index construction can be done in two phase: apply analyzer rule and index compression respectively. For fast index construction in-memory blocked sort-based Index partitions approach is used. Inverted index data structure used to provide efficient query evaluation. Special attention is given to a particular case of desktop search system such as Windows Desktop Search and Apple spotlight. As one of the main result of this paper demonstrate the time complexity and space complexity of indexed file.

Comparative Analysis of Motion base Image Segmentation using Machine Learning Techniques

Indian journal of science and technology, Mar 14, 2016

A Review on Optical Character Recognition Techniques

International journal of computer applications, Feb 15, 2017

At present scenario, there is growing demand for the software system to recognize characters in a... more At present scenario, there is growing demand for the software system to recognize characters in a computer system when information is scanned through paper documents. This paper presents detailed review in the field of Optical Character Recognition. Various techniques are determined that have been proposed to realize the center of character recognition in an optical character recognition system. OCR (Optical Character Recognition) translates images of typewritten or handwritten characters into the electronically editable format and it preserves font properties. Different techniques for preprocessing and segmentation have been surveyed and discussed in this paper.

Various License Plate Detection and Recognition Methods using Computer Vision and Machine Learning

ITM Web of Conferences

With the increasing advancements in the technology, our lives have become significantly more conv... more With the increasing advancements in the technology, our lives have become significantly more convenient. We now have automated many things. One example of such things is the automated number plate recognition system. There are many ways to perform the ANPR (Automatic Number Plate Recognition). Performing ANPR in wild still remains a big challenge. This review focuses on some techniques that have tried to overcome this challenge.

Comparative Analysis for Moving Object Detection in Vibrant Background Using SVM, LS SVM and Random Forest Classifier

CRC Press eBooks, Jun 28, 2022

A Hybrid approach for Optical Character Verification

International Journal for Scientific Research and Development, 2017

At present scenario, there is growing demand for the software system to recognize characters in a... more At present scenario, there is growing demand for the software system to recognize characters in a computer system when information is scanned through paper documents. This paper presents detailed review in the field of Optical Character Recognition. Various techniques are determined that have been proposed to realize the center of character recognition in an optical character recognition system. OCR (Optical Character Recognition) translates images of typewritten or handwritten characters into the electronically editable format and it preserves font properties. Where OCV (Optical Character Verification) is a hybrid approach of OCR and pattern matching. Different techniques for pre-processing and segmentation have been surveyed and discussed. Proposed methodology and the dataset are presented here. Intermediate results have shown in this paper.

Optimized index construction for large text collections using blocked sort-based indexing

2014 IEEE International Conference on Advanced Communications, Control and Computing Technologies, 2014

Indexes are the key technology corroborate efficient text search. This paper, present an optimize... more Indexes are the key technology corroborate efficient text search. This paper, present an optimized index construction strategies for file system search engine. Index construction can be done in two phase: apply analyzer rule and index compression respectively. For fast index construction in-memory blocked sort-based Index partitions approach is used. Inverted index data structure used to provide efficient query evaluation. Special attention is given to a particular case of desktop search system such as Windows Desktop Search and Apple spotlight. As one of the main result of this paper demonstrate the time complexity and space complexity of indexed file.

Comparative Analysis of Motion base Image Segmentation using Machine Learning Techniques

Indian Journal of Science and Technology, 2016

The region-based segmentation method classifies the image in number of classes or regions. These ... more

A Parallel Hybrid Approach With MPI And OpenMP

International journal of scientific research in science, engineering and technology, Oct 31, 2017

Parallel Hybrid Computing has been an emerging research field in Parallel Computing and Multi Cor... more Parallel Hybrid Computing has been an emerging research field in Parallel Computing and Multi Core and Many Core Programming in recent years. It is the hybrid approach to make the combination of MPI (Message Passing Interface) and OpenMP(Open Multi-Processor) programming structure to enhance the speed and running time of the program. In the last few years, hybrid approach became the eye catching approach for the large and iterative programs to enhance the performance and reduce the delay. With PHC we can get the advantage of shared memory paradigm with the distributed memory system. with this hybrid approach we can enhance the performance of the sorting algorithm, iterative programs like Pi and Prime Numbers, and all computing algorithms. Paper will mainly focus on the comparison of the Pure MPI, OpenMP and the Hybrid Version of both approaches.

A Survey On Efficient Frequent Pattern Mining Techniques

International journal of scientific research in science, engineering and technology, 2018

Data Mining is the technique to abstract the useful data from the large dataset for different per... more Data Mining is the technique to abstract the useful data from the large dataset for different perspectives. Frequent pattern mining has become an important data mining technique to find the frequent patterns from the data set that appears frequently. Frequent Pattern Technique is widely used in financial, retail, telecommunication and many more. The major concern of these industries is faster processing of a very large amount of data. Various techniques and algorithms have been proposed for this purpose. Apriori, FP-tree are the pioneer techniques among them. In this paper, we have analysed algorithms for finding frequent patterns with the purpose of discovering how these algorithms can be used to obtain frequent patterns over large transactional databases with most efficient way in various aspects. This has been presented in the form of a comparative study of the following algorithms: Apriori, Frequent Pattern (FP) Growth, dNC-ECPM Algorithm, OCFP–growth, IA-TJ-FGTT(Important Attri...