Ashish M. Kothari
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Ashish M. Kothari has a total of 22 co-authors. Here are the top co-authors by number of shared publications:
- Rohit M. Thanki
KRiAN GmbH, Tappenbeck, Germany
18 shared publications - Ved Vyas Dwivedi
C. U. Shah University
7 shared publications Dwivedi Ved Vyas
2 shared publicationsVed Vyas Dwivedi
2 shared publicationsBalvant J Makwana
EC Department, Government Engineering College, Rajkot
1 shared publicationDeven N Trivedi
C. U. Shah University, Near Kothariya Village, Wadhwan City, Gujarat, India
1 shared publication
Latest Publications
- RetractionOpen Access
- 28 November 2024
- Published by Springer Nature in Journal of Ambient Intelligence and Humanized Computing
- Vol. 15 (S1) , 243
Retraction Note: Hybrid domain watermarking technique for copyright protection of images using speech watermarks
- 28 November 2024
- Published by Springer Nature in Journal of Ambient Intelligence and Humanized Computing
- Vol. 15 (S1) , 243
RetractionOpen AccessRetraction Note: Hybrid domain watermarking technique for copyright protection of images using speech watermarks
- Open Access
- 7 May 2024
- Published by Qeios Ltd
Review of: "CNN-Based Road Damage Detection"
The paper is well written and organized to take advantage of both software and hardware. The methods and algorithms used seem fine for the achievement of the objectives. The paper can be accepted for possible publication.
- 7 May 2024
- Published by Qeios Ltd
Open AccessReview of: "CNN-Based Road Damage Detection"
The paper is well written and organized to take advantage of both software and hardware. The methods and algorithms used seem fine for the achievement of the objectives. The paper can be accepted for possible publication.
- 4 October 2022
- Published by National Aerospace University - Kharkiv Aviation Institute in RADIOELECTRONIC AND COMPUTER SYSTEMS
- No. 3,p. 161-173
A novel approach for semantic segmentation of automatic road network extractions from remote sensing images by modified UNet
Accurate and up-to-date road maps are crucial for numerous applications such as urban planning, automatic vehicle navigation systems, and traffic monitoring systems. However, even in the high resolutions remote sensing images, the background and roads look similar due to the occlusion of trees and buildings, and it is difficult to accurately segment the road network from complex background images. In this research paper, an algorithm based on deep learning was proposed to segment road networks from remote sensing images. This semantic segmentation algorithm was developed with a modified UNet. Because of the lower availability of remote sensing images for semantic segmentation, the data augmentation method was used. Initially, the semantic segmentation network was trained by a large number of training samples using traditional UNet architecture. After then, the number of training samples is reduced gradually, and measures the performance of a traditional UNet model. This basic UNet model gives better results in the form of accuracy, IOU, DICE score, and visualization of the image for the 362 training samples. The idea here is to simply extract road data from remote sensing images. As a result, unlike traditional UNet, there is no need for a deeper neural network encoder-decoder structure. Hence, the number of convolutional layers in the modified UNet is lower than that in the standard UNet. Therefore, the complexity of the deep learning architecture and the training time required by the road network model was reduced. The model performance measured by the intersection over union (IOU) was 93.71% and the average segmentation time of a single image was 0.28 sec. The results showed that the modified UNet could efficiently segment road networks from remote sensing images with identical backgrounds. It can be used under various situations.
- 4 October 2022
- Published by National Aerospace University - Kharkiv Aviation Institute in RADIOELECTRONIC AND COMPUTER SYSTEMS
- No. 3,p. 161-173
A novel approach for semantic segmentation of automatic road network extractions from remote sensing images by modified UNet
Accurate and up-to-date road maps are crucial for numerous applications such as urban planning, automatic vehicle navigation systems, and traffic monitoring systems. However, even in the high resolutions remote sensing images, the background and roads look similar due to the occlusion of trees and buildings, and it is difficult to accurately segment the road network from complex background images. In this research paper, an algorithm based on deep learning was proposed to segment road networks from remote sensing images. This semantic segmentation algorithm was developed with a modified UNet. Because of the lower availability of remote sensing images for semantic segmentation, the data augmentation method was used. Initially, the semantic segmentation network was trained by a large number of training samples using traditional UNet architecture. After then, the number of training samples is reduced gradually, and measures the performance of a traditional UNet model. This basic UNet model gives better results in the form of accuracy, IOU, DICE score, and visualization of the image for the 362 training samples. The idea here is to simply extract road data from remote sensing images. As a result, unlike traditional UNet, there is no need for a deeper neural network encoder-decoder structure. Hence, the number of convolutional layers in the modified UNet is lower than that in the standard UNet. Therefore, the complexity of the deep learning architecture and the training time required by the road network model was reduced. The model performance measured by the intersection over union (IOU) was 93.71% and the average segmentation time of a single image was 0.28 sec. The results showed that the modified UNet could efficiently segment road networks from remote sensing images with identical backgrounds. It can be used under various situations.
- Open Access
- 1 February 2022
- Published by Springer Nature in Journal of King Saud University Computer and Information Sciences
- Vol. 34 (2) , 72-85
Zoom based image super-resolution using DCT with LBP as characteristic model
- 1 February 2022
- Published by Springer Nature in Journal of King Saud University Computer and Information Sciences
- Vol. 34 (2) , 72-85
Open AccessZoom based image super-resolution using DCT with LBP as characteristic model
- 13 December 2021
- Published by Novosibirsk State Technical University in Metal Working and Material Science
- Vol. 23 (4) , 21-32
Application of digital image processing technique in the microstructure analysis and the machinability investigation
Introduction. The world is at the stage of creating an interdisciplinary approach that will be implemented in metallurgical research. The paper formulates the technique of image analysis in the study of processing at different depths from the mold-metal interface. The purpose of the work. Processing of a cast-iron workpiece within the first 3.5 mm of thickness from the mold-metal interface is a serious problem of solid processing. The study of machinability at different depths is a key requirement of the industry for ease of processing. Machinability will determine a number of factors, including tool consumption, workpiece surface quality, energy consumption, etc. The method of investigation. Image analysis is performed to determine the percentage of graphite in etched and non-etched samples. K-means clustering allows to create a new image from a given one with a clear separation of white and black areas by converting a digital image into a binary image using a threshold value for segmentation. The volume fraction of perlite, the volume fraction of graphite and the average size of graphite flakes in microns are used as input variables for the machinability of cast iron. Results and discussion. The output, that is, the segmented image, will be the input function for calculating the workability index using formulas. Thus, microstructural analysis will help predict the workability index of grey cast iron ASTM A48 Class 20. Using this method and the program, based on the microstructure, it is possible to predict in advance the characteristics of the machining of the part, taking into account possible changes in the casting process itself.
- 13 December 2021
- Published by Novosibirsk State Technical University in Metal Working and Material Science
- Vol. 23 (4) , 21-32
Application of digital image processing technique in the microstructure analysis and the machinability investigation
Introduction. The world is at the stage of creating an interdisciplinary approach that will be implemented in metallurgical research. The paper formulates the technique of image analysis in the study of processing at different depths from the mold-metal interface. The purpose of the work. Processing of a cast-iron workpiece within the first 3.5 mm of thickness from the mold-metal interface is a serious problem of solid processing. The study of machinability at different depths is a key requirement of the industry for ease of processing. Machinability will determine a number of factors, including tool consumption, workpiece surface quality, energy consumption, etc. The method of investigation. Image analysis is performed to determine the percentage of graphite in etched and non-etched samples. K-means clustering allows to create a new image from a given one with a clear separation of white and black areas by converting a digital image into a binary image using a threshold value for segmentation. The volume fraction of perlite, the volume fraction of graphite and the average size of graphite flakes in microns are used as input variables for the machinability of cast iron. Results and discussion. The output, that is, the segmented image, will be the input function for calculating the workability index using formulas. Thus, microstructural analysis will help predict the workability index of grey cast iron ASTM A48 Class 20. Using this method and the program, based on the microstructure, it is possible to predict in advance the characteristics of the machining of the part, taking into account possible changes in the casting process itself.
- Open Access
- 1 March 2021
- Published by IGI Global Scientific Publishing in International Journal of Digital Crime and Forensics
- Vol. 13 (2) , 35-56
Fragile Watermarking Framework for Tamper Detection of Color Biometric Images
Application of fragile watermarking on biometric images stored at a server or cloud ensures proper authentication and tamper detection when access to the servers was shared. In this paper, a hybrid domain fragile watermarking technique for authenticity of color biometric images, using hybridization of various transforms such as discrete cosine transform (DCT), fast discrete curvelet transform (FDCuT), and singular value decomposition (SVD) is proposed. The hybrid transform coefficients are modified according to the scrambled color watermark to obtain watermarked color biometric image. The security of this technique is strengthened with the usage of Arnold scrambling, and by using multiple secret keys. The proposed technique is analyzed on FEI Brazilian face database. The experimental results show that this technique performs better than the existing fragile watermarking techniques.
- 1 March 2021
- Published by IGI Global Scientific Publishing in International Journal of Digital Crime and Forensics
- Vol. 13 (2) , 35-56
Open AccessFragile Watermarking Framework for Tamper Detection of Color Biometric Images
Application of fragile watermarking on biometric images stored at a server or cloud ensures proper authentication and tamper detection when access to the servers was shared. In this paper, a hybrid domain fragile watermarking technique for authenticity of color biometric images, using hybridization of various transforms such as discrete cosine transform (DCT), fast discrete curvelet transform (FDCuT), and singular value decomposition (SVD) is proposed. The hybrid transform coefficients are modified according to the scrambled color watermark to obtain watermarked color biometric image. The security of this technique is strengthened with the usage of Arnold scrambling, and by using multiple secret keys. The proposed technique is analyzed on FEI Brazilian face database. The experimental results show that this technique performs better than the existing fragile watermarking techniques.
- 19 July 2020
- Published by Springer Nature
- Vol. 1235, 290-297
Color Image Watermarking in Encryption Domain
Image watermarking is one of the techniques used for copyright protection of digital images. In this paper, copyright protection of color images using watermarking is presented with the help of data encryption. The main contribution in this algorithm is that the color image converts into an encrypted image using compressive sensing (CS) based encryption and then watermark logo is inserted into the encrypted image to get the watermarked encrypted image. After that, CS based decryption is applied to the watermarked version of an encrypted image to get a watermarked color image. Experimental results of this algorithm show that this technique effectively works for copyright protection of color images and provide better robustness compared to existing algorithms available in the literature.
- 19 July 2020
- Published by Springer Nature
- Vol. 1235, 290-297
Color Image Watermarking in Encryption Domain
Image watermarking is one of the techniques used for copyright protection of digital images. In this paper, copyright protection of color images using watermarking is presented with the help of data encryption. The main contribution in this algorithm is that the color image converts into an encrypted image using compressive sensing (CS) based encryption and then watermark logo is inserted into the encrypted image to get the watermarked encrypted image. After that, CS based decryption is applied to the watermarked version of an encrypted image to get a watermarked color image. Experimental results of this algorithm show that this technique effectively works for copyright protection of color images and provide better robustness compared to existing algorithms available in the literature.
- 1 June 2019
- Published by Elsevier in Journal of Information Security and Applications
- Vol. 46, 231-249
Hybrid and blind watermarking scheme in DCuT – RDWT domain
- 1 June 2019
- Published by Elsevier in Journal of Information Security and Applications
- Vol. 46, 231-249
Hybrid and blind watermarking scheme in DCuT – RDWT domain
- Retracted
- 16 April 2019
- Published by Springer Nature in Journal of Ambient Intelligence and Humanized Computing
- Vol. 11 (5) , 1835-1857
RETRACTED ARTICLE: Hybrid domain watermarking technique for copyright protection of images using speech watermarks
- 16 April 2019
- Published by Springer Nature in Journal of Ambient Intelligence and Humanized Computing
- Vol. 11 (5) , 1835-1857
RetractedRETRACTED ARTICLE: Hybrid domain watermarking technique for copyright protection of images using speech watermarks
- 15 April 2019
- Published by The World Academy of Research in Science and Engineering in International Journal of Advanced Trends in Computer Science and Engineering
- Vol. 8 (2) , 182-186
SVD Based Performance Improvement in Hiding a Message Behind an Image
- 15 April 2019
- Published by The World Academy of Research in Science and Engineering in International Journal of Advanced Trends in Computer Science and Engineering
- Vol. 8 (2) , 182-186
SVD Based Performance Improvement in Hiding a Message Behind an Image