Conference Papers

Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506

Browse

Has files: NoSubject: search.filters.subject.Super-resolution

Search Results

Now showing 1 - 7 of 7
  • No Thumbnail Available
    Item
    Automatic video object plane segmentation and super resolution: A novel approach
    (2009) Pais, A.R.; Pasupuleti, N.; Gidnavar, R.; D'Souza, J.
    Detecting, segmenting and super resolving of moving object is an important subject in computer visual analysis and surveillance. This paper discusses automatic VOP (Video Object Plane) segmentation and super resolving of VOP. We have proposed a new method for automatic generation of VOP by using block matching and quantization techniques. Edge model based approach is used to generate the HR (High Resolution) image with sharper VOP boundaries. Experimental results show that the proper VOP is segmented and super resolution of VOP yields lesser blurring compared to bicubic, bilinear interpolation schemes. Copyright © 2009 by IICAI.
  • No Thumbnail Available
    Item
    Single image super resolution from compressive samples using two level sparsity based reconstruction
    (Elsevier B.V., 2015) Nath, A.G.; Nair, M.S.; Rajan, J.
    Super Resolution based on Compressed Sensing (CS) considers low resolution (LR) image patch as the compressive samples of its high resolution (HR) patch. Compressed sensing based image acquisition systems acquire less number of random linear measurements without first collecting all the pixel values. But using these compressive measurements directly to reconstruct the image causes quality issues. In this paper an image super-resolution method with two level sparsity based reconstruction via patch based image interpolation and dictionary learning is proposed. The first level reconstruction generates a low resolution image from random samples and the interpolation scheme used in this algorithm reduces the HR-LR patch coherency due to neighborhood issue which is a major drawback of single image super resolution algorithms. The dictionary based reconstruction phase generates the high resolution image from the low resolution output of the first level reconstruction phase. The experimental results proved that the proposed two level reconstruction scheme recovers more details of the image and yields improved results from very few samples (around 35-45%) than the state-of-the-art algorithms which uses low resolution image itself as input. The results are compared by considering both PSNR values and visual perception. © 2015 The Authors.
  • No Thumbnail Available
    Item
    Neighbor embedding based super-resolution using residual luminance
    (Institute of Electrical and Electronics Engineers Inc., 2015) Mishra, D.; Majhi, B.; Sa, P.K.
    Resolution plays a crucial role for study of information in an image. Therefore to enhance the resolution of an image, there are so many techniques have been proposed with respect to the reference images. In this paper, we proposed a new scheme for single image super-resolution based on the neighbor embedding method. Many feature selection methods have been proposed for the learning based super-resolution using manifold learning. Here a new feature selection has been proposed by combining first-order gradient and residual of the luminance component, inspired by Gaussian pyramid. In this Neighbor Embedding based Super-Resolution using the Residual Luminance (NESRRL) method the high resolution targeted image is estimated by the training image pairs. This approach imposes the local compatibility and smoothness constraints between patches in the estimated high resolution image. The experimental results show the comparisons of qualitative performance of proposed method with different existing methods using Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index (SSIM). © 2014 IEEE.
  • No Thumbnail Available
    Item
    Single depth image super-resolution via high-frequency subbands enhancement and bilateral filtering
    (Institute of Electrical and Electronics Engineers Inc., 2016) Balure, C.S.; Ramesh Kini, M.; Bhavsar, A.
    This paper addresses the problem of super-resolution (SR) from a single low-resolution (LR) depth image to a high-resolution (HR) depth image. A simple yet effective method has been proposed using Discrete Wavelet Transform (DWT), Stationary Wavelet Transform (SWT), and by utilizing the gradient information of the interpolated LR image. We propose an intermediate stage to enhance the high-frequency subbands to recover the HR image for both noiseless and noisy scenarios. The proposed method has been validated on Middlebury dataset for different upsampling factors (i.e. 2, 4, 8) and is shown to be superior when compared with some related DWT and SWT based SR methods. We also demonstrate encouraging performance of the approach on noisy depth images. © 2016 IEEE.
  • No Thumbnail Available
    Item
    An Improved and Secure Visual Secret Sharing (VSS) scheme for Medical Images
    (Institute of Electrical and Electronics Engineers Inc., 2019) Mhala, N.C.; Pais, A.R.
    Nowadays, medical information is being shared over the communication networks due to ease of technology. The patient's medical information has to be securely communicated over the network for automatic diagnosis. Most of the communication networks are prone to attacks from an intruder thus compromising the security of patients data. Therefore, there is a need to transmit medical images securely over the network. Visual secret sharing scheme can be used to transmit the medical images over the network securely. Visual Secret Sharing (VSS) scheme generates multiple shares to share secret information among n participants. To recover the secret information, all shares should be stacked together. In our previous work [9], we proposed a VSS based technique to recover secret images with the contrast of 70-80% known as Randomized Visual Secret Sharing (RVSS) scheme. However, RVSS scheme suffers from problems like 1) Generation of blocking artifacts in the recovered images. 2) It recovers medical images with a maximum contrast of 30-40%, hence it is not suitable for medical images.In this paper, we propose a modified RVSS scheme to recover the medical images with improved contrast. The proposed scheme introduces the idea of using super-resolution concept to improve the contrast of reconstructed medical images. The reconstruction quality of the medical images is evaluated using Human Visual System (HVS) based parameters. Additionally, the performance of the proposed system is evaluated using the existing Computer Aided Diagnosis (CAD) systems. The experimental results showed that the proposed system is able to reconstruct the secret image with the contrast of almost 85-90% and similarity of almost 77%. AIso, the reconstructed images using the proposed system achieves the similar classification accuracy as that of existing CAD systems. © 2019 IEEE.
  • No Thumbnail Available
    Item
    A Novel Fingerprint Image Enhancement based on Super Resolution
    (Institute of Electrical and Electronics Engineers Inc., 2020) Muhammed, A.; Pais, A.R.
    Fingerprint is a most common and broadly accepted biometric trait used for personal authentication. In fingerprint-based authentication, the feature extraction module extract features, and these extracted characteristics are used for authentication. In fingerprints, the feature extraction module heavily depends on the status of the image. However, in practice, always getting a good quality fingerprint image is not possible. Moreover, a notable number of fingerprints collected are of poor quality. The accurate extraction of fingerprint characteristics from a lesser quality fingerprint image is a challenging problem. Fingerprint enhancement is introduced to resolve this issue. Hence in this paper, we introduce a fingerprint enhancement technique using a Deep Convolution Neural Network (DCNN), which improves image quality. The proposed method consists of super-resolution, followed by filtering and enhancement. The proposed method provides better results as compared with the conventional fingerprint enhancement methods. The experimental results determine that the proposed strategy improves the visual clarity of low-quality images and reduces the error rates during the fingerprint matching. © 2020 IEEE.
  • No Thumbnail Available
    Item
    Optimizing Super-Resolution Generative Adversarial Networks
    (Springer Science and Business Media Deutschland GmbH, 2023) Jain, V.; Annappa, B.; Dodia, S.
    Image super-resolution is an ill-posed problem because many possible high-resolution solutions exist for a single low resolution (LR) image. There are traditional methods to solve this problem, they are fast and straightforward, but they fail when the scale factor is high or there is noise in the data. With the development of machine learning algorithms, their application in this field is studied, and they perform better than traditional methods. Many Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs) have been developed for this problem. The Super-Resolution Generative Adversarial Networks (SRGAN) have proved to be significant in this area. Although the SRGAN produces good results with 4 upscaling, it has some shortcomings. This paper proposes an improved version of SRGAN with reduced computational complexity and training time. The proposed model achieved an PPSNR of 29.72 and SSIM value of 0.86. The proposed work outperforms most of the recently developed systems. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.