Edge-based image interpolation approach for video sensor network

Image resolution enhancement is a usable process for many image processing applications such as geoscience studies, astronomy and geographical information systems. One of the traditional methods used to increase the image resolution is image interpolation but the potential problem associated with it is to magnify the image many times without loss in image clarity. However, all the classical linear interpolation techniques like bilinear, bi-cubic interpolation methods generate blurred image. By employing dual-tree complex wavelet transform (DT-CWT) on a edge directional interpolation, it is possible to recover the high frequency components which provides an image with good visual clarity and thus super resolved high resolution images are obtained. The obtained simulation results comply with the above stated claim. A performance comparison of it is made with the recent work discussed in.

Satellite Image Resolution Enhancement Using Dyadic-integer Coefficients Based Bi-orthogonal Wavelet Filters

In this paper, an undecimated double density dual tree discrete wavelet transform (UDDDT-DWT) based image resolution enhancement technique is proposed. The critically sampled discrete wavelet transform (DWT) suffers from the drawbacks of being shift-variant and lacking the capacity to process directional information in images. The double density dual tree wavelet transform (DDDT-WT) is an approximately shift-invariant transform capturing directional information. The UDDDT-DWT is an improvement of the DDDT-DWT, making it exactly shift-invariant. The method uses a forward and inverse (UDDDT-DWT) to construct a high-resolution (HR) image from the given lower-resolution (LR) image. The HR image is reconstructed from the LR image using the inverse UDDDT-DWT. Results are presented and discussed through comparisons between state-of-the-art resolution enhancement methods.

In computer vision and image processing, image interpolation has been used widely to perform magnification by predicting missing details from a sampled image. Numerous techniques have been proposed to interpolate images and produce a better and clearly defined image. However, many conventional polynomial-based interpolation methods tend to smooth or blur out image details around the edges. Although a number of different interpolation techniques are available, all of them are somewhat limited in terms of their robustness and accuracy. In this paper, an edge-directed Lagrange time delay interpolation model is developed to confront the problem. The proficiency of the new technique is compared to some of the well-known conventional interpolation techniques. The newly proposed technique is shown to provide much better resultant images compared to the other conventional interpolation methods in terms of subjective and objective evaluations of mean squared error (MSE) and structural similarity (SSIM) tests.

Image super-resolution is a research endeavour that has gained notoriety in computer vision. The research goal is to increase the spatial dimensions of an image using corresponding low-resolution and high-resolution image pairs to enhance the perceptual quality. The challenge of maintaining such perceptual quality lies in developing appropriate algorithms that learn to reconstruct higher-quality images from their lower-resolution counterparts. Recent methods employ deep learning algorithms to reconstruct textural details prevalent in low-resolution images. Since corresponding image pairs are non-trivial to collect, researchers attempt super-resolution by creating synthetic low-resolution representations of high-resolution images. Unfortunately, such methods employ ineffective downscaling operations to achieve synthetic low-resolution images. These methods fail to generalize well on real-world images that may suffer different degradations. A different angle is offered to solve the task of image super-resolution by investigating the plausibility of learning the degradational operation using generative adversarial networks. A two-stage generative adversarial network along with two architectural variations is proposed to solve the task of real-world super-resolution from low-resolution images with unknown degradations. It is demonstrated that learning to downsample images in a weakly supervised manner is an impactful and viable approach for super-resolution.

This paper interprets image interpolation as a channel decoding problem and proposes a tanner graph based interpolation framework, which regards each pixel in an image as a variable node and the local image structure around each pixel as a check node. The pixels available from low-resolution image are received whereas other missing pixels of highresolution image are erased, through an imaginary channel. Local image structures exhibited by the low-resolution image provide information on the joint distribution of pixels in a small neighborhood, and thus play the same role as parity symbols in the classic channel coding scenarios. We develop an efficient solution for the sum-product algorithm of belief propagation in this framework, based on a gaussian auto-regressive image model. Initial experiments show up to 3dB gain over other methods with the same image model. The proposed framework is flexible in message processing at each node and provides much room for incorporating more sophisticated image modelling techniques.

Satellite images are corrupted by noise in its acquisition and transmission. The removal of noise from the image by attenuating the high frequency image components, removes some important details as well. In order to retain the useful information and improve the visual appearance, an effective denoising and resolution enhancement techniques are required. In this research, Hybrid Directional Lifting (HDL) technique is proposed to retain the important details of the image and improve the visual appearance. The Discrete Wavelet Transform (DWT) based interpolation technique is developed for enhancing the resolution of the denoised image. The performance of the proposed techniques are tested by Land Remote-Sensing Satellite (LANDSAT) images, using the quantitative performance measure, Peak Signal to Noise Ratio (PSNR) and computation time to show the significance of the proposed techniques. The PSNR of the HDL technique increases 1.02 dB compared to the standard denoising technique and the DWT based interpolation technique increases 3.94 dB. From the experimental results it reveals that newly developed image denoising and resolution enhancement techniques improve the image visual quality with rich textures.

Local spatial properties based image interpolation scheme using SVMs

In this paper, we propose a new super-resolution technique based on interpolation of the high-frequency subband images obtained by discrete wavelet transform (DWT) and the input image. The proposed technique uses DWT to decompose an image into different subband images. Then the high-frequency subband images and the input low-resolution image have been interpolated, followed by combining all these images to generate a new super-resolved image by using inverse DWT. The proposed technique has been tested on Lena, Elaine, Pepper, and Baboon. The quantitative peak signal-to-noise ratio (PSNR) and visual results show the superiority of the proposed technique over the conventional and state-of-art image resolution enhancement techniques. For Lena's image, the PSNR is 7.93 dB higher than the bicubic interpolation.

This paper proposes a novel edge-directed sub-pixel extraction method for still images. The basic idea is to decompose the mixed pixels present at the low spatial resolution images into four sub pixels using edge-directed models, thus to localize the mixed pixels in a smaller region. The pixels marked as edge pixels in low resolution images are mixture of different end-members with different spectral signature and a set of corresponding fractions, abundances. According to the linear spectrum mixing model, that is, the reflectance spectrum of a mixture is a systematic combination of the component reflectance spectra in the mixture and the systematic is basically linear, the sub-pixels can be extracted by quad-decomposition geometrically depend on the decomposition models directed by different edge-types in the 3 x 3 neighborhood centered on the mixed pixels. The algorithm can be used in image interpolation and image super-resolution construction. Super-resolution experiments demonstrate that the new algorithm substantially improves the subjective quality of the enlarged images over the conventional bilinear and bicubic interpolation.

Image interpolation using interpolative classified vector quantization

An Image Resolution Enhancement Technique based on Interpolation of the high frequency sub-band of colour images obtained by Discrete Wavelet Transform and the input colour image is proposed in this paper. Interpolation determines the intermediate values on the basis of observed values. One of the commonly used interpolation technique is Bicubic Interpolation. The edges are enhanced by introducing an intermediate stage by using Stationary Wavelet Transform. It is designed to overcome the lack of Translation-Invariance of Discrete Wavelet Transform. This is widely used in Signal Denoising and Pattern Recognition. Discrete Wavelet Transform is applied in order to decompose an input colour image into different sub-bands. Then the high frequency sub-bands as well as the input colour image are interpolated separately. The interpolated high frequency sub-bands and the Stationary Wavelet Transform high frequency sub-bands have the same size which means they can be added with each other. The new corrected high frequency sub-bands can be interpolated further for higher enlargement. Then all these sub-bands are combined with interpolated input image for new high resolution image by using Inverse Discrete Wavelet Transform. This has been done by MATLAB. The Peak Signal-Noise Ratio was obtained upto 5dB greater than the conventional and state-of-art image resolution enhancement techniques.

Wavelet transform is the result of inheritance and development of Fourier transform, which has a certain capability to analyze non-stationary signals. It is particularly suitable for image processing, such as non-stationary signals. In this paper, we propose an image resolution enhancement technique based on wavelet transform. The edges are enhanced by introducing an intermediate stage using stationary wavelet transform (SWT). The quantitative and visual results show the superiority of the proposed technique over the conventional image resolution enhancement techniques based on interpolation.

As an important image processing technique, image interpolation has wide applications, such as digital photos, video communication, medical imaging, object identification. A new adaptive image amplification technique is presented in this paper. The proposed method is able to cope with arbitrary integer multiples image amplification. For a low-resolution (LR) image and a given integer magnification factor n, every LR pixel is corresponding to a n ×n block of unknown HR pixels. The LR image is firstly zoomed n times by Bicubic interpolation. Then for every LR pixel belonging to edge area, weighted least-square estimation technique is adopted to compute its eight orientation model parameters for its associated n ×n block. Subsequently, each pixel in the associated n ×n block is substituted by a linear weighted summation using its eight surrounding HR pixels and the obtained model parameters. Finally, a correction strategy is implemented based on every n × n block to refine each tentative HR pixel. Compared with other image zooming methods, experimental results demonstrate the superiority of the proposed image zooming algorithm.

Spatial resolution in modern ultrasound imaging systems is limited by the high cost of large aperture transducer arrays, which require a large number of transducer elements and electronic channels. A new technique to enhance the spatial resolution of pulse-echo imaging systems is presented. The method attempts to build an image that could be obtained with a transducer array aperture larger than that physically available. We consider two images of the same object obtained with two different apertures, the full aperture and a subaperture, of the same transducer. A suitable artificial neural network (ANN) is trained to reproduce the relationship between the image obtained with the transducer full aperture and the image obtained with a subaperture. The inputs of the neural network are portions of the image obtained with the subaperture (low resolution image), and the target outputs are the corresponding portions of the image produced by the full aperture (high resolution image). After the network is trained, it can produce images with almost the same resolution of the full aperture transducer, but using a reduced number of real transducer elements. All computations are carried out on envelope-detected decimated images; for this reason, the computational cost is low and the method is suitable for real-time applications. The proposed method was applied to experimental data obtained with the ultrasound synthetic aperture focusing technique (SAFT), giving quite promising results. Real-time implementation on a modern, full-digital echographic system is currently being developed.