Lanczos Interpolation Research Articles

Classification of Coffee Leaf Diseases using the Convolutional Neural Network (CNN) EfficientNet Model

Coffee leaf disease is a problem that needs attention because it affects the quality and productivity of the coffee harvest and is detrimental to farmers. Therefore, a system is needed to identify types of coffee leaf diseases using artificial intelligence. There are four types of coffee leaf diseases, namely Miner leaf, Phoma leaf, Rust leaf, and Nodisease leaf. The research used the EfficientNet Architecture Convolutional Neural Network (CNN) method to detect types of disease on coffee leaves. This method was chosen because it is capable and reliable in processing digital images for pattern recognition. The dataset used is 1,464 images with dimensions of 2048 x 1024 pixels with RGB type which are divided into 1,264 training data and 400 testing data. Several architectures used in EfficientNet are EfficientNet B0, EfficientNet B1, EfficientNet B2, EfficientNet B3, EfficientNet B4. Parameters used are Lanczos resampling, Epoch 25, Learning Rate 0.0001, Loss Function Sparse Categorical Cross Entropy, Optimizer Adam. The results of training data testing, namely the CNN EfficientNet B1 Architecture Model method, got the best accuracy of 97% and a loss of 0.1328 and testing data testing got an accuracy of 0.97% and a loss of 0.1328. The architecture of the EfficientNet B1 model is better than other architectural models, namely VGG16, ResNet50, MobileNetv2, EfficientNet B0, EfficientNet B2, EfficientNet B3, EfficientNet B4, EfficientNet B5, EfficientNet B6, EfficientNet B7.

Reasonable mechanical model on shallow tunnel excavation to eliminate displacement singularity caused by unbalanced resultant

When considering initial stress field in geomaterial, nonzero resultant of shallow tunnel excavation exists, which produces logarithmic items in complex potentials, and would further lead to a unique displacement singularity at infinity to violate geo-engineering fact in real world. The mechanical and mathematical reasons of such a unique displacement singularity in the existing mechanical models are elaborated, and a new mechanical model is subsequently proposed to eliminate this singularity by constraining far-field ground surface displacement, and the original unbalanced resultant problem is converted into an equilibrium one with mixed boundary conditions. To solve stress and displacement in the new model, the analytic continuation is applied to transform the mixed boundary conditions into a homogenerous Riemann-Hilbert problem with extra constraints, which is then solved using an approximate and iterative method with good numerical stability. The Lanczos filtering is applied to the stress and displacement solution to reduce the Gibbs phenomena caused by abrupt change of the boundary conditions along ground surface. Several numerical cases are conducted to verify the proposed mechanical model and the results strongly validate that the proposed mechanical model successfully eliminates the displacement singularity caused by unbalanced resultant with good convergence and accuracy to obtain stress and displacement for shallow tunnel excavation. A parametric investigation is subsequently conducted to study the influence of tunnel depth, lateral coefficient, and free surface range on stress and displacement distribution in geomaterial.

Sensitivity Analysis of a Damage Detection Method Through High-Resolution Photos on Various Statically Deflected Beams

In this manuscript, the authors deeply investigate and test a modern technique that allows the analysis of a structure starting from a photo to identify and locate damage present on it, rapidly and non-destructively without any physical interaction with the analyzed structure. The technique suitability is tested on four statically deformed beams, on which notches represent the defects. The core of the proposed method is the correlation between the curvature that each beam presents under load conditions and its flexural stiffness. The proposed methodology consists in taking a photo of the inflected beam; subsequently, the acquired photo is manipulated with specifically designed image processing tools, and the second derivative of the beam axis is estimated using two distinct numerical differentiator filters (Lanczos filter and Gaussian wavelets) along with suitable processing to reduce border distortions. The tests conducted demonstrate that it is possible, with an opportune static deflection amplitude, to accurately detect the position of the notch with the proposed procedure; a sensitivity analysis is also conducted by testing the procedure with different beam thicknesses, notch positions, and amplitude of the static deflection. Although the authors realize that the technique can generally require sensibly large displacements, the results seem promising.

KaRIn Noise Reduction Using a Convolutional Neural Network for the SWOT Ocean Products

The SWOT (Surface Water Ocean Topography) mission will provide high-resolution and two-dimensional measurements of sea surface height (SSH). However, despite its unprecedented precision, SWOT’s Ka-band Radar Interferometer (KaRIn) still exhibits a substantial amount of random noise. In turn, the random noise limits the ability of SWOT to capture the smallest scales of the ocean’s topography and its derivatives. In that context, this paper explores the feasibility, strengths and limits of a noise-reduction algorithm based on a convolutional neural network. The model is based on a U-Net architecture and is trained and tested with simulated data from the North Atlantic. Our results are compared to classical smoothing methods: a median filter, a Lanczos kernel smoother and the SWOT de-noising algorithm developed by Gomez-Navarro et al. Our U-Net model yields better results for all the evaluation metrics: 2 mm root mean square error, sub-millimetric bias, variance reduction by factor of 44 (16 dB) and an accurate power spectral density down to 10–20 km wavelengths. We also tested various scenarios to infer the robustness and the stability of the U-Net. The U-Net always exhibits good performance and can be further improved with retraining if necessary. This robustness in simulation is very encouraging: our findings show that the U-Net architecture is likely one of the best candidates to reduce the noise of flight data from KaRIn.

Zero-sample mural superresolution reconstruction for enhanced perceptual quality

Aiming at the problem of texture loss and poor perceptual quality in low-resolution mural images, this paper proposes a zero-sample mural superresolution reconstruction method called EPZSSR to enhance perceptual quality, and the specific model is obtained by training the image. The algorithm takes the zero-shot superresolution method as the framework, randomly cuts the original image into a 128 * 128 size, performs Gaussian blurring on the image, uses Lanczos interpolation to downsample the smooth image to reduce artifacts, and uses convolutional attention module and skip connection to optimize the network structure. SmoothL1Loss is used to enhance the robustness of the model, and the PI value is introduced as the perceptual quality evaluation index. The experimental results show that compared with other superresolution reconstruction algorithms, the peak signal-to-noise ratio of the algorithm in this paper is increased by 0.98–3.23 dB on average. The mural texture reconstruction effect is better, the PI value is reduced by 0.56 on average, the mural perception quality is better, and the running time is reduced by 89.68 s on average. It has a certain value for mural superresolution reconstruction.

A Study on Analysis Method of Sea Level Change in Korean Coastal Zone

In this study, the analysis of sea level change using Lanczos filtering, a filtering method of Fourier transform, was performed. The high frequency(short period) of tide was excluded by low-pass filter of Lanczos filtering, the analysis of sea level change was carried out considering only low frequency(long period). First, all tide data provided by the Korea Hydrographic and Oceanographic Administration(KHOA) were preprocessed to delete the error value, the astronomical tide and the meteorological tide repeated in short period were excluded by Lanczos filtering. The Mean Sea Level(MSL) for each tidal station was calculated using these results, the MSL tended to rise at most point in the Korean coastal zone. And the rate of sea level change, which means annual sea level change(mm/year), was estimated and compared with previous study results. The rate of sea level change was higher than values in past study, this means that future sea level change may be greater than previous expected. The analysis method applied to this study can be effectively used to improve the accuracy of prediction of sea level change.

Denoising of Impulse Noise using Partition-Supported Median, Interpolation and DWT in Dental X-Ray Images

The impulse noise often damages the human dental X-Ray images, leading to improper dental diagnosis. Hence, impulse noise removal in dental images is essential for a better subjective evaluation of human teeth. The existing denoising methods suffer from less restoration performance and less capacity to handle massive noise levels. This method suggests a novel denoising scheme called "Noise removal using Partition supported Median, Interpolation, and Discrete Wavelet Transform (NRPMID)" to address these issues. To effectively reduce the salt and pepper noise up to a range of 98.3 percent noise corruption, this method is applied over the surface of dental X-ray images based on techniques like mean filter, median filter, Bi-linear interpolation, Bi-Cubic interpolation, Lanczos interpolation, and Discrete Wavelet Transform (DWT). In terms of PSNR, IEF, and other metrics, the proposed noise removal algorithm greatly enhances the quality of dental X-ray images.

Reproducibility of CT-based radiomic features against image resampling and perturbations for tumour and healthy kidney in renal cancer patients

Computed Tomography (CT) is widely used in oncology for morphological evaluation and diagnosis, commonly through visual assessments, often exploiting semi-automatic tools as well. Well-established automatic methods for quantitative imaging offer the opportunity to enrich the radiologist interpretation with a large number of radiomic features, which need to be highly reproducible to be used reliably in clinical practice. This study investigates feature reproducibility against noise, varying resolutions and segmentations (achieved by perturbing the regions of interest), in a CT dataset with heterogeneous voxel size of 98 renal cell carcinomas (RCCs) and 93 contralateral normal kidneys (CK). In particular, first order (FO) and second order texture features based on both 2D and 3D grey level co-occurrence matrices (GLCMs) were considered. Moreover, this study carries out a comparative analysis of three of the most commonly used interpolation methods, which need to be selected before any resampling procedure. Results showed that the Lanczos interpolation is the most effective at preserving original information in resampling, where the median slice resolution coupled with the native slice spacing allows the best reproducibility, with 94.6% and 87.7% of features, in RCC and CK, respectively. GLCMs show their maximum reproducibility when used at short distances.

Effect of the Pixel Interpolation Method for Downsampling Medical Images on Deep Learning Accuracy

Background: High-resolution medical images often need to be downsampled because of the memory limitations of the hardware used for machine learning. Although various image interpolation methods are applicable to downsampling, the effect of data preprocessing on the learning performance of convolutional neural networks (CNNs) has not been fully investigated. Methods: In this study, five different pixel interpolation algorithms (nearest neighbor, bilinear, Hamming window, bicubic, and Lanczos interpolation) were used for image downsampling to investigate their effects on the prediction accuracy of a CNN. Chest X-ray images from the NIH public dataset were examined by downsampling 10 patterns. Results: The accuracy improved with a decreasing image size, and the best accuracy was achieved at 64 × 64 pixels. Among the interpolation methods, bicubic interpolation obtained the highest accuracy, followed by the Hamming window.

A Multiscale Clustering Approach for Non-IID Nominal Data.

Multiscale brings great benefits for people to observe objects or problems from different perspectives. Multiscale clustering has been widely studied in various disciplines. However, most of the research studies are only for the numerical dataset, which is a lack of research on the clustering of nominal dataset, especially the data are nonindependent and identically distributed (Non-IID). Aiming at the current research situation, this paper proposes a multiscale clustering framework based on Non-IID nominal data. Firstly, the benchmark-scale dataset is clustered based on coupled metric similarity measure. Secondly, it is proposed to transform the clustering results from benchmark scale to target scale that the two algorithms are named upscaling based on single chain and downscaling based on Lanczos kernel, respectively. Finally, experiments are performed using five public datasets and one real dataset of the Hebei province of China. The results showed that the method can provide us not only competitive performance but also reduce computational cost.

Inverse-Scaled Lanczos Filtering for Image Sharpening

The edge response in retinal image is the first step for human vision recognizing the outside world. A variety of receptive field models for describing the impulse response have been proposed. Which satisfies the uncertain principle? occupied the interest from a point of minimizing the product (Δx)(Δ w) both in spatial and spectral. Among the typical edge response models, finally Gabor function and 2nd. Gaussian Derivative GD2 remained as strong candidates. While famous D. Marr and R. Young support GD2, many vision researchers prefer Gabor. The retinal edge response model is used for image sharpening.<br/> Different from the conventional image sharpening filters, this paper proposes a novel image sharpening filter by modifying the Lanczos resampling filter. The Lanczos filter is used for image scaling to resize digital images. Usually it works to interpolate the discrete sampled points like as a kind of smoothing filter not as sharpening. The Lanczos kernel is given by the product of sampling Sinc function and the scaled Sinc function. The scaled Sinc function expanded by the scale "s" plays a role of window function. The author noticed that the inverse scaling of Lanczos window can be used not for smoothing but for sharpening filter.<br/> This paper demonstrates how the proposed model works effectively in comparison with Gabor and GD2.

On robustness of camera identification algorithms

In this paper we consider the problem of a privacy threat enabling tracing digital cameras by the analysis of pictures they produced. As thousands of images are processed at a mass scale, the threat may apply to most users of digital cameras. We consider a state-of-the-art algorithm for digital camera identification proposed in Lucas et al. (IEEE Trans Inf Forensics Secur 1(2):205–214, 2006) and discuss strategies that can be used to bypass it, in order to make information about the camera unavailable. It turns out that many natural strategies like Gaussian blur, adding artificial noise or removing pixels’ least significant bit from the image does not prevent the identification of a camera unless a huge loss of image details is suffered. On the other hand, we show a method to bypass the camera identification with a just marginally more complex, yet not intuitive, method namely cropping the image on the edges and resizing to the original size using Lanczos resampling.

Image resolution enhancement via sparse interpolation on wavelet domain

The image processing algorithms collectively known as super-resolution (SR) have proven effective in producing high-quality imagery from low-resolution (LR) images. This paper focuses on a novel image resolution enhancement method employing the wavelet domain techniques. In order to preserve more edge information, additional edge extraction step is proposed employing high-frequency (HF) sub-band images - low-high (LH), high-low (HL), and high-high (HH) - via the Discrete Wavelet Transform (DWT). In the designed procedure, the LR image is used in the sparse interpolation for the resolution-enhancement obtaining low-low (LL) sub- band. Additionally, all sub-bands (LL, LH, HL and HH) are performed via the Lanczos interpolation. Finally, the estimated sub-band images are used to form the new high-resolution (HR) image using the inverse DWT (IDWT). Experimental results on real data sets have confirmed the effectiveness of the proposed framework in terms of objective criteria as well as in subjective perception.

THE IMPACT OF USE DIFFERENT TYPE OF IMAGE INTERPOLATION METHODS ON THE ACCURACY OF THE RECONSTRUCTION OF SKULL ANATOMICAL MODEL

The paper presents comparative studies concerning the use different methods of image resampling on the accuracy of the reconstruction of oral and maxillofacial geometry. The study was conducted on 14 different patients. In order to extract the oral and maxillofacial models from DICOM data, a region-growing algorithm was used. Thresholds were set above 200 HU to select only craniofacial tissue. After an oral and maxillofacial tissue was segmented from DICOM data, a marching cubes algorithm was used for computing isosurfaces. Model with [Formula: see text][Formula: see text]mm voxel was chosen as the gold standard for models of [Formula: see text][Formula: see text]mm structure and improved with image resampling filters. In the study 7 different kernels were used allowing for filtration. The image resampling filters minimize maximum positive deviations, especially in the occipital, mandible and zygomatic bone area, and maximum negative deviations in the area of the maxilla and nasal bone. Lanczos filtering is the best method of interpolation as compared to other used methods, due to significantly increased visibility of the edges of the segmented structures. As a result of applying this method, partial volume effect artifact was minimized. The distributions and statistical parameters of resampled DICOM data prove that on this stage of data editing, it is possible to increase the accuracy of segmentation and reconstruction of the geometry.

M-GAN: Retinal Blood Vessel Segmentation by Balancing Losses Through Stacked Deep Fully Convolutional Networks

Until now, the human expert segments retinal blood vessels manually in fundus images to inspect human retinal-related diseases, such as diabetic retinopathy and vascular occlusion. Recently, many studies were conducted for automatic retinal vessel segmentation from fundus images through supervised and unsupervised methods to minimize user intervention. However, most of them lack in segmentation robustness and cannot optimize loss functions so that results of the segmentation have made lots of fake or thin branches. This article proposes a new conditional generative adversarial network called M-GAN to conduct accurate and precise retinal vessel segmentation by balancing losses through stacked deep fully convolutional networks. It consists of a newly designed M-generator with deep residual blocks for more robust segmentation and an M-discriminator with a deeper network for more efficient training of the adversarial model. In particular, a multi-kernel pooling block is added between the stacked layers to support the scale-invariance of vessel segmentations of different sizes. The M-generator has down-sampling layers to extract features and up-sampling layers to make segmented retinal blood vessel images from the extracted features. The M-discriminator also has a deeper network similar to the down-sampling of the M-generator, but the final layer is constructed as a fully connected layer for decision making. We conduct pre-processing of the input image using automatic color equalization (ACE) to make the retinal vessels of the input fundus image more clear and perform post-processing that makes the vessel branches smooth and reduces false-negatives using a Lanczos resampling method. To verify the proposed method, we used DRIVE, STARE, HRF, and CHASE-DB1datasets and compared the proposed M-GAN with other studies. We measured accuracy, the intersection of union (IoU), F1 score, and Matthews correlation coefficient (MCC) for comparative analysis. Results of comparison proved that the proposed M-GAN derived superior performance than other studies.

Convolutional Neural Networks With Intermediate Loss for 3D Super-Resolution of CT and MRI Scans

Computed Tomography (CT) scanners that are commonly-used in hospitals and medical centers nowadays produce low-resolution images, e.g. one voxel in the image corresponds to at most one-cubic millimeter of tissue. In order to accurately segment tumors and make treatment plans, radiologists and oncologists need CT scans of higher resolution. The same problem appears in Magnetic Resonance Imaging (MRI). In this paper, we propose an approach for the single-image super-resolution of 3D CT or MRI scans. Our method is based on deep convolutional neural networks (CNNs) composed of 10 convolutional layers and an intermediate upscaling layer that is placed after the first 6 convolutional layers. Our first CNN, which increases the resolution on two axes (width and height), is followed by a second CNN, which increases the resolution on the third axis (depth). Different from other methods, we compute the loss with respect to the ground-truth high-resolution image right after the upscaling layer, in addition to computing the loss after the last convolutional layer. The intermediate loss forces our network to produce a better output, closer to the ground-truth. A widely-used approach to obtain sharp results is to add Gaussian blur using a fixed standard deviation. In order to avoid overfitting to a fixed standard deviation, we apply Gaussian smoothing with various standard deviations, unlike other approaches. We evaluate the proposed method in the context of 2D and 3D super-resolution of CT and MRI scans from two databases, comparing it to related works from the literature and baselines based on various interpolation schemes, using 2× and 4× scaling factors. The empirical study shows that our approach attains superior results to all other methods. Moreover, our subjective image quality assessment by human observers reveals that both doctors and regular annotators chose our method in favor of Lanczos interpolation in 97.55% cases for an upscaling factor of 2× and in 96.69% cases for an upscaling factor of 4×. In order to allow others to reproduce our state-of-the-art results, we provide our code as open source at https://github.com/lilygeorgescu/3d-super-res-cnn.

Medical image interpolation based on 3D Lanczos filtering

ABSTRACTSeveral medical imaging techniques have been employed to create visual representations of internal anatomic structures, such as organs and tissues, for clinical procedures. Diagnosis and treatment of diseases have benefited from the use of imaging modalities, such as computed tomography, magnetic resonance imaging, ultrasonography, mammography, and photon emission tomography. Interpolation is a method for generating new data points on a regular grid from a set of known data points, as result of some image transformations, such as scaling and rotation. Several filters have been proposed to implement image interpolation. In this paper, we develop and evaluate a resampling method based on the three-dimensional Lanczos kernel. Experiments are conducted on several medical images and videos to demonstrate the effectiveness of the proposed interpolation approach.

Single Image Super-Resolution Algorithm Possessing Edge and Contrast Preservation

In real time surveillance video applications, it is often required to identify a region of interest in a degraded low resolution (LR) image. State-of-the-art super-resolution (SR) techniques produce images with poor illumination and degraded high frequency details. In this paper, we present a different approach for SISR by correcting the dual-tree complex wavelet transform (DT-CWT) subbands using the multi-stage cascaded joint bilateral filter (MSCJBF) and singular value decomposition (SVD). The proposed method exploits geometric regularity for implementing the covariance-based interpolation in the spatial domain. We decompose the interpolated LR image into different image and wavelet coefficients by employing DT-CWT. To preserve edges, we alter the wavelet sub-bands with the high frequency details obtained from the MSCJBF. Simultaneously, we retain uniform illumination by improving the image coefficients using SVD. In addition, the wavelet sub-bands undergo lanczos interpolation prior to the subband refinement. Experimental results demonstrate the effectiveness of our method.

Investigation of the Impact of Unsteady Turbulence Effects on the Aeroelastic Analysis of a Low-Pressure Turbine Rotor Blade

Abstract The impact of the unsteadiness in the considered turbulence quantities on the numerical prediction of the aeroelastic behavior of a low-pressure turbine (LPT) rotor blade is evaluated by means of a numerical study. In this context, one of the main objectives of this work is to compare different nonlinear harmonic balance (HB) approaches—one neglecting and one considering the unsteadiness in the employed turbulence models—with a conventional nonlinear solver of the unsteady Reynolds-averaged Navier–Stokes (URANS) equations in the time domain. In order to avoid unphysical oscillations in the turbulence quantities caused by the Gibbs phenomenon in the chosen HB approach, a filter method based on the Lanczos filter is developed. The developed filter method is applied in the course of the HB simulations considering the unsteadiness in the underlying turbulence model. Furthermore, the impact of its application on the solution of the flow field and on the unsteady surface pressure of the rotor blade, in particular, is discussed in the context of this work.

Numerical Analysis on the Effects of Binary Interaction between Typhoons Tembin and Bolaven in 2012

The binary interaction is one of the most challenging factors to improve the forecast accuracy of multiple tropical cyclones (TCs) in close vicinity. The effect of binary interaction usually results in anomalous track and variable intensity of TCs. A typical interaction type, one-way influence mode, has been investigated by many studies which mainly focused on the anomalous track and record-breaking precipitation, such as typhoons Morakot and Goni. In this paper, a typical case of this type, typhoons Tembin and Bolaven, occurred in the western North Pacific in August 2012, was selected to study how one typhoon impacts the track and intensity of the other one. The vortex of Tembin or Bolaven and the monsoon circulation were removed by a TC bogus scheme and a low-pass Lanczos filter, respectively, to carry out the numerical experiments. The results show that the presence of monsoon made the binary interaction more complex by affecting the tracks and the translation speeds of the TCs. The influence of Bolaven on the track of Tembin was more significant than the influence of Tembin on the track of Bolaven, and the looping track of Tembin was also affected by the external surrounding circulation associated with Bolaven. The absence of Tembin was not conducive to the development of Bolaven due to stronger vertical wind shear condition and the less kinetic energy being transported to upper troposphere. Note that the above analysis demonstrates the interacting processes between TCs and sheds some light on the prediction of binary TCs.