False Textures Research Articles

DVDR-SRGAN: Differential Value Dense Residual Super-Resolution Generative Adversarial Network

In the field of single-image super-resolution reconstruction, GAN can obtain the image texture more in line with the human eye. However, during the reconstruction process, it is easy to generate artifacts, false textures, and large deviations in details between the reconstructed image and the Ground Truth. In order to further improve the visual quality, we study the feature correlation between adjacent layers and propose a differential value dense residual network to solve this problem. We first use the deconvolution layer to enlarge the features, then extract the features through the convolution layer, and finally make a difference between the features before being magnified and the features after being extracted so that the difference can better reflect the areas that need attention. In the process of extracting the differential value, using the dense residual connection method for each layer can make the magnified features more complete, so the differential value obtained is more accurate. Next, the joint loss function is introduced to fuse high-frequency information and low-frequency information, which improves the visual effect of the reconstructed image to a certain extent. The experimental results on Set5, Set14, BSD100, and Urban datasets show that our proposed DVDR-SRGAN model is improved in terms of PSNR, SSIM, and LPIPS compared with the Bicubic, SRGAN, ESRGAN, Beby-GAN, and SPSR models.

Non-transfer Deep Learning of Optical Coherence Tomography for Post-hoc Explanation of Macular Disease Classification.

Deep transfer learning is a popular choice for classifying monochromatic medical images using models that are pretrained by natural images with color channels. This choice may introduce unnecessarily redundant model complexity that can limit explanations of such model behavior and outcomes in the context of medical imaging. To investigate this hypothesis, we develop a configurable deep convolutional neural network (CNN) to classify four macular disease conditions using retinal optical coherence tomography (OCT) images. Our proposed non-transfer deep CNN model (acc: 97.9%) outperforms existing transfer learning models such as ResNet-50 (acc: 89.0%), ResNet-101 (acc: 96.7%), VGG-19 (acc: 93.3%), Inception-V3 (acc: 95.8%) in the same retinal OCT image classification task. We perform post-hoc analysis of the trained model and model extracted image features, which reveals that only eight out of 256 filter kernels are active at our final convolutional layer. The convolutional responses of these selective eight filters yield image features that efficiently separate four macular disease classes even when projected onto two-dimensional principal component space. Our findings suggest that many deep learning parameters and their computations are redundant and expensive for retinal OCT image classification, which are expected to be more intense when using transfer learning. Additionally, we provide clinical interpretations of our misclassified test images identifying manifest artifacts, shadowing of useful texture, false texture representing fluids, and other confounding factors. These clinical explanations along with model optimization via kernel selection can improve the classification accuracy, computational costs, and explainability of model outcomes.

Dark Channel Prior Based Image Dehazing using Different Transmission Map Refinement Methods

Due to existence of haze, the image quality is degraded in the environment. Removal of haze is called dehazing. To dehaze an image Dark Channel Prior is recommended. Dark Channel Prior is an observation, that an image has few pixels whose intensity value is very small or near to zero in most non-sky patches. Such pixels are referred to as dark pixels. Dehazing through Dark Channel Prior is accomplished using four major steps. The steps include estimating atmospheric light, estimating transmission map, refinement of transmission map and image reconstruction. Incorrect estimation of transmission map may lead to some problems. These problems include false textures and blocking artifacts. Many methods are developed to further sharpen transmission map. Here transmission map is refined using soft matting, guided filter and bilateral filter. The comparison of dehazing methods has become difficult due to scarce availability of ground truth images .So we used I-HAZE, a new data set containing 35 picture pairs of hazy pictures and their respective ground truth pictures. A significant benefit of I-HAZE data set is that it allows us to compare different refinement methods used for dehazing with SSIM, PSNR and RMSE which are used for the measurement of finally obtained reconstructed image quality after the removal of haze.

Image denoising based on sparse representation and gradient histogram

Various image priors, such as sparsity prior, non-local self-similarity prior and gradient histogram prior, have been widely used for noise removal, while preserving the image texture. However, the gradient histogram prior used for texture enhancement sometimes generates false textures in the smooth areas. In order to address these problems, the authors propose a robust algorithm combining gradient histogram with sparse representation to obtain good estimates of the sparse coding coefficients of the latent image and realising image denoising while preserving the texture. The proposed model is solved by having a balance between over-enhancement and over-smoothing of the texture in order to preserve the natural texture appearance. Experimental results demonstrate the efficiency and effectiveness of the proposed method.

An Image De-Noising Algorithm Based on K-SVD and BM3D

The existence of noise affects the quality of the image seriously. The image de-noising algorithm based on KSVD appears fuzzy, where weak texture smooth area also can appear false textures, at the same time, when the noise was very big, the de-noising effect would not always be ideal. This paper proposed an image de-noising method based on K-SVD dictionary and BM3D. The algorithm can solve image weak texture fuzzy problems and weak edges effectively. The experimental results show that, compare with K-SVD de-noising algorithm, this algorithm has a good de-noising ability, which keeping the detail and the edge character of the image better.

Fractional-order bidirectional diffusion for image up-sampling

Following the recently proposed total variation (TV) up-sampling method, we present a novel image up-sampling algorithm based on fractional-order bidirectional diffusion. For the bidirectional diffusion, the forward diffusion occurs on the light side of edge, while the backward diffusion proceeds on the dark side. This bidirectional diffusion can reduce the edge width and avoid the appearances of false edge or texture and block effects. Moreover, the fractional-order derivative is used to avoid strong contrast near the edge in the interpolated images. The experiments show that, unlike the TV up-sampling (TVUP) method, the proposed algorithm does not suffer from the drift of edges, block effect in the smooth regions, false edges, and false texture.

A Noise-Aware Coding Scheme for Texture Classification

Texture-based analysis of images is a very common and much discussed issue in the fields of computer vision and image processing. Several methods have already been proposed to codify texture micro-patterns (texlets) in images. Most of these methods perform well when a given image is noise-free, but real world images contain different types of signal-independent as well as signal-dependent noises originated from different sources, even from the camera sensor itself. Hence, it is necessary to differentiate false textures appearing due to the noises, and thus, to achieve a reliable representation of texlets. In this proposal, we define an adaptive noise band (ANB) to approximate the amount of noise contamination around a pixel up to a certain extent. Based on this ANB, we generate reliable codes named noise tolerant ternary pattern (NTTP) to represent the texlets in an image. Extensive experiments on several datasets from renowned texture databases, such as the Outex and the Brodatz database, show that NTTP performs much better than the state-of-the-art methods.

NLM algorithm with weight update

The nonlocal means (NLM) filter is one of the most popular denoising approaches and there have been many improvements regarding its weight function and parameter optimisation. However, those improvements have not removed artifacts such as the false texture pattern, which occurs when the smoothing parameter in the weight function is small. The smoothness in the flat region and the sharpness in the texture region without the artifacts can be ensured by accurately estimated weights, which can be calculated with sufficient similar patches. In this reported work, a NLM algorithm with weight update is used to exploit the weights from relatively similar locations as well as the weights from the centre location. Experimental results demonstrate that the proposed method performed well.

Modified Algorithm with Bidirectional Diffusion for TV Image Interpolation

In this paper a method of image interpolation based on bidirectional diffusion is proposed. With this method, the edge width of interpolated images is effectively reduced and crisp and smooth edges are obtained. In this bidirectional diffusion, forward diffusion occurs in brighter lateral on edge ramp and backward diffusion proceeds in darker lateral. And the intensity of diffusion is adjusted adaptively according to the image features, which avoid the appearance of artifacts and false textures in the interpolated image. Numerical experiments on real images show that images interpolated with the proposed method have smaller edge width and are almost artifact-free.

Texture Difference Perception between Different Halftone Algorithms

Characterization of combination of different halftone algorithm is important for assess image halftone quality. A number of researches attempt to develop halftoning algorithm in order to improve halftone image quality. A main factor that influences the combination of AM and FM halftone transformation is the texture difference emerging in the gray level between AM and FM area. The paper purposes to investigate the false texture difference occurred in AM/FM halftone transformation by subjective evaluation. The relationships between the transformations of difference halftone algorithm and texture difference perception are discussed.

Design of Tone-Dependent Color-Error Diffusion Halftoning Systems

Grayscale error diffusion introduces nonlinear distortion (directional artifacts and false textures), linear distortion (sharpening), and additive noise. Tone-dependent error diffusion (TDED) reduces these artifacts by controlling the diffusion of quantization errors based on the input graylevel. We present an extension of TDED to color. In color-error diffusion, which color to render becomes a major concern in addition to finding optimal dot patterns. We propose a visually meaningful scheme to train input-level (or tone-) dependent color-error filters. Our design approach employs a Neugebauer printer model and a color human visual system model that takes into account spatial considerations in color reproduction. The resulting halftones overcome several traditional error-diffusion artifacts and achieve significantly greater accuracy in color rendition.

Application of Imaging Techniques to Geometry Analysis of Aggregate Particles

This paper presents image analysis techniques by which to characterize the texture, angularity, and form of aggregate particles used in highway construction and geotechnical applications. For texture analysis, wavelet decomposition in gray scale images of particles is performed. The results demonstrate that multiscale wavelet representation is a powerful tool by which to capture the texture and to differentiate true texture from false texture caused by variations of natural color on a particle surface. Angularity and form analyses of particles are done using binary images. A gradient-based method is employed to describe angularity. This method is shown to differentiate between particles with different angularity characteristics. Form analysis of the particles includes computing the shape factor and sphericity index, which are based on measurements of the shortest, intermediate, and longest axis of the particle. Particle thickness is measured using the feature of an autofocus microscope. The width and length are calculated by an eigenvalue decomposition method of two-dimensional particle projections. Details of an interactive software developed to compute the different aggregate shape factors are discussed. The results indicate that these calculated values of the particle dimensions match very closely the values measured manually using a digital caliper.

False pyroclastic textures in altered silicic lavas, with implications for volcanic-associated mineralization

There is strong emphasis in recent literature that Zn-Cu-Pb massive sulfide deposits associated with felsic volcanism (the kuroko-type deposits) are directly related to submarine calderas formed by large submarine pyroclastic eruptions. Altered silicic volcanic rocks with the appearance of welded and nonwelded pyroclastic flow deposits, volcaniclastic debris flows, and massive- to thin-bedded tuffaceous rocks form a major component of a Silurian succession that hosts Zn-Cu-Pb massive sulfide deposits at Benambra, southeastern Australia. However, critical evaluation of rock textures indicates that these silicic volcanic rocks are mainly lavas and associated autoclastic facies with remarkably deceptive false pyroclastic and volcaniclastic textures, including apparent pumice fiamme and glass shards. Intercalated sediments indicate that most of the silicic volcanics were emplaced in subaqueous settings ranging from relatively shallow water to deep marine.The false textures were produced mainly in originally glassy lava by the combined effects of devitrification, perlitic fracture, and pervasive hydrothermal alteration. These processes are related to the emplacement and cooling history of individual lavas and to a regional, essentially synvolcanic, hydrothermal system. Tectonic foliation of the more altered and mechanically weakened rocks has led to partial dismembering of phenocrysts and groundmass fabrics, thereby further enhancing clastic appearance.Subsequent to recognition of widespread false pyroclastic textures at Benambra, similar features have been identified in three other Australian massive sulfide districts. These results suggest that a reinvestigation of host rocks is warranted at other deposits where submarine pyroclastics are documented. Noncaldera settings combining subsidence in an extensional volcano-tectonic basin, passive submarine lava effusion within the basin, and ambient sedimentation derived from basin margins (including resedimented pyroclastic debris) may be more relevant models in a number of massive sulfide districts.