Performance analysis of texture characterization techniques for lung nodule classification

Lung cancer ranks very high on a global index for cancer-related casualties. With early detection of lung cancer, the rate of survival increases to 80-90%. The standard method for diagnosing lung cancer from Computed Tomography (CT) scans is by manual annotation and detection of the cancerous regions, which is a tedious task for radiologists. This paper proposes a machine learning approach for multi-class classification of the lung nodules into solid, semi-solid, and Ground Glass Object texture classes. We employ feature extraction techniques, such as gray-level co-occurrence matrix, Gabor filters, and local binary pattern, and validate the performance on the LNDb dataset. The best performing classifier displays an accuracy of 94% and an F1-score of 0.92. The proposed approach was compared with related work using the same dataset. The results are promising, and the proposed method can be used to diagnose lung cancer accurately.

Multi-Camera Workflow Applied to a Cultural Heritage Building: Alhambra’s Torre de la Cautiva from the Inside

It is increasingly necessary to generate accessible and navigable digital representations of historical or heritage buildings. This article explains the workflow that was applied to create such a digital component for one of the least accessible areas of the Alhambra palace in Granada, the so-called Torre de la Cautiva (Tower of the Captive). The main goal of this process was to create affordable, photorealistic 3D models that contribute to the dissemination of cultural heritage, the decision making for its conservation and restoration, and public engagement and entertainment. With enough preparation, the time spent gathering data following a Structure from Motion (SfM) approach can be significantly reduced by using a multi-camera (low cost DSLR) photogrammetric strategy. Without the possibility of artificial lighting, it was essential to use RAW images and calibrate the color in the scene for material and texture characterization. Through processing, the amount of data was reduced by optimizing the model’s topology. Thus, a photorealistic result was obtained that could be managed and visualized in immersive Visual Reality (VR) environments, simulating different historical periods and environmental and lighting conditions. The potential of this method allows, with slight modifications, the creation of HBIMs and the adaptation to VR systems development, whose current visualization quality is below the resolution of actionable models in rendering engines.

Effect of Secondary Cold Reduction Rates on Microstructure, Texture and Earing Behavior of Double Reduction Tinplate

In order to evaluate the effect of secondary cold reduction rate on the drawing performance of double reduction tinplate and explain the mechanism, a detailed investigation into the microstructural characterization, dissolved carbon atoms, texture characterization by an X-ray powder diffractometer (XRD) and electron backscatter diffraction (EBSD), and earing behavior were carried out with different secondary cold reduction rates of 15%, 20% and 25% for double reduction tinplate. The experimental results indicate that 15% secondary cold reduction rate could obtain a better drawing performance because there are no holes and cracks at the microstructure, and the content of dissolved carbon atom is relatively low; at the same time, it has a better texture distribution and low earing coefficient.

Texture characterization of some large hailstones with an automated technique

Hailstone structures have been studied for over a century, but so far mainly by manual optical means. This paper presents new texture and microstructure data (i.e. crystal lattice orientations, grain sizes and shapes) measured with an Automatic Ice Texture Analyzer, which gives access to high spatial and angular resolutions. The hailstones show two main characteristics: (1) they are structured with several concentric layers composed of alternating fine equiaxed grains and coarse elongated and radially oriented grains, and (2) they show two texture types with c-axes oriented either parallel or perpendicular to the radial direction. Such textures are compared with the ones observed in lake S1 and S2 ices, respectively. The S1 texture type (with c-axes parallel to the columnar crystals that grew in the radial direction) may result from epitaxial growth from a polycrystalline embryo, while the S2 texture (c-axes in the plane perpendicular to the column direction) may result from the growth from an embryo made of a few crystals with mainly one crystallographic orientation. Our novel high-resolution maps and measurements of both microstructure and texture may help to shed new light on the long-term discussion on the growth mechanisms of large hailstones.

Effect of Cut-Off, Evaluation Length, and Measurement Area in Profile and Areal Surface Texture Characterization of As-Built Metal Additive Manufactured Components

Surface texture characterization of components built using additive manufacturing (AM) remains a challenge. The presence of various asperities and random roughness distributions across a surface poses several challenges to users in selecting an appropriate cut-off wavelength (λc), evaluation length (ln), and measurement area. This paper investigates a modified framework for surface texture characterization of AM components. First, the surface asperities in an AM component were identified through scanning electron microscope (SEM) analyses. The maximum diameter (φm) of the surface asperities were determined through image processing and were used as cut-off for surface texture evaluation. Second, another set of surface texture results were extracted using standard measurement procedures per ISO 4287, 4288, 25178-1, -2, and -3. Third, the investigative measurement framework’s effectiveness and suitability were explored by comparing the results with ISO standard results. Last, the effects of using non-standard cut-off wavelength, evaluation length, and measurement area during surface texture characterization were studied, and their percentage deviations from the standard values were discussed. The key findings prove that (a) the evaluation length could be compromised instead of cut-off, (b) measurement area must be 2.5 times the maximum asperity size present in the surface, and (c) it is possible to identify, distinguish, and evaluate specific features from the AM surface by selecting appropriate filters, thereby characterizing them specifically. The investigations and the obtained results serve as valuable data for users to select appropriate measurement settings for surface texture evaluation of AM components.

An effective shape descriptor for content based image retrieval

In this work, we present a modified version of the Generic Fourier Descriptor (GFD) that operates on edge information within natural images from the COREL image database for the purpose of shape-based image retrieval. By incorporating an edge-texture characterization (ETC) measure, we reduced the complexity inherent in oversensitive edge maps typical of most gradient-based detectors that otherwise tend to contaminate the shape feature description. We find that the proposed techniques not only improve overall retrieval in terms of shape, but more importantly, provide a more accurate similarity ranking measure of retrieved results, demonstrating the need for greater consideration for dominant internal and external shape details. A feature database combined by color moments, color histograms, Gabor wavelet and shape features is applied in our image retrieval system. Relevance feedback has also been considered, bridging the gap between the high level concepts and the low level visual features. The experimental results indicate that dynamically updating weights associated with feature compenents by users' feedback greatly improves retrieval performance.

An effective shape descriptor for content based image retrieval

In this work, we present a modified version of the Generic Fourier Descriptor (GFD) that operates on edge information within natural images from the COREL image database for the purpose of shape-based image retrieval. By incorporating an edge-texture characterization (ETC) measure, we reduced the complexity inherent in oversensitive edge maps typical of most gradient-based detectors that otherwise tend to contaminate the shape feature description. We find that the proposed techniques not only improve overall retrieval in terms of shape, but more importantly, provide a more accurate similarity ranking measure of retrieved results, demonstrating the need for greater consideration for dominant internal and external shape details. A feature database combined by color moments, color histograms, Gabor wavelet and shape features is applied in our image retrieval system. Relevance feedback has also been considered, bridging the gap between the high level concepts and the low level visual features. The experimental results indicate that dynamically updating weights associated with feature compenents by users' feedback greatly improves retrieval performance.

Using Macrotexture Map to Visualize Texture Heterogeneity in Polycrystalline Aggregate

A new tool, macrotexture map, was developed to represent and visualize texture heterogeneity in polycrystalline aggregate. This is a critical tool for microstructure representation, useful in risk analysis, performance simulation, and hotspot identification. In contrast to orientation imaging microscope (OIM) map where each color represents a crystal orientation, each color in this macrotexture map represents a texture. Different color represent different texture and similar texture shall have similar color. Macrotexture map provide a unique function to quantitatively evaluate texture heterogeneity of large components, leading to a first-hand understanding of property heterogeneity and anisotropy. For an experienced user, these maps serve the same purpose in identifying high risk locations in the investigated component as medical imaging maps do for diagnosis purpose. This method will also serve as a starting point in mesoscale simulation with meshing sensitivity based on the texture heterogeneity. It will provide a bridge between texture characterization and behavior simulation of components with texture heterogeneity. Macrotexture map will offer a linkage between crystal plasticity simulation in small length scale and finite element/difference simulation in large length scale.