Contour estimation using global shape constraints and local forces

Excessive settlement or differential settlement of subgrade will lead to the deterioration of line operational conditions, the reduction of passenger comfort, and even endanger the safety of traffic. Therefore, it is of great significance to study the settlement prediction of subgrade. In order to predict the settlement of foundation under the next level of loading earlier during the embankment construction process, a new method of predicting settlement of soft soil subgrade is proposed. Firstly, based on monitoring results of soft soil foundation, the consolidation parameters of soil layer are back-calculated according to the three-point method. Then, combined with the theory of the consolidation degree of graded loading, the formula that can predict settlement under different loading conditions are derived. Eventually, the practical application of the method is verified by the prediction and comparative analysis of measured settlements based on engineering examples. The result of research shows that the method can predict the foundation settlement after loading during construction of engineering fill. This method has obvious advantages over the traditional curve fitting method and can guide the actual engineering construction.

Improving the quality of water intake monitoring data is an urgent issue in current water management. The industrial water intake monitoring data obtained during the National Water Resources Monitoring Capacity Building Project promotion project was taken as a sample, and the common abnormal categories of water intake monitoring data were summarized, and the strategy of "rough screening-fine identification-reconstruction" was proposed. Considering the seasonal fluctuation law of water monitoring data, the multiscale industrial water monitoring abnormal data identification models were constructed based on segmented 3σ criterion, wavelet transform, and Fourier function. Moreover, the least squares support vector machine (LSSVM) model with adaptive inertia function and particle swarm optimization (PSO) was used to reconstruct the recovered anomaly data. The results indicate that the segmented 3σ criterion performs well for the rough processing of water intake monitoring data, identifying 26 data points that fall outside the corresponding threshold intervals. The Fourier function can effectively reduce the information loss associated with the wavelet transform, thereby improving the accuracy of abnormal data identification; based on verification feedback from monitoring users, 31 of the 38 detected abnormal points were confirmed as "demand-driven anomalies," yielding an identification accuracy of 81.6%. Furthermore, the inertia function-particle swarm optimization LSSVM model meets the high-precision requirements for abnormal data reconstruction and recovery, and its reconstruction accuracy is higher than that of the LSSVM, the PSO-LSSVM, and the traditional curve fitting method. Specifically, the inertia function-particle swarm optimization LSSVM achieves an average fitting error of 0.0286, representing reductions of 46.2% and 44.4% compared with the LSSVM (0.0532) and PSO-LSSVM (0.0514), respectively; moreover, when compared with the ground-truth values obtained from verification feedback, the reconstruction error rate is below 5%. Overall, the proposed multiscale mining and reconstruction strategy for industrial water intake monitoring abnormal data can provide a valuable methodological reference for enhancing the decision support capability of data in the National Water Resources Monitoring Capacity Building Project.

In this paper, we investigate the descriptive capabilities of a series of well known shape descriptors, using classical Multi-Dimensional Scaling (MDS). Sammon Mapping is applied to a data-set of six hundred shapes from a real-world multi-component trademark data-set, that have been described using traditional perceptual descriptors, Fourier descriptors and Rosin’s descriptors. The maps generated offer considerable insight into how the descriptors encode shape, there is evidence to suggest that traditional perceptual descriptors offer better discriminating performance than Fourier and Rosin’s, and also strong indications that an intelligently chosen combination of descriptors could offer increased discriminating capacity over any single descriptor.The mapping techniques discussed in this paper use an arbitrary distance measure, and hence can be easily adapted to map many other forms of image descriptors. This makes them extremely useful for evaluating descriptor performance and also for visual browsing of image databases.

Shape descriptors have demonstrated encouraging potential in retrieving images based on image content. A number of shape descriptors have been reported in the literature. Nevertheless, most of the reported descriptors still face accuracy and computational challenges. Fourier descriptors are considered to be promising descriptors as they are based on sound theoretical foundation, and possess computational efficiency and attractive invariance properties. In this paper, we propose a novel Fourier descriptor based on contour curvature. The proposed descriptor takes an unconventional view of the curvature-scale-space representation of a shape contour as it treats it as a 2-D binary image (hence referred to as Curvature-Scale Image, or CSI). The invariant descriptor is derived from the 2-D Fourier transform of the Curvature-Scale Image. This allows the descriptor to capture detailed dynamics of the shape curvature and enhance the efficiency of the shape matching process. Experiments using images from the MPEG-7 database have been conducted to compare the performance of the proposed descriptor with the Curvature- Scale-Space Descriptor (CSSD), the Generic Fourier Descriptor (GFD), and the 1-D Fourier Descriptor (1-FD). The proposed descriptor demonstrated superior performance.

The Fourier descriptors (FDs) is a classical but still popular method for contour matching. The key idea is to apply the Fourier transform to a periodic representation of the contour, which results in a shape descriptor in the frequency domain. FDs are most commonly used to compare object silhouettes and object contours; the authors instead use this well-established machinery to describe local regions to be used in an object-recognition framework. Many approaches to matching FDs are based on the magnitude of each FD component, thus ignoring the information contained in the phase. Keeping the phase information requires us to take into account the global rotation of the contour and shifting of the contour samples. The authors show that the sum-of-squared differences of FDs can be computed without explicitly de-rotating the contours. The authors compare correlation-based matching against affine-invariant Fourier descriptors (AFDs) and WARP-matched FDs and demonstrate that correlation-based approach outperforms AFDs and WARP on real data. As a practical application the authors demonstrate the proposed correlation-based matching on a road sign recognition task.

Recently, rapid repair of damaged blade has become the focus of considerable interest for extending its service life. However, due to the defects caused by high temperature and pressure of operations as well as foreign object impact, the turbine blades often undergo the deviations of the actual part profile from its design model, such that this nominal Computer Aided Design (CAD) model cannot be directly used in the process of repair for tool path generation of laser cladding and Numerical Control (NC) machining, thus to nicely repair the damaged or worn blades, it is necessary to reconstruct the surface model of the actual blade. This paper develops a deformable template-based approach to recovering the surface of blade from the cross-sectional profiles. The mathematical model for cross-sectional profile reconstruction is first established and is then solved by an alternate iteration optimization strategy consisting of registration and deformation of the template curve. Since the proposed method can automatically transform and deform the template curve to best fit the cross-sectional points, the compatibility conditions between different sections are automatically satisfied and there is no need for the data preprocessing such as data sorting, parameterization, etc. which are necessary for the traditional surface fitting methods. Undoubtedly, this considerably simplifies the reconstruction problem of the damaged blade and nicely adapts to blade part-to-part variation. Moreover, a method of closest point computation that combines the arithmetic for Bernstein-form polynomials and Bézier curve subdivision is also given based on bintree decomposition to improve the iteration processes of 2D profile reconstruction. Then, according to these reconstructed sectional profiles, the actual blade surface is reconstructed by surface skinning operations. Finally, the proposed method is tested on a sample blade, and the experimental results show that our method can precisely reconstruct the surface of the damaged blade, especially for the blades with area defects.

Fourier descriptors (FDs) is a classical but still popular method for contour matching. The key idea is to apply the Fourier transform to a periodic representation of the contour, which results in a shape descriptor in the frequency domain. Fourier descriptors have mostly been used to compare object silhouettes and object contours; we instead use this well established machinery to describe local regions to be used in an object recognition framework. We extract local regions using the Maximally Stable Extremal Regions (MSER) detector and represent the external contour by FDs. Many approaches to matching FDs are based on the magnitude of each FD component, thus ignoring the information contained in the phase. Keeping the phase information requires us to take into account the global rotation of the contour and shifting of the contour samples. We show that the sum-of-squared differences of FDs can be computed without explicitly de-rotating the contours. We compare our correlation based matching against affine-invariant Fourier descriptors (AFDs) and demonstrate that our correlation based approach outperforms AFDs on real world data.

In this paper, we present a novel set of features which mainly comprises Fourier descriptors to facilitate the classification of white matter (WM) fiber tracts reconstructed from diffusion tensor imaging (DTI) tractography. Particularly, Fourier descriptors composing the feature vectors are different frequency terms of the discrete Fourier transform of the Euclidean coordinates of the step points of WM fiber tracts. This approach can effectively reduce the computing cost in WM fiber tract classification due to the decreased feature vector dimensionality. As integrals, Fourier descriptor features make the algorithm less sensitive to DTI noise. The fact that Fourier descriptors achieve spatial independent representation and normalization of WM fiber tracts makes it useful for group studies of WM fiber tracts. Specially, a new strategy for using Fourier descriptors is proposed. Experimental results with real datasets show that this technique is feasible and may be an alternative to existing approaches.

Segmentation followed by shape descriptors represents a common and fundamental approach used in many image processing systems. Active contour models have been used for shape description as a promising method. In particular, the deformable template, which is a kind of active contour model, has been used for various shape description problems. But active contour models, including the deformable template model, suffer from some common difficulties. We propose new approaches in which we represent an object using a weighted Gaussian approximation to find the best candidate template and minimize an appropriately designed cost function to deform the template after finding a best fit candidate in the multiscale representation of the image. This framework can be applied to many real-time applications such as object based video coding, and the estimation of facial features in face recognition.

Automated determination of poplar chip size distribution based on combined image and multivariate analyses

Chenopodium oahuense is a polymorphic Hawaiian endemic plant inhabiting several xeric habitats. Considerable variability in leaf morphology has made comprehensively describing its diversity difficult. Chenopodium oahuense subsp. ilioense is differentiated from C. oahuense subsp. oahuense by smaller, less-lobed, succulent leaves, smaller seeds, and prostrate to scandent habit. The lacking quantification of leaf shape, succulence, and previously unknown heteroblastic leaf transition in C. oahuense subsp. ilioense complicates the morphological boundaries separating subspecies. This study used landmark analyses, elliptical Fourier descriptors (EFDs), and traditional shape descriptors measured from 1585 greenhouse-grown plant leaves collected over 18 weeks. Principal component analyses visualized correlations in leaf shape, and linear discriminant analyses predicted classifications, either subspecific or heteroblastic. Identity determination and heteroblastic change visualization were limited in landmark analyses. On the basis of EFDs and shape descriptors, C. oahuense subsp. ilioense was determined to be morphologically differentiated from C. oahuense subsp. oahuense with the Pu'u Ka Pele population as intermediate. The EFDs depicted heteroblastic change, predominantly in lobing. All analyses were restricted in correctly attributing a leaf to the week collected. Shape descriptors generally represented significant heteroblastic change over the growth period. These analyses support significant differentiation between the subspecies, particularly from shape descriptors. Furthermore, we quantified the morphological intermediacy of the Pu'u Ka Pele population. Results suggest this population could be the result of incomplete lineage sorting or a recent hybridization of the two subspecies. Hawaiian Chenopodium is a polymorphic lineage notable for future research in adaptive radiations, phenotypic plasticity, and heteroblasty.

Decomposing Strain Maps Using Fourier-Zernike Shape Descriptors

There exist many methods for image digital description of objects as one of the important steps before image classification for object recognition. Examples of such methods are momentum expressions based, Fourier descriptors (DFT), geometric features, structural approaches, statistical approaches ? PCA, SVD and LDA, MPEG-7 for shape description, … and many other known methods based on neural networks, genetic algorithms, hidden Markov chains and so on. These descriptors represent the image code by introducing the image characteristics through what is called the feature vector. Biometrics images are examples of rather specific objects, the human biometric features, and require special care in selecting their descriptor. In the talk the speaker will consider some easy to compute feature vectors based on the modified Circulant Toeplitz matrix minimal eigenvalues. Theory and practice will be considered with biometrics images as examples.

Application of artificial intelligence to forecast hydrocarbon production from shales

Selective Laser Melting (SLM) is a powder-bed additive manufacturing technique whose part quality depends critically on feedstock morphology. However, conventional powder characterization methods are low-throughput and qualitative, failing to capture the heterogeneity of industrial-scale batches. We present an automated, machine learning framework that couples high-throughput imaging with shape extraction and clustering to profile metallic powder morphology at scale. We evaluate three complementary approaches. The first uses an orientation-invariant variational autoencoder (VAE) to learn shape features that are insensitive to particle rotation, after which the resulting latent representations are grouped using clustering methods. The second applies classic shape descriptors – centroid distance functions, Fourier descriptors, and Zernike moments – followed by k-means or Gaussian mixture clustering to group particles with similar outlines. The third treats each particle’s radial profile as a continuous function and clusters these functional representations using the GPmix algorithm, allowing subtle differences in lobing and eccentricity to be captured. Across a dataset of approximately 126,000 powder images (0.5 – 102 μ m diameter), internal validity metrics identify the Fourier-descriptor + k-means pipeline as the most effective, achieving the lowest Davies–Bouldin index and highest Calinski–Harabasz score while maintaining sub-millisecond runtime per particle on a standard desktop workstation. The VAE provides orientation-invariant representations but at significantly higher computational cost, whereas the functional-data approach captures finer shape variations but requires a more complex configuration. Although the present work focuses on establishing the morphological-clustering framework, the resulting shape groups form a basis for future studies examining their relationship to flowability, packing density, and SLM part quality. Overall, this unsupervised learning framework enables rapid, automated assessment of powder morphology and supports tracking of shape evolution across reuse cycles, offering a path toward real-time feedstock monitoring in SLM workflows. • Automated machine learning framework for large-scale Selective Laser Melting (SLM) powder morphology profiling. • Evaluated three unsupervised pipelines: using variational autoencoder, shape descriptors, and functional clustering. • Fourier-descriptor + K-means achieved best accuracy, speed, and efficiency. • Framework distinguishes spherical, lobed, elongated, and defective particles. • Enables real-time monitoring of powder quality across reuse cycles in SLM.