Geometric Feature Extraction Research Articles

Numerical Method for Geometrical Feature Extraction and Identification of Patient-Specific Aorta Models in Pediatric Congenital Heart Disease

An algorithm providing information on the key geometric features of an aorta extracted from multi-slice computed tomography images is proposed. Using the numerical method, the aorta’s geometric characteristics, such as vessel cross-sectional areas and diameters, as well as distances between arteries, can be determined. This step is crucial for training the meta-model necessary to construct an expert system with a significantly reduced volume of data and for identifying key relationships between diagnoses and geometric and hydrodynamic features. This methodology is expected to be part of an innovative decision-making software product for clinical implementation. Based on clinical and anamnestic data as well as calculations, the software will provide the shunt parameters (in particular, its diameter) and installation position to ensure regular blood flow.

Lightweight integration of 3D features to improve 2D image segmentation

Scene understanding has made tremendous progress over the past few years, as data acquisition systems are now providing an increasing amount of data of various modalities (point cloud, depth, RGB...). However, this improvement comes at a large cost on computation resources and data annotation requirements. To analyze geometric information and images jointly, many approaches rely on both a 2D loss and 3D loss, requiring not only 2D per pixel-labels but also 3D per-point labels. However, obtaining a 3D groundtruth is challenging, time-consuming and error-prone. In this paper, we show that image segmentation can benefit from 3D geometric information without requiring a 3D groundtruth, by training the geometric feature extraction and the 2D segmentation network jointly, in an end-to-end fashion, using only the 2D segmentation loss. Our method starts by extracting a map of 3D features directly from a provided point cloud by using a lightweight 3D neural network. The 3D feature map, merged with the RGB image, is then used as an input to a classical image segmentation network. Our method can be applied to many 2D segmentation networks, improving significantly their performance with only a marginal network weight increase and light input dataset requirements, since no 3D groundtruth is required.

Labelled Indoor Point Cloud Dataset for BIM Related Applications

BIM (building information modelling) has gained wider acceptance in the AEC (architecture, engineering, and construction) industry. Conversion from 3D point cloud data to vector BIM data remains a challenging and labour-intensive process, but particularly relevant during various stages of a project lifecycle. While the challenges associated with processing very large 3D point cloud datasets are widely known, there is a pressing need for intelligent geometric feature extraction and reconstruction algorithms for automated point cloud processing. Compared to outdoor scene reconstruction, indoor scenes are challenging since they usually contain high amounts of clutter. This dataset comprises the indoor point cloud obtained by scanning four different rooms (including a hallway): two office workspaces, a workshop, and a laboratory including a water tank. The scanned space is located at the Electrical and Computer Engineering department of the Faculty of Engineering of the University of Porto. The dataset is fully labelled, containing major structural elements like walls, floor, ceiling, windows, and doors, as well as furniture, movable objects, clutter, and scanning noise. The dataset also contains an as-built BIM that can be used as a reference, making it suitable for being used in Scan-to-BIM and Scan-vs-BIM applications. For demonstration purposes, a Scan-vs-BIM change detection application is described, detailing each of the main data processing steps.

A Convolutional Neural Network for Parameter Estimation of the Bi-GTD Model

In this paper, a novel parameter estimation method based on a convolutional neural network (CNN) is proposed to extract geometrical features of radar objects. The CNN’s design is inspired by the inversion process of a physically relevant model, called the geometrical theory of diffraction (GTD) model, whose bistatic form can be used to describe the bistatic scattering response from the target in the netted radar system. This model-inspired inversion method can automatically compensate for phase errors between multiple signal channels and obtain better parameter estimation performance than traditional methods, such as the orthogonal matching pursuit (OMP), the estimation of signal parameters via rotational invariance techniques (ESPRIT) and the multiple signal classification (MUSIC). The experimental results not only verify the validity of the proposed intelligent inversion method but also demonstrate the interpretability and generalization ability of the CNN, whose architecture is designed based on mathematical derivation.

An Improved Machine Learning Algorithm for Silver Nanoparticle Images: A Study on Computational Nano-Materials

Objectives: To study the determination and classification of Scanning Electron Microscopy (SEM) images of silver nanoparticles using digital image processing techniques and classifying the images using various machine learning classifiers namely; SVM, K-NN, and PNN classifiers. Methods: Segmentation techniques namely; Fuzzy C-Means (FCM) and K-Means were applied to extract geometric features from SEM images of silver nanoparticles. The size of nanoparticles was determined and classified based on various nano applications. The categorization of silver nanoparticles was done based on the geometrical feature value, i.e. area, and was further classified into three categories such as; 0-100 nm, 101-200 nm, and 201-500 nm using both the segmentation techniques. Further, the proposed method extracts texture features such as; kurtosis, skewness, and entropy and used SVM, K-NN, and PNN classifiers to classify the silver nanoparticle SEM images. The proposed results are analyzed and interpreted by chemical experts to observe the effectiveness of the proposed method. Findings: To improve the classification accuracy of Silver nanoparticle images, textural features such as; kurtosis, skewness, and entropy were extracted and SVM, K-NN, and PNN classifiers are used to classify the silver nanoparticle images. The performance of the proposed algorithm is evaluated based on performance evaluation measures such as; specificity, sensitivity, true predictive value, and false positive value. Higher TPV, specificity, and sensitivity values show promising results. It is observed that the SVM classifier has 75.00%, K-NN has 62.50%, and PNN yielded an accuracy of 93.00%. Novelty: Traditional synthesis and characterization techniques are inefficient and time-consuming and require more space and infrastructure to analyze the size of the nanoparticle. The current study circumvents these limitations. The silver nanoparticle images used in the present study are synthesized using rudanti fruit seeds, and the process is named as green synthesis process or biological method. This method is widely used because it does not use anytoxic chemicals, is inexpensive, andis environmentally friendly. In the previous work, only geometrical features namely; Area (pore size), and range-wise pore size were calculated depending on the application-oriented. The present study employs two phases; the training phase (knowledge base) and the testing phase(classification). The training phase includes segmentation, feature extraction, and categorization depending on the applications. The testing phase extracts the textural features such as; kurtosis, skewness, and entropy, and classifies the silver nanoparticles by using SVM, K-NN, and PNN classifiers. Keywords: Silver nanoparticles; Image segmentation; FCM; KMeans; SVM; KNN; PNN

Motion Planning of Manipulator by Points-Guided Sampling Network

This paper proposes a network called points-guided sampling net (PGSN) to guide the sampling process in sampling-based motion planner by utilizing the geometric information of obstacles. The geometric information is extracted from the point cloud of obstacles. By analyzing the properties of the point cloud, we propose a VAE feature extraction net that incorporates the variational autoencoder (VAE) framework with unique architectures designed for point clouds. Furthermore, we design a multi-modal sampling net to model the probability distribution of the states based on training trajectories taken from different environments. Based on PGSN, we propose a sampling-based motion planning algorithm called the point-guided rapidly-exploring random tree (PG-RRT). Three experiments are conducted to verify the proposed PGSN: Exp I shows the proposed VAE feature extraction net can successfully extract geometric features from the inputted point cloud; Exp II verifies the multi-modal sampling net successfully chooses corresponding mode with respect to extracted features; Exp III demonstrates the efficacy of our PG-RRT algorithm by showing PG-RRT outperforms other algorithms. Moreover, we provide theoretical analysis and insights towards understanding our model. Note to Practitioners—Obstacles cause lots of the sampling space invalid, thus the traditional sampling-based motion planning (SBMP) algorithm is usually unable to generate a trajectory within a reasonable short period of time. To improve the success rate and efficiency of SBMP, this paper proposes a novel deep neural network called points-guided sampling net (PGSN). PGSN is designed to exploit: (1) environmental point clouds and (2) training trajectories from multiple environments with different obstacles. In the first step, the point clouds include important geometric information. To utilize this information, we adopt a variational autoencoder approach which combines an encoder and a decoder together to extract geometric features more accurately from point clouds. In the second step, trajectories from multiple environments have a multi-modal property which can be represented by a truncated multivariate Gaussian mixture model. We propose a multi-modal sampling net to learn optimal parameters of this model from the training trajectories, and to select corresponding mode based on the extracted features. Experiments demonstrate that the proposed algorithm is feasible and can achieve higher success rate than the state-of-the-art methods. Our method uses a single frame of point cloud to improve efficiency, therefore multiple point clouds from different perspective maybe needed when objects occlude with each other.

Contour Primitive of Interest Extraction Network Based on Dual-Metric One-Shot Learning for Vision Measurement

Although many existing vision measurement systems have achieved high performances, they are object-specific and have limitations in flexibility. Toward intelligent vision measurement that can be conveniently reused for novel objects, this article focuses on the image geometric feature extraction with one-shot learning ability. We propose a contour primitive of interest (CPI) extraction network with dual metric (CPieNet-DM), which can obtain a designated CPI in a query image of a novel object under the guidance of only one annotated support image. First, the dual-metric learning mechanism is proposed, which not only utilizes inter-image similarity as guidance but also leverages the intra-image coherency of CPI pixels to facilitate the inference. Second, a neural network is designed to infer the CPI map based on the dual metric, which also predicts the CPI's geometric parameters. Moreover, the dual context aggregator is plugged in to provide the awareness of both images’ contexts. Third, the network training is jointly supervised by the multiple tasks of dual-metric learning, geometric parameters regression, and CPI extraction. The online hard example mining is utilized to improve the training outcome. The effectiveness of the proposed methods is validated with a series of experiments.

Recognition and Pose Estimation Method for Stacked Sheet Metal Parts

To address issues such as detection failure and the difficulty in locating gripping points caused by the stacked placement of irregular parts in the automated sheet metal production process, a highly robust method for the recognition and pose estimation of parts is proposed. First, a decoding framework for parts of a two-dimensional code is established. The morphological closed operation and topology of contours are used to locate the two-dimensional code, and the type of the part is decoded according to the structure of the two-dimensional code extracted by the projection method. Second, the recognition model of the occluded part type is constructed. The edge information of parts is extracted. The contour convex hull is used to split the part contours, and the similarity of segmented contours is calculated based on the Fourier transform. Finally, the occluded parts are located. The corner points of the metal parts are extracted by the adjacency factor of the differential chain code sequence and the contour radius of curvature. The transformation matrix between the part and the standard template is calculated using similar contour segments and contour corner points. A stereo vision system is built to detect and localize the irregular sheet metal parts for experiments, including detection and information extraction experiments of the two-dimensional laser-generated code and detection and positioning experiments of parts under different occlusion rates. The experimental results show that the decoding framework can accurately decode the two-dimensional code made by a laser under low-contrast conditions, the average recognition rate can reach 93% at multiple occlusion rates, the geometric feature extraction algorithm is more accurate than common algorithms and no pseudo-corner points, the localization error is less than 0.8 mm, and the pose angle error is less than 0.6°. The methods proposed in this paper have high accuracy and robustness.

Geometric prior guided hybrid deep neural network for facial beauty analysis

AbstractFacial beauty analysis is an important topic in human society. It may be used as a guidance for face beautification applications such as cosmetic surgery. Deep neural networks (DNNs) have recently been adopted for facial beauty analysis and have achieved remarkable performance. However, most existing DNN‐based models regard facial beauty analysis as a normal classification task. They ignore important prior knowledge in traditional machine learning models which illustrate the significant contribution of the geometric features in facial beauty analysis. To be specific, landmarks of the whole face and facial organs are introduced to extract geometric features to make the decision. Inspired by this, we introduce a novel dual‐branch network for facial beauty analysis: one branch takes the Swin Transformer as the backbone to model the full face and global patterns, and another branch focuses on the masked facial organs with the residual network to model the local patterns of certain facial parts. Additionally, the designed multi‐scale feature fusion module can further facilitate our network to learn complementary semantic information between the two branches. In model optimisation, we propose a hybrid loss function, where especially geometric regulation is introduced by regressing the facial landmarks and it can force the extracted features to convey facial geometric features. Experiments performed on the SCUT‐FBP5500 dataset and the SCUT‐FBP dataset demonstrate that our model outperforms the state‐of‐the‐art convolutional neural networks models, which proves the effectiveness of the proposed geometric regularisation and dual‐branch structure with the hybrid network. To the best of our knowledge, this is the first study to introduce a Vision Transformer into the facial beauty analysis task.

Image classification using gaussian curvature based local tetra pattern descriptor with enhanced feature selection technique

A Gaussian Curvature-based Local Tetra Descriptor (GCLTrP) is proposed in this paper to incorporate geometric discriminative feature extraction using a hybrid combination of Gaussian Curvature (GC) and Local Terta Pattern (LTrP). The texture of an image is locally discriminant, capturing the equivalent binary response from the gaussian curvature. The extracted feature value is fed into the Enhanced Grey Wolf Optimization (EGWO), a lightweight metaheuristic searching algorithm that selects the best optimal textural features. The proposed GCLTrP with EGWO method’s effective performance is validated using the benchmarks dataset, and the results are tested using the performance evaluation metric. In comparison to other cutting-edge methods, the proposed method achieves the highest overall classification accuracy of 100% on the Brodatz and RS datasets. In terms of computational redundancy and noise reduction, the proposed technique outperforms the other existing techniques.

Extraction of Blood Vessels Geometric Shape Features with Catheter Localization and Geodesic Distance Transform for Right Coronary Artery Detection.

Extraction of Blood Vessels Geometric Shape Features with Catheter Localization and Geodesic Distance Transform for Right Coronary Artery Detection.

Learning efficient facial landmark model for human attractiveness analysis

Existing geometric features on facial attractiveness analysis only focus on the ratios and distances, which is incomplete to represent all the information of a face. In this paper, we introduce a new category of feature, i.e., the angle features, to describe the angle of different organs such as the chin and eyes, which help boost the analysis performance in experiment. In addition, existing facial beauty analysis papers usually apply existing landmark models and extract their own different geometric feature sets on the landmarks. On the one hand, the geometric features are quite chaotic between different papers. On the other hand, most of the landmarks in the existing landmark model are useless for geometric feature extraction which wastes a lot of computational resources. To tackle these issues, we suggest to define a common geometric feature set and learn a special landmark model for attractiveness analysis. Specially, we collect all the available geometric features from the previous jobs and introduce a genetic feature selection algorithm to select the most effective geometric features. Furthermore, we introduce a special landmark model which exactly covers all the extracted geometric features. The experiments show that our method with the introduced angle features and the common feature set can outperform state-of-art facial beauty estimation methods with geometric features.

Rapid seismic risk assessment of bridges using UAV aerial photogrammetry

This paper proposes a framework for rapid seismic risk assessment of bridges using aerial photogrammetric surveys conducted by Unmanned Aerial Vehicles (UAVs). First, the data acquisition process for the photogrammetric 3D reconstruction of an asset and the subsequent procedure for computer-vision-based automatic extraction of visible geometric features are presented. Then, the extracted features are combined in the structural models to perform a seismic risk assessment in terms of capacity-to-demand ratios. Uncertainties are related to the material and geometric hidden variables. Monte Carlo simulation is performed, assuming uniform (uninformed) distributions for the input variables affected by epistemic uncertainty. The ranges of such distributions are estimated based on engineering judgment and knowledge of construction practices. The analysis provides an estimate of the epistemic uncertainties of the capacity-to-demand ratios, with limited available information obtained from UAV aerial photogrammetry. The feasibility and applicability of the proposed framework are demonstrated through a case study of a simply supported bridge in the Italian highway network. The 3D reconstruction process is validated by comparing the aerial survey with a laser scanner survey. The seismic risk of the selected bridge is assessed using the geometric information obtained from the UAV aerial photogrammetric survey. Finally, the paper discusses the applicability of the proposed methodology for bridge management purposes in an automated framework and at network scale.

Appearance design defect detection of rigid plastic products based on geometric feature extraction

Appearance design defect detection of rigid plastic products based on geometric feature extraction

Appearance design defect detection of rigid plastic products based on geometric feature extraction

Appearance design defect detection of rigid plastic products based on geometric feature extraction

DGM-VINS: Visual–Inertial SLAM for Complex Dynamic Environments With Joint Geometry Feature Extraction and Multiple Object Tracking

Most current state-of-the-art simultaneous localization and mapping (SLAM) algorithms perform well in static environments. However, their applications in real-world scenarios are limited by the assumption that environments are static because their performance becomes unstable in complex dynamic environments. To enhance system stability and localization accuracy in complex dynamic scenes, this article presents a novel visual-inertial SLAM system called DGM-VINS. In DGM-VINS, a joint geometric dynamic feature extraction module (JGDFE) is designed that can combine the advantages of multiple geometric constraints and effectively reduce the limitations of a single geometric constraint in the application process. In addition, a temporal instance segmentation module (TISM) is presented to establish the temporal correlation of instance objects in consecutive frames, which effectively addresses the instance segmentation issue in complex environments. The inertial measurement unit (IMU) is utilized for motion prediction and consistency detection to improve localization accuracy in challenging environments with weak textures. The proposed methodology is tested in various public datasets and actual scenarios, and the results demonstrate superior accuracy and robustness than existing methods in complex dynamic scenarios.

Discussion on Related Key Technologies in Distributed Remote Sensing Image Processing

Due to the rapid development of science and technology, remote sensing identification means have been applied in all aspects, especially in the civil and military, and have become an important means to obtain important information. Remote sensing image processing (processing of remote sensing image data) is a series of operations, such as radiation correction and geometric correction, image finishing, projection transformation, inlay, feature extraction, classification and various thematic processing, in order to achieve the desired purpose. Remote sensing image processing can be divided into two categories: one is to use optical, photographic and electronics methods to process remote sensing simulation images (photo, film), referred to as optical processing; the other is to use computer for a series of operations to obtain certain expected results, called remote sensing digital image processing. This paper will focus on the distributed remote sensing image processing method, and further study the possible challenges and key technologies in the practical application, in order to provide theoretical help for the practical work of workers in the same industry.

D3GATTEN: Dense 3D Geometric Features Extraction and Pose Estimation Using Self-Attention

Point-cloud processing for extracting geometric features is difficult due to the highly non-linear rotation variance and measurement noise corrupting the data. To address these challenges, we propose a new architecture, called Dense 3D Geometric Features Extraction And Pose Estimation Using Self-Attention (D3GATTEN), which allows us to extract strong 3D features. Later on these can be used for point-cloud registration, object reconstruction, pose estimation, and tracking. The key contribution of our work is a new architecture that makes use of the self-attention module to extract powerful features. Thoughtful tests were performed on the 3DMatch dataset for point-cloud registration and on TUM RGB-D dataset for pose estimation achieving 98% Feature Matching Recall (FMR). Our results outperformed the existing state-of-the-art in terms of robustness specification for point-cloud alignment and pose estimation. Our code and test data can be accessed at link: <uri xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">https://github.com/tamaslevente/trai/tree/master/d3gatten</uri> .

Image Processing and Feature Extraction for Hull Structure GMAW Based on Weld Pool Visual Sensing

Image processing and feature information extraction based on the visual perception of the weld pool are considered essential components of intelligent welding quality monitoring of hull structure gas metal arc welding (GMAW). The unstable characteristics, such as large spatter, much smoke, and strong arc light during hull structure GMAW, lead to the blurring of image acquisition and the difficulty of contour extraction of the weld pool. The present study is aimed at addressing the practical issues from two perspectives, i.e., a spectrum‐visual‐sensing acquisition system and an image‐processing and feature extraction algorithm. First of all, by analyzing the light energy distribution law and acquiring the optical parameters relevant to the cut‐off composite dimming and near‐infrared narrowband filtering, spectral sensing is employed in establishing models of arc light radiation to detect the strength of continuously distinctive spectral lines. Besides, an appropriate high‐speed charge‐coupled device (CCD) camera is selected to build a visual acquisition system, which can reduce the external interference of the arc light on the image acquisition of the weld pool. Afterwards, the implementation of an image‐processing fusion model based on the spatial information fuzzy C‐means (FCM) clustering analysis and Sobel edge detection operator accompanies the investigation of the geometric aspects of the weld pool image. In terms of clear segmentation of the interest region, the edge detection and accurate extraction of the target contour are successfully obtained. In the subsequent section, the Hough transform analysis is adopted to establish the geometric feature extraction model of the weld pool, with corner detection, conversion, and camera calibration as the core technology. Additionally, the left and right views of the image contour are calibrated to achieve the lossless conversion of corner pixels and physical coordinates. Finally, three other distinct image‐processing methods are designed to compare the segmentation effect of the edge contour with the fusion model, and then, the extraction accuracy of the geometric features of the weld pool is verified. The interference of the arc light and smoke has been demonstrated to be substantially diminished, which is attributable to the visual‐sensing system during image acquisition of the weld pool. The results of edge fusion of the weld pool image show that based on the GMAW using the FCM‐Sobel fusion method, a superior extraction accuracy of geometric features characterized by smoothness, continuousness, no breakpoints, and less noise has fulfilled the engineering requirements.

Efficient and Lightweight Neuron Morphology Classification Using Gabor Convolutional Networks

Due to the complex and diverse morphology of neurons, achieving accurate classification of neurons is not an easy task. Most existing methods for neuron classification are based on traditional machine learning algorithms, which not only have high design costs but also poor transferability. Gabor Orientation Filters (GoFs), obtained by modulating Gabor filters and traditional convolutional kernels, can better extract geometric features of neurons and enhance the adaptability of deep learning features to GoFs' orientation and scale changes. Therefore, this paper proposes an efficient and lightweight Gabor Convolutional Network (GCN) for neuron morphology classification based on the geometric shape data of neurons and the spatial transformation capabilities of Gabor filters. The network consists of five consecutive Gabor convolutional layers for extracting geometric features of neurons and two linear layers for neuron morphology classification. Experimental results show that this method achieves accuracy rates of 94.26%, 87.16%, and 87.72% for binary classification and 96.52%, 85.96%, and 86.73% for 12-class classification on the Img_raw, Img_resample, and Img_XYalign neuron morphology datasets, respectively. Meanwhile, compared with classical classification models, the experimental results show that this method can capture enough neuron morphology features with fewer convolutional layers, and it has good classification accuracy and fewer model parameters.