Registration of 3D spatial data and models is a fundamental task in applications such as mapping, positioning and virtual/augmented reality. Most of the existing 3D registration methods such as ...iterative closest point (ICP) and recent learning‐based methods are dedicated to point cloud registration, and rely heavily on point‐wise correspondences, which limits their ability to address registration problems across different data types. Since man‐made objects and buildings usually contain many planar surfaces, it is possible to use the planes for accurate registration of different data and models. In this paper, a unified registration framework is proposed consisting of a plane extraction module, which can extract planes from various forms of spatial data such as point clouds or surface‐based 3D models, and a registration module, which performs automatic registration based on the extracted planes. Tests show that the proposed method can handle small‐overlap registration across all these data types with high success rates. The result of point cloud registration also indicates that the method achieves better accuracy as compared to ICP.
Résumé
Le recalage de données et de modèles spatiaux 3D est une opération fondamentale dans des applications comme la cartographie, le positionnement et la réalité virtuelle/augmentée. La plupart des méthodes de recalage 3D existantes, telles que la méthode du point le plus proche itéré (ICP) et les méthodes récentes basées sur l'apprentissage, sont dédiées au recalage de nuages de points et reposent essentiellement sur des correspondances ponctuelles, ce qui limite leur capacité à résoudre les problèmes de recalage pour des données de types différents. Étant donné que les objets artificiels et les bâtiments contiennent généralement de nombreuses surfaces planes, il est possible d'utiliser les plans pour un recalage précis de différents modèles et données. Cet article propose une méthode unifiée de recalage, composée d'un module d'extraction de plans, qui peut extraire des plans de différentes formes de données spatiales telles que les nuages de points ou les modèles surfaciques 3D, et d'un module de recalage, qui effectue un recalage automatique basé sur les plans extraits. Les tests montrent que la méthode proposée peut effectuer le recalage en présence de petits chevauchements sur tous ces types de données avec des taux de succès élevés. Le résultat du recalage des nuages de points indique également que cette méthode est plus précise que la méthode ICP.
Zusammenfassung
Die Registrierung von räumlichen 3D‐Daten und ‐Modellen ist eine grundlegende Aufgabe in Anwendungen wie Kartierung, Positionierung und virtuelle/erweiterte Realität. Die meisten der bestehenden 3D‐Registrierungsmethoden wie Iterative Closest Point (ICP) und neuere lernbasierte Methoden sind der Punktwolkenregistrierung gewidmet und beruhen stark auf punktweisen Korrespondenzen, was ihre Fähigkeit einschränkt, Registrierungsprobleme über verschiedene Datentypen hinweg anzugehen. Da von Menschenhand geschaffene Objekte und Gebäude normalerweise viele planare Oberflächen enthalten, ist es möglich, die Ebenen für eine genaue Registrierung verschiedener Daten und Modelle zu verwenden. In diesem Artikel wird ein vereinheitlichtes Registrierungs‐Framework vorgeschlagen, das aus einem Ebenenextraktionsmodul besteht, das Ebenen aus verschiedenen Formen von räumlichen Daten wie Punktwolken oder oberflächenbasierten 3D‐Modellen extrahieren kann, und einem Registrierungsmodul, das eine automatische Registrierung auf der Grundlage der extrahierte Flugzeuge. Tests zeigen, dass das vorgeschlagene Verfahren eine Registrierung mit geringer Überlappung über alle diese Datentypen hinweg mit hohen Erfolgsraten handhaben kann. Das Ergebnis der Punktwolkenregistrierung zeigt auch, dass das Verfahren im Vergleich zu ICP eine bessere Genauigkeit erreicht.
Resumen
La captura y generación de datos y modelos espaciales 3D es una tarea fundamental en aplicaciones de cartografía, posicionamiento y realidad virtual o aumentada. La mayoría de los métodos de registro 3D existentes, como el punto más cercano iterativo (ICP) y los métodos recientes basados en el aprendizaje, están orientados al registro de nubes de puntos y en gran medida dependen de las correspondencias de puntos, lo que limita su capacidad para abordar los problemas de registro en tipos de datos diferentes. Dado que los objetos y edificios hechos por el hombre suelen contener muchas superficies planas, es posible usar los planos para el registro preciso de datos diversos y modelos. En este documento, se propone un método de registro unificado que consta de un módulo de extracción de planos, que puede extraer planos de diversos tipos de datos espaciales, como nubes de puntos o modelos 3D basados en superficies, y un módulo de registro, que realiza un registro automático basado en los planos extraídos. Las pruebas muestran que el método propuesto puede manejar el registro de todos los tipos de datos a pesar de tener escasa superposición con tasas de éxito elevadas. El resultado del registro de nubes de puntos también indica que el presente método alcanza mayor precisión en comparación con ICP.
摘要
三维空间信息配准在诸如高精地图,定位以及VR/AR等领域有广泛应用。多数现有的3D配准方法如ICP和近期的基于机器学习的配准算法均只能完成点云的配准,且此类算法多依赖于点对点相关性,因此很难拓展到除点云外的其他形式的空间信息的配准。在包含大量人造物体或建筑的应用场景中通常存在很多平面结构,这些平面信息可以用于将存储于不同格式的空间信息和模型进行配准。本文提出的统一配准架构包含两个模块: 基于从点云和多边形3D模型等不同格式中提取平面信息的平面提取模块和自动将平面信息配准的配准模块。实验结果显示此配准方法可以实现不同格式数据间的低重合度配准并获得高配准成功率。在点云配准测试中的结果表明此方法可以得到高于ICP的准确度。
Plasma electrolytic polishing (PeP) is mainly used to improve the surface quality and thus the performance of electrically conductive parts. It is usually used as an anodic process, i.e., the ...workpiece is positively charged. However, the process is susceptible to high current peaks during the formation of the vapour–gaseous envelope, especially when polishing workpieces with a large surface area. In this study, the influence of the anode immersion speed on the current peaks and the average power during the initialisation of the PeP process is investigated for an anode the size of a microreactor mould insert. Through systematic experimentation and analysis, this work provides insights into the control of the initialisation process by modulating the anode immersion speed. The results clarify the relationship between immersion speed, peak current, and average power and provide a novel approach to improve process efficiency in PeP. The highest peak current and average power occur when the electrolyte splashes over the top of the anode and not, as expected, when the anode touches the electrolyte. By immersion of the anode while the voltage is applied to the anode and counterelectrode, the reduction of both parameters is over 80%.
Vast majority of already-installed grid-connected wind turbines is fixed-speed wind generators (FSWGs). Their initialisations for power-system dynamic simulations are mostly performed using ...conventional power-flow approach, causing an unavoidable reactive-power discrepancy between power-flow result and generator dynamic model. To eliminate this problem, unified Newton–Raphson (NR) power-flow approach together with wind-turbine aerodynamic power coefficient is applied for accurate initialisation of the FSWG under passive-stall regulation. However, FSWG initialisations relating to two other active-stall and active-pitch regulations under the same approach have not yet been addressed. Therefore, in this study, the unified power-flow approach is straightforwardly extended to yield efficient and precise dynamic initialisations of active-stall- and active-pitch-regulated FSWGs. In the proposed model, variables of generators and wind turbines are fully solved all together with the power coefficients. The obtained power-flow result is compared with the published result for model verification. Efficiency of the extended algorithm is demonstrated using IEEE-300bus and 118bus networks with large groups of wind power plants. The simulation results reveal that the extended algorithm not only gives precise steady-state initialisation of the FSWGs but also retains NR quadratic convergence characteristics.
A situation where the set of initial solutions lies near the position of the true optimality (most favourable or desirable solution) by chance can increase the probability of finding the true ...optimality and significantly reduce the search efforts. In optimisation problems, the location of the global optimum solution is unknown a priori, and initialisation is a stochastic process. In addition, the population size is equally important; if there are problems with high dimensions, a small population size may lie sparsely in unpromising regions, and may return suboptimal solutions with bias. In addition, the different distributions used as position vectors for the initial population may have different sampling emphasis; hence, different degrees of diversity. The initialisation control parameters of population-based metaheuristic algorithms play a significant role in improving the performance of the algorithms. Researchers have identified this significance, and they have put much effort into finding various distribution schemes that will enhance the diversity of the initial populations of the algorithms, and obtain the correct balance of the population size and number of iterations which will guarantee optimal solutions for a given problem set. Despite the affirmation of the role initialisation plays, to our knowledge few studies or surveys have been conducted on this subject area. Therefore, this paper presents a comprehensive survey of different initialisation schemes to improve the quality of solutions obtained by most metaheuristic optimisers for a given problem set. Popular schemes used to improve the diversity of the population can be categorised into random numbers, quasirandom sequences, chaos theory, probability distributions, hybrids of other heuristic or metaheuristic algorithms, Lévy, and others. We discuss the different levels of success of these schemes and identify their limitations. Similarly, we identify gaps and present useful insights for future research directions. Finally, we present a comparison of the effect of population size, the maximum number of iterations, and ten (10) different initialisation methods on the performance of three (3) population-based metaheuristic optimizers: bat algorithm (BA), Grey Wolf Optimizer (GWO), and butterfly optimization algorithm (BOA).
The K-means algorithm gets widely used due to its simplicity and effectiveness. But it is sensitive to the selection of initial cluster centres. In this paper, we proposed an initialisation method to ...select initial clustering centres for K-means algorithm. Furthermore, because of the boundedness of the initialisation method, we modified it and designed a Tissue-like P system to realise the new method. The experiments are operated on five UCI datasets and the results proved that the new initialisation method based on the designed Tissue-like P system is effective.
Background:
Digital Image Correlation (DIC) is based on the matching, between reference and deformed state images, of features contained in patterns that are deposited on test sample surfaces. These ...features are often suitable for a single scale, and there is a current lack of multiscale patterns capable of providing reliable displacement measurements over a wide range of scales.
Objective:
Here, we aim to demonstrate that a pattern based on a fractal (self-affine) surface would make a suitable pattern for multiscale DIC.
Methods:
A method to numerically generate patterns directly from a desired auto-correlation function is introduced. It is then enhanced by a Mean Intensity Gradient (MIG) improvement process based on grey level redistribution. Numerical experiments at multiple scales are performed for two different imposed displacement fields and results for one of the patterns generated are compared with those obtained for a random pattern and a Perlin noise one.
Results:
The proposed pattern is shown to lead to DIC errors comparable to those found with the two others for the first scales, but has much greater robustness. More importantly, the pattern generated here exhibits stable errors and robustness with respect to the scale whereas these two outputs become significantly degraded for the other two patterns as the scale increases.
Conclusions:
As a result, scale invariance properties of the pattern based on fractal surfaces correspond to scale invariance in DIC errors as well. This is of great interest regarding the use of such patterns in multiscale DIC.
Face alignment has made great progress in recent years and the cascade regression framework is one of the main contributors. However, the performance of this framework is unsatisfactory on heavily ...occluded faces or those far from the frontal pose. This is because regression is sensitive to hidden landmarks and unified initialisation can often lead to the method falling into local minima. The authors propose a new pipeline of salient-to-inner-to-all to progressively compute the locations of landmarks. Additionally, a feedback process is utilised to improve the robustness of regression. They bring out a pose-invariant shape retrieval method to generate the discriminative initialisation. Experiments are performed on two benchmarks, and the experimental results demonstrate that the proposed method has a considerable improvement on the cascade regression model, and achieves favourable results compared with the state-of-the-art deep learning-based methods.
This study presents a geometric deformable model-based segmentation approach to segmentation of the intima and media-adventitial (MA) borders in sequential intravascular ultrasound (IVUS) images. The ...initial estimation of the vessel borders was done manually only for the first frame of each sequence. After the border initialisation, pre-processing including edge preservation, noise reduction, and dead zone preservation was successively performed on each IVUS frame. To improve segmentation performance, the image masks were determined preliminarily by local binary pattern-based mask initialisation. Then, the inner and outer borders were approximated using a modified distance regularised level set evolution model. The results showed superior performance of the suggested approach for estimating intima and MA layers from the IVUS images. The corresponding correlation coefficients of area, vessel perimeter, maximum vessel diameter, and maximum lumen diameter were r = 0.782, r = 0.716, r = 0.956, and r = 0.874 for the 20 MHz images, respectively, and r = 0.990, r = 0.995, r = 0.989, and r = 0.996 for the 45 MHz images, respectively. In addition, linear regression analysis indicated that the manual segmentation had significantly high similarity at r > 0.967 and r > 0.993 for 20 and 45 MHz images, respectively.
There exists a need to generate well‐designed software systems because of the extensive adoption of object‐oriented programming in software growth. Thus, the total software maintenance cost is ...decreased and the component's reusability is augmented. However, the software system's internal structure worsens owing to extended maintenance activities. For enhancing the system's overall internal structure without varying its external behaviour, restructuring is an extensively utilised solution in this circumstance. Thus, utilising the Deterministic Initialisation Method based K‐Means (DIM‐K‐Means) and Median Absolute Deviation‐based Elastic Net Regulariser Neural Network (MAD‐ENRNN) classifier, a framework called Distributed Object‐Oriented (DOO) software restructuring model is created by the study. Five steps are undertaken by the developed framework. Centred on source code along with change history, the interactions amongst the classes are initially pre‐processed where the dependencies of disparate classes are detected and formulated into a graphical structure. After that, from the graph, the extraction of significant features is done. Utilising a multi variant objective‐based Aquila optimiser, the most pertinent features are selected as of the extracted features. Next, for minimising the complexity, the selected features are created into clusters. Then, the formed clusters are offered to the classifier named MAD‐ENRNN. The DOO software is effectively restructured by MAD‐ENRNN. The proposed methodology's performance is contrasted with the prevailing systems in an experimental evaluation. The outcomes displayed that the proposed framework is capable of restructuring the DOO software with improved accuracy of 9.94% when analogised to the top‐notch methods.
Solving Mismatch problem to software and hardware.