Shape Database Research Articles

Sketch Augmentation-Driven Shape Retrieval Learning Framework Based on Convolutional Neural Networks.

In this article, we present a deep learning approach to sketch-based shape retrieval that incorporates a few novel techniques to improve the quality of the retrieval results. First, to address the problem of scarcity of training sketch data, we present a sketch augmentation method that more closely mimics human sketches compared to simple image transformation. Our method generates more sketches from the existing training data by (i) removing a stroke, (ii) adjusting a stroke, and (iii) rotating the sketch. As such, we generate a large number of sketch samples for training our neural network. Second, we obtain the 2D renderings of each 3D model in the shape database by determining the view positions that best depict the 3D shape: i.e., avoiding self-occlusion, showing the most salient features, and following how a human would normally sketch the model. We use a convolutional neural network (CNN) to learn the best viewing positions of each 3D model and generates their 2D images for the next step. Third, our method uses a cross-domain learning strategy based on two Siamese CNNs that pair up sketches and the 2D shape images. A joint Bayesian measure is used to measure the output similarity from these CNNs to maximize inter-class similarity and minimize intra-class similarity. Extensive experiments show that our proposed approach comprehensively outperforms many existing state-of-the-art methods.

New Set of Non-separable Orthogonal Invariant Moments for Image Recognition

It is known that the rotation, scaling and translation invariant property of image moments has a high significance in image recognition. For this reason, the seven invariant moments presented by Hu are widely used in the field of image analysis. These moments are of finite order; therefore, they do not comprise a complete set of image descriptors. For this reason, we introduce in this paper another series of invariant moments of infinite order, which are based on normalized central moments. The non-orthogonal property of these moments causes the redundancy of information. To overcome this problem, we propose a new construction technique of non-separable orthogonal polynomials in two variables based on a recurrence formula and we present a new set of orthogonal moments, which are invariant to translation, scaling and rotation. The presented approaches are tested in several well-known computer vision datasets including moment’s invariability, image retrieval and classification of objects, this latter based on fuzzy K-means clustering algorithm. The performance of these invariant moments for classification and image retrieval is compared with some recent invariant moments such as invariants of multi-channel orthogonal radial-substituted Chebyshev moments, invariants of quaternion radial-substituted Chebyshev moments, invariants of rotational moments in Radon space and Legendre–Fourier moments in Radon space. The experimental results made using four databases of images, namely Columbia Object Image Library (COIL-20) database, MPEG7-CE shape database, COIL-100 database and ORL database, show that our orthogonal invariant moments have done better than the other descriptors tested.

Rotation scaling and translation invariants by a remediation of Hu’s invariant moments

Due to the invariance to translation, rotation and scaling, the seven invariant moments presented by Hu (Visual pattern recognition by moment invariants, IRE Transactions on Information Theory, vol. 8, February 1962, pp. 179–187) are widely used in the field of pattern recognition. The set of these moments is finite; therefore, they do not comprise a complete set of image descriptors. To solve this problem, we introduce in this paper a new set of invariant moments of infinite order. The non-orthogonal property causes the redundancy of information. For this reason, we propose a new set of orthogonal polynomials in two variables, and we present a set of orthogonal moments, which are invariant to rotation, scale and translation. The presented approaches are tested by the invariability of the moments, the image retrieval and the classification of the objects. In this framework, using the proposed orthogonal moments, we present two classification systems. The first based on the Fuzzy C-Means Clustering algorithm (FCM) and the second based on the Radial Basis Functions Neural Network (RBF). The performance of our invariant moments is compared with Legendre invariant moments, Tchebichef-Krawtchouk (TKIM), Tchebichef-Hahn (THIM), Krawtchouk-Hahn (KHIM), Hu invariant moments, the descriptor of histogram of oriented gradients (HOG), the adaptive hierarchical density histogram features (AHDH) and with descriptors of color and texture Hist, HSV, FOS and SGLD. The experimental tests are performed on seven image databases: Columbia Object Image Library (COIL-20) database, MPEG7-CE shape database, MNIST handwritten digit database, MNIST fashion image database, ImageNet database, COIL-100 database and ORL database. The obtained results show the efficiency and superiority of our orthogonal invariant moments.

Relational data paradigms: What do we learn by taking the materiality of databases seriously?

Although databases have been well-defined and thoroughly discussed in the computer science literature, the actual users of databases often have varying definitions and expectations of this essential computational infrastructure. Systems administrators and computer science textbooks may expect databases to be instantiated in a small number of technologies (e.g., relational or graph-based database management systems), but there are numerous examples of databases in non-conventional or unexpected technologies, such as spreadsheets or other assemblages of files linked through code. Consequently, we ask: How do the materialities of non-conventional databases differ from or align with the materialities of conventional relational systems? What properties of the database do the creators of these artifacts invoke in their rhetoric describing these systems—or in the data models underlying these digital objects? To answer these questions, we conducted a close analysis of four non-conventional scientific databases. By examining the materialities of information representation in each case, we show how scholarly communication regimes shape database materialities— and how information organization paradigms shape scholarly communication. These cases show abandonment of certain constraints of relational database construction alongside maintenance of some key relational data organization strategies. We discuss the implications that these relational data paradigms have for data use, preservation, and sharing, and discuss the need to support a plurality of data practices and paradigms.

Quantifying the importance of a realistic tomato (Solanum lycopersicum) leaflet shape for 3-D light modelling.

Leaflet shapes of tomato plants (Solanum lycopersicum) have been reduced to simple geometric shapes in previous functional-structural plant models (FSPMs) in order to facilitate measurements and reduce the time required to reconstruct the plant virtually. The level of error that such simplifications introduce remains unaddressed. This study therefore aims to quantify the modelling error associated with simplifying leaflet shapes. Realistic shapes were implemented in a static tomato FSPM based on leaflet scans, and simulation results were compared to simple geometric shapes used in previous tomato FSPMs in terms of light absorption and gross photosynthesis, for both a single plant and a glasshouse scenario. The effect of simplifying leaflet shapes in FSPMs leads to small but significant differences in light absorption, alterations of canopy light conditions and differences in photosynthesis. The magnitude of these differences depends on both the type of leaflet shape simplification used and the canopy shape and density. Incorporation of realistic shapes requires a small increase in initial measurement and modelling work to establish a shape database and comes at the cost of a slight increase in computation time. Our findings indicate that the error associated with leaflet shape simplification is small, but often unpredictable, and is affected by plant structure but also lamp placement, which is often a primary optimization goal of these static models. Assessment of the cost-benefit of realistic shape inclusion shows relatively little drawbacks for a decrease in model uncertainty.

STUDY OF EXTERNAL EYE DIMENSIONS AS POSSIBLE TOOL FOR SEX IDENTIFICATION IN EGYPTIAN POPULATION

Recognition of sex through eye dimensions is important especially that these dimensions is fixed after puberty and not liable for reshape especially in cases of intersex, mutilation of the genitalia ,transsexual cases and mass disasters.use of eye recognition as a biometric identification tool is accurate .aim of the study : is to differentiate between eye shape database of the males and females populations as possible tool for sex identification in male and females population in egypt . Methods: in our study A sum of 100 subjects (49males and 51 females) ranging from 15 to 30 years old were randomly chosen from ophthalmology outpatient clinic, Beni-Suef University, Egypt, interpalpebral distance , intercanthal distance and eye width were taken, The result Throughout the study, was revealed the gender plays a substantial role across the varieties of inter eye distance, data collected can be used as preliminary database for further research in the field of forensic medicine for sex identification.

3D shape classification with NNLS coding and optimal projections technique

Using graph-based encoding techniques and well-known shape descriptors a framework is presented here which is checked for 3D shape classification performance. A data-driven feature extraction procedure, taking the form of a simple projection operator, is in the core of our framework. Provided a shape database, a graph encapsulating the structural relationships among all the available shapes is first constructed and then is employed to define low-dimensional sparse projections. The weights in the graph, reflecting the database structure, are calculated so as to approximate each shape as a sparse linear combination of the remaining dataset objects. NNLS (nonnegative least squares) coding method is employed and fully exploited in this stage. Sparse coding with L2graph is also included in the framework. By way of solving a generalized eigenanalysis problem, a linear matrix operator is designed by means of optimal projections that will act as the feature extractor. A trained SVM (support vector machine) in the final stage of our framework makes the class prediction. Two popular, inherently high dimensional descriptors, namely ShapeDNA and Global Point Signature (GPS), with a modification emphasizing the smaller eigenvalues are employed in our experimentations with SHREC10, SHREC11 and SCHREC15 datasets. Classification results are very promising and outperform state of the art methods, providing evidence about the highly discriminative nature of the 3D shape representation produced in the proposed multistage framework.

Surface structuring using multi-stage grinding strategies based on geometric physically-based process simulations

Abstract For various industrial applications, e.g., the manufacturing of forming tools for the automotive sector, surface structures of functional areas play a decisive role. NC grinding processes on machining centers represent a flexible solution for finishing the free-form surfaces of forming tools. However, the surface topographies are influenced by several factors like the distribution of the grains and the wear of the tools. In this paper, a geometric physically-based simulation system is applied to predict the surface structures for different multi-stage grinding strategies. Point clouds were used for modeling the individual grains, which allows the simulation of different states of tool wear and the resulting influence on the surface topographies and process forces. Due to the high number of grains on the grinding tools, a stochastic model for the grain distributions and a database of different grain shapes were created based on representative measurements. The simulated surfaces are compared to experimental results in order to validate this approach.

Object-aware guidance for autonomous scene reconstruction

To carry out autonomous 3D scanning and online reconstruction of unknown indoor scenes, one has to find a balance between global exploration of the entire scene and local scanning of the objects within it. In this work, we propose a novel approach, which provides object-aware guidance for autoscanning, for exploring, reconstructing, and understanding an unknown scene within one navigation pass. Our approach interleaves between object analysis to identify the next best object (NBO) for global exploration, and object-aware information gain analysis to plan the next best view (NBV) for local scanning. First, an objectness-based segmentation method is introduced to extract semantic objects from the current scene surface via a multi-class graph cuts minimization. Then, an object of interest (OOI) is identified as the NBO which the robot aims to visit and scan. The robot then conducts fine scanning on the OOI with views determined by the NBV strategy. When the OOI is recognized as a full object, it can be replaced by its most similar 3D model in a shape database. The algorithm iterates until all of the objects are recognized and reconstructed in the scene. Various experiments and comparisons have shown the feasibility of our proposed approach.

Identification of a species diagnostic character for instar and juvenile mud crabs (Genus Scylla)

A species diagnostic character in instar and juvenile Scylla spp. allows for more focused use of resources in farms, enhanced research in wild populations and improved culture practices into preferred species. Species identification using molecular methods entails extraction of soft tissue that can be fatal to early developmental morphs. This prevents downstream applications that require live samples. In this study, a potential species diagnostic character from the frontal lobe spine shapes of 177 wild-caught Scylla crablets was generated. The species of the samples were identified using the molecular internal transcribed spacer 1 and 16S rDNA markers. Fourier transformation was employed on the images of their spines using the software SHAPE and the results underwent discriminant and principal component analyses. This method was able to assign 92.4% of Scylla serrata, 96.2% of S. olivacea and 90.1% of S. tranquebarica to the right species. The frontal lobe spine shapes of each species group were then traced and used as a means for species identification on 50 cultured and 100 wild-caught Scylla crablets through Resemble.js. This method was then able to assign 85.9% of the instars and 84.7% to the correct species, verified using molecular markers. Results of this study show great potential for the use of the frontal lobe spines as a morphological diagnostic character for instars and juveniles. Increased accuracy can be achieved by expanding the reference shape database and inclusion of more Scylla populations across the region.

Data driven composite shape descriptor design for shape retrieval with a VoR-Tree

We develop a data driven method (probability model) to construct a composite shape descriptor by combining a pair of scale-based shape descriptors. The selection of a pair of scale-based shape descriptors is modeled as the computation of the union of two events, i.e., retrieving similar shapes by using a single scale-based shape descriptor. The pair of scale-based shape descriptors with the highest probability forms the composite shape descriptor. Given a shape database, the composite shape descriptors for the shapes constitute a planar point set. A VoR-Tree of the planar point set is then used as an indexing structure for efficient query operation. Experiments and comparisons show the effectiveness and efficiency of the proposed composite shape descriptor.

On the definition of shape parts: a dominant sets approach

In the present paper, a novel graph-based approach to the shape decomposition problem is addressed. The shape is appropriately transformed into a visibility graph enriched with local neighborhood information. A two-step diffusion process is then applied to the visibility graph that efficiently enhances the information provided, thus leading to a more robust and meaningful graph construction. Inspired by the notion of a clique as a strict cluster definition, the dominant sets algorithm is invoked, slightly modified to comport with the specific problem of defining shape parts. The cluster cohesiveness and a node participation vector are two important outputs of the proposed graph partitioning method. Opposed to most of the existing techniques, the final number of the clusters is determined automatically, by estimating the cluster cohesiveness on a random network generation process. Experimental results on several shape databases show the effectiveness of our framework for graph-based shape decomposition.

Fit, stability and comfort assessment of custom-fitted bicycle helmet inner liner designs, based on 3D anthropometric data

Research has demonstrated that a better-fitted bicycle helmet offers improved protection to the rider during an impact. Nowadays, bicycle helmets in the market that range in size from small/medium to medium/large might not fit the diverse range of human head shapes and dimensions. 3D scanning was used to create 3D head shape databases of 20 participants who volunteered for the study. We developed new custom-fitted helmet inner liners, based on the 3D head shape of two sub-groups of participants, to map their head sizes and contours closely to the conventional Medium (M) and Large (L) sizes as described in from AS/NZS 2512.1: 2009. The new custom-fitted helmet was compared with the helmet available in the market place in a dynamics stability test and from participants' subjective feedback. A significant reduction in the angle of helmet rotation on the headform in the lateral direction was recorded for the custom-fitted helmet. A Wilcoxon signed-rank test was conducted to evaluate participants’ feedback on the helmets according to different area definitions. The overall fit and comfort and the top region of the new helmet were significantly improved. However, no difference was found at the significant level of 0.05 for the front and rear region of the new helmet.

Modeling indoor scenes with repetitions from 3D raw point data

Automatic modeling of indoor scenes from raw point data has received considerable attention due to its wide applications in computer graphics and robotics. The raw point data, however, always suffer from incompleteness, noise and anisotropy in density, which make rapid reconstruction fairly challenging. To overcome these challenges, in this paper, we explore the repeatability and regularity of man-made structures, which is the crux of our automatic reconstruction of indoor scenes. As observed, repetitive structures oftentimes exhibit in indoor scenes, such as classrooms, meeting rooms, and auditoria. We detect repetitions hierarchically and extract each repetitive object separately in these scenes. The object is represented with a set of key points which are extracted by leveraging both the local and global information of the data. By retrieving the most similar models from the shape database, we align them with the input data to obtain a high-quality virtual representation of the scene, which is quite faithful to the original geometry. In contrast to previous methods, we discover the high-level structure of the scene and can obtain a complete reconstruction efficiently even in the presence of noise and incomplete scans. A variety of indoor scenes have been tested to verify the effectiveness and the robustness of our proposed method.

Hierarchical complexity measures for effective shape-based image retrieval

A novel descriptor, called HCMD, which is based on hierarchical complexity measures, is proposed for generic shape based image retrieval. HCMD belongs to region-based methods and iteratively partitions a shape into smaller blocks along various directions. The geometrical properties of these smaller blocks, which are derived from each iterative cut, are measured to form a hierarchical description for the shape. The descriptor has the ability to characterize a shape from coarse to fine, and can effectively capture its complex inner structural features. Shape matching based on HCMD is independent of the rotation, scaling, translation, and mirror transform of the shape. It has low computational complexity and can effectively handle both the contour and region shapes. Three standard test sets, namely, the MPEG-7 CE-1 contour shape database, MPEG-7 CE-2 region shape database, and COIL-20 database, are used to evaluate the performance of the proposed HCMD, and extensive comparisons with state-of-the-art approaches, including five region-based descriptors, four point-set based descriptors, and two curve-based descriptors, are conducted. All experimental results indicate that the proposed HCMD outperforms these approaches in terms of their comprehensive performance based on the retrieval rates, retrieval efficiency, and general applications.

Shape retrieval through normalized B-splines curves

This paper proposes a new technique for 2D shape modeling and retrieval based only on curves defined from shape boundary. Firstly, a shape representation system is build based on the decomposition of the outline into its constituent parts and their geometric description. The process decomposition is done using high curvature points located along the boundary. These obtained parts are then described by parametric curves using the B-spline approximation and normalized in order to eliminate scaling transformation. Finally, the resulting curves allow matching of shapes and retrieving that is robust to rotation, scale change and deformation. Experiments conducted on a variety of shape databases including Kimia-99, Kimia-216, MPEG-7 and our database created from a selection of ETH-80 shape database, illustrate the performance of the proposed approach when compared with existing algorithms in literature. Obtained results are presented and discussed.

Shape Based Image Retrieval using Lower Order Zernike Moments

Shape is one of the significant features of Content Based Image Retrieval (CBIR). This paper proposes a strong and successful shape feature, which is based on a set of orthogonal complex moments of images known as Zernike moments. For shape classification Zernike moment (ZM) is the dominant solution. The radial polynomial of Zernike moment produces the number of concentric circles based on the order. As the order increases number of circles will increases, due to this the local information of an image will be ignored. In this paper, we introduced a novel method for radial polynomial where local information of an image given importance. We succeeded to extract the local features and shape features at very a low order of polynomial compared to the state of traditional ZM.The proposed method gives an advantage of a lower order, less complex, and lower dimension feature vector.For more similar images we find that simple Euclidian distance approximately zero. Proposed method tested on a MPEG-7 CE-1 shape database, Coil-100 databases. Experiments demonstrated that it is outperforming in identifying the shape of an object in the image and reduced the retrieving time and complexity of calculations.

An efficient unsupervised diffusion clustering algorithm with application to shape decomposition based on visibility context

In this paper the two and three dimensional of a single shape partitioning problem is revisited by means of an iterative, non supervised, very fast and effective algorithm. The notion of visibility context is used as the shape signature, which actually provides a physical meaning to the representation. The visibility serves as a means to manipulate on the shape parts applying clustering techniques to the corresponding graph. Therefore, the decomposition problem is re-casted as a clustering problem. An unsupervised Diffusion Clustering Algorithm is proposed, which efficiently achieves to capture the functional shape parts. Although the proposed algorithm is developed and fits very well to the specific problem of the shape partitioning, its utility is undoubtedly much more general. Experimental results conducted on two and three dimensional shape databases are very promising.

Classification of two-dimensional figures using skeleton-geodesic histograms of thicknesses and distances

The paper considers a problem of shape representation and classification. We propose a skeleton-geodesic histogram of thicknesses and distances for this purpose. It is based on the statistics of pair distances between shape elements. It is computed using skeleton-geodesic distances and thickness differences between pairs of skeleton edges. This differs from conventional geodesic histograms that are calculated for all figure points. The switch to the skeleton edges and areas of their attraction significantly speeds up the calculation of skeleton-geodesic histogram of thicknesses and distances, while maintaining many useful properties inherent in usual geodesic histograms. Extensive experimentation has been conducted on the most difficult binary shape database. Obtained classification results indicate the high potential of the proposed descriptor.

Optimized shape semantic graph representation for object understanding and recognition in point clouds

To understand and recognize the three-dimensional (3-D) objects represented as point cloud data, we use an optimized shape semantic graph (SSG) to describe 3-D objects. Based on the decomposed components of an object, the boundary surface of different components and the topology of components, the SSG gives a semantic description that is consistent with human vision perception. The similarity measurement of the SSG for different objects is effective for distinguishing the type of object and finding the most similar one. Experiments using a shape database show that the SSG is valuable for capturing the components of the objects and the corresponding relations between them. The SSG is not only suitable for an object without any loops but also appropriate for an object with loops to represent the shape and the topology. Moreover, a two-step progressive similarity measurement strategy is proposed to effectively improve the recognition rate in the shape database containing point-sample data.