Problem Of Shape Recognition Research Articles

3DSliceLeNet: Recognizing 3D Objects Using a Slice-Representation

Convolutional Neural Networks (CNNs) have become the default paradigm for addressing classification problems, especially, but not only, in image recognition. This is mainly due to their high success rate. Although a number of approaches currently apply deep learning to the 3D shape recognition problem, they are either too slow for online use or too error-prone. To fill this gap, we propose 3DSliceLeNet, a deep learning architecture for point cloud classification. Our proposal converts the input point clouds into a two-dimensional representation by performing a slicing process and projecting the points to the principal planes, thus generating images that are used by the convolutional architecture. 3DSliceLeNet successfully achieves both high accuracy and low computational cost. A dense set of experiments has been conducted to validate our system under the ModelNet challenge, a large-scale 3D Computer Aided Design (CAD) model dataset. Our proposal achieves a success rate of 94.37% and an Area under Curve (AUC) of 0.978 on the ModelNet-10 classification task.

Isometry Invariant Shape Recognition of Projectively Perturbed Point Clouds by the Mergegram Extending 0D Persistence

Rigid shapes should be naturally compared up to rigid motion or isometry, which preserves all inter-point distances. The same rigid shape can be often represented by noisy point clouds of different sizes. Hence, the isometry shape recognition problem requires methods that are independent of a cloud size. This paper studies stable-under-noise isometry invariants for the recognition problem stated in the harder form when given clouds can be related by affine or projective transformations. The first contribution is the stability proof for the invariant mergegram, which completely determines a single-linkage dendrogram in general position. The second contribution is the experimental demonstration that the mergegram outperforms other invariants in recognizing isometry classes of point clouds extracted from perturbed shapes in images.

A generic interval branch and bound algorithm for parameter estimation

In engineering sciences, parameter estimation is a challenging problem consisting in computing the parameters of a parametric model that fit observed data. The system is defined by unknown parameters and sometimes internal constraints. The observed data provide constraints on the parameters. This problem is particularly difficult when some observation constraints correspond to outliers and/or the constraints are non convex. The ransac randomized algorithm can efficiently handle it, but is non deterministic and must be specialized for every problem. This paper presents the first generic interval branch and bound algorithm that produces a model maximizing the number of observation constraints satisfied within a given tolerance. This tool is inspired by the IbexOpt branch and bound algorithm for constrained global optimization (NLP) and is endowed with an improved version of a relaxed intersection operator applied to observations. Experiments have been carried out on two different computer vision problems. They highlight a significant speedup w.r.t. Jaulin et al.’s interval method in 2D and 3D shape recognition problems (having three parameters). We have also obtained promising results on a stereo vision problem where the essential matrix (five parameters) is estimated exactly at a good accuracy in hours for models having a thousand points, a typical size for such problems.

An Improved 3D Shape Recognition Method Based on Panoramic View

Recognition of three-dimensional (3D) shape is a remarkable subject in computer vision systems, because of the lack of excellent shape representations. With the development of 2.5D depth sensors, shape recognition is becoming more important in practical applications. Many methods have been proposed to preprocess 3D shapes, in order to get available input data. A common approach employs convolutional neural networks (CNNs), which have become a powerful tool to solve many problems in the field of computer vision. DeepPano, a variant of CNN, converts each 3D shape into a panoramic view and shows excellent performance. It is worth paying attention to the fact that both serious information loss and redundancy exist in the processing of DeepPano, which limits further improvement of its performance. In this work, we propose a more effective method to preprocess 3D shapes also based on a panoramic view, similar to DeepPano. We introduce a novel method to expand the training set and optimize the architecture of the network. The experimental results show that our approach outperforms DeepPano and can deal with more complex 3D shape recognition problems with a higher diversity of target orientation.

A Novel Signal Similarity Evaluation Algorithm and its Application in Irregular Shape Recognition

Irregular shape recognition has been a tough problem in image understanding field. One type of feature extraction techniques involves transforming shapes into signals through certain process, and thus the shape recognition problem becomes comparing similarity between signals. This paper proposes a novel signal similarity evaluation scheme that may classify signals into different categories to fulfill the shape recognition task. Experiments on pavement guide marks recognition are conducted to study the potential of the proposed method. Results demonstrate its ability in irregular shape recognition applications with higher accuracy and faster speed.

Shape recognition by bag of skeleton-associated contour parts

Contour and skeleton are two complementary representations for shape recognition. However combining them in a principal way is nontrivial, as they are generally abstracted by different structures (closed string vs graph), respectively. This paper aims at addressing the shape recognition problem by combining contour and skeleton according to the correspondence between them. The correspondence provides a straightforward way to associate skeletal information with a shape contour. More specifically, we propose a new shape descriptor, named Skeleton-associated Shape Context (SSC), which captures the features of a contour fragment associated with skeletal information. Benefited from the association, the proposed shape descriptor provides the complementary geometric information from both contour and skeleton parts, including the spatial distribution and the thickness change along the shape part. To form a meaningful shape feature vector for an overall shape, the Bag of Features framework is applied to the SSC descriptors extracted from it. Finally, the shape feature vector is fed into a linear SVM classifier to recognize the shape. The encouraging experimental results demonstrate that the proposed way to combine contour and skeleton is effective for shape recognition, which achieves the state-of-the-art performances on several standard shape benchmarks.

Fragmentary shape recognition: A BCI study

Recently, Brain–Computer Interface (BCI) has emerged as a potential modality that utilizes natural and intuitive human mechanisms of thinking process to enable interactions in CAD interfaces. Before BCI could become a mainstream mode of HCI for CAD interfaces; fundamental studies directed towards understanding how humans mentally represent and process the geometry are needed. The outlined work in this paper presents an objective user study to understand shape recognition process in the humans. Specifically, we focus on the fundamental task of fragmentary shape identification. The problem of fragmentary shape recognition can be defined as follows: given a partial and incomplete minimalistic representation of a given shape, can one recognize the actual complete shape or object? In user studies, each subject was progressively (in stages) shown more informative fragmented images of an object to be recognized. During each stage of the experiment, the brain activity of users in the form of electroencephalogram (EEG) signals was recorded with a BCI headset. The recorded signals are then processed to objectively study the fragmentary shape recognition process. The results of user studies conclusively show that the measured brain activities of subjects can serve as a very accurate proxy to estimate subjects fragmentary shape recognition process.

3D morphology of a random field from its 2D cross-section

Abstract The two aspect ratios of randomly oriented triaxial ellipsoids (representing isosurfaces of an isotropic 3D random field) can be determined from a single 2D cross-section of their sample using the probability density function (PDF) of the filamentarityF of individual structures seen in cross-section (F = 0 for a circle and F = 1 for a line). The PDF of F has a robust form with a sharp maximum and truncation at larger F, and the most probable and maximum values of F are uniquely and simply related to the two aspect ratios of the triaxial ellipsoids. The parameters of triaxial ellipsoids of randomly distributed sizes can still be recovered from the PDF of F. This method is applicable to many shape-recognition problems, here illustrated by the neutral hydrogen density in the turbulent interstellar medium of the Milky Way. The gas distribution is shown to be filamentary with axis ratios of about 1:2:20.

Demisting the Hough Transform for 3D Shape Recognition and Registration

In applying the Hough transform to the problem of 3D shape recognition and registration, we develop two new and powerful improvements to this popular inference method. The first, intrinsic Hough, solves the problem of exponential memory requirements of the standard Hough transform by exploiting the sparsity of the Hough space. The second, minimum-entropy Hough, explains away incorrect votes, substantially reducing the number of modes in the posterior distribution of class and pose, and improving precision. Our experiments demonstrate that these contributions make the Hough transform not only tractable but also highly accurate for our example application. Both contributions can be applied to other tasks that already use the standard Hough transform.

Multi-circle detection on images inspired by collective animal behavior

Hough transform (HT) has been the most common method for circle detection that delivers robustness but adversely demands considerable computational efforts and large memory requirements. As an alternative to HT-based techniques, the problem of shape recognition has also been handled through optimization methods. In particular, extracting multiple circle primitives falls into the category of multi-modal optimization as each circle represents an optimum which must be detected within the feasible solution space. However, since all optimization-based circle detectors focus on finding only a single optimal solution, they need to be applied several times in order to extract all the primitives which results on time-consuming algorithms. This paper presents an algorithm for automatic detection of multiple circular shapes that considers the overall process as a multi-modal optimization problem. In the detection, the approach employs an evolutionary algorithm based on the way in which the animals behave collectively. In such an algorithm, searcher agents emulate a group of animals which interact to each other using simple biological rules. These rules are modeled as evolutionary operators. Such operators are applied to each agent considering that the complete group maintains a memory which stores the optimal solutions seen so-far by applying a competition principle. The detector uses a combination of three non-collinear edge points as parameters to determine circle candidates (possible solutions). A matching function determines if such circle candidates are actually present in the image. Guided by the values of such matching functions, the set of encoded candidate circles are evolved through the evolutionary algorithm so that the best candidate (global optimum) can be fitted into an actual circle within the edge-only image. Subsequently, an analysis of the incorporated memory is executed in order to identify potential local optima which represent other circles. Experimental results over several complex synthetic and natural images have validated the efficiency of the proposed technique regarding accuracy, speed and robustness.

A Gromov-Hausdorff Framework with Diffusion Geometry for Topologically-Robust Non-rigid Shape Matching

In this paper, the problem of non-rigid shape recognition is studied from the perspective of metric geometry. In particular, we explore the applicability of diffusion distances within the Gromov-Hausdorff framework. While the traditionally used geodesic distance exploits the shortest path between points on the surface, the diffusion distance averages all paths connecting the points. The diffusion distance constitutes an intrinsic metric which is robust, in particular, to topological changes. Such changes in the form of shortcuts, holes, and missing data may be a result of natural non-rigid deformations as well as acquisition and representation noise due to inaccurate surface construction. The presentation of the proposed framework is complemented with examples demonstrating that in addition to the relatively low complexity involved in the computation of the diffusion distances between surface points, its recognition and matching performances favorably compare to the classical geodesic distances in the presence of topological changes between the non-rigid shapes.

Vibration mode shape recognition using image processing

Currently the most widely used method for comparing mode shapes from finite elements and experimental measurements is the modal assurance criterion (MAC), which can be interpreted as the cosine of the angle between the numerical and measured eigenvectors. However, the eigenvectors only contain the displacement of discrete coordinates, so that the MAC index carries no explicit information on shape features. New techniques, based upon the well-developed philosophies of image processing (IP) and pattern recognition (PR) are considered in this paper. The Zernike moment descriptor (ZMD), Fourier descriptor (FD), and wavelet descriptor (WD) are the most popular shape descriptors due to their outstanding properties in IP and PR. These include (1) for the ZMD-rotational invariance, expression and computing efficiency, ease of reconstruction and robustness to noise; (2) for the FD—separation of the global shape and shape-details by low and high frequency components, respectively, invariance under geometric transformation; (3) for the WD—multi-scale representation and local feature detection. Once a shape descriptor has been adopted, the comparison of mode shapes is transformed to a comparison of multidimensional shape feature vectors. Deterministic and statistical methods are presented. The deterministic problem of measuring the degree of similarity between two mode shapes (possibly one from a vibration test and the other from a finite element model) may be carried out using Pearson's correlation. Similar shape feature vectors may be arranged in clusters separated by Euclidian distances in the feature space. In the statistical analysis we are typically concerned with the classification of a test mode shape according to clusters of shape feature vectors obtained from a randomised finite element model. The dimension of the statistical problem may often be reduced by principal component analysis. Then, in addition to the Euclidian distance, the Mahalanobis distance, defining the separation of the test point from the cluster in terms of its standard deviation, becomes an important measure. Bayesian decision theory may be applied to formally minimise the risk of misclassification of the test shape feature vector. In this paper the ZMD is applied to the problem of mode shape recognition for a circular plate. Results show that the ZMD has considerable advantages over the traditional MAC index when identifying the cyclically symmetric mode shapes that occur in axisymmetric structures at identical frequencies. Mode shape recognition of rectangular plates is carried out by the FD. Also, the WD is applied to the problem of recognising the mode shapes in the thin and thick regions of a plate with different thicknesses. It shows the benefit of using the WD to identify mode-shapes having both local and global components. The comparison and classification of mode shapes using IP and PR provides a ‘toolkit’ to complement the conventional MAC approach. The selection of a particular shape descriptor and classification method will depend upon the problem in hand and the experience of the analyst.

Disconnected Skeleton: Shape at Its Absolute Scale

We present a new skeletal representation along with a matching framework to address the deformable shape recognition problem. The disconnectedness arises as a result of excessive regularization that we use to describe a shape at an attainably coarse scale. Our motivation is to rely on stable properties the shape instead of inaccurately measured secondary details. The new representation does not suffer from the common instability problems of the traditional connected skeletons, and the matching process gives quite successful results on a diverse database of 2D shapes. An important difference of our approach from the conventional use of skeleton is that we replace the local coordinate frame with a global Euclidean frame supported by additional mechanisms to handle articulations and local boundary deformations. As a result, we can produce descriptions that are sensitive to any combination of changes in scale, position, orientation and articulation, as well as invariant ones.

Decomposition of two-dimensional shapes for efficient retrieval

This paper presents a novel approach to address the problem of generic 2D shape recognition. We propose a morphological method to decompose a binary shape into entities in correspondence with their protrusions. Each entity is associated with a set of perceptual features that can be used in indexing into image databases. The matching process, based on the softassign algorithm, has produced encouraging results, showing the potential of the developed method in a variety of computer vision and pattern recognition domains. The results demonstrate its robustness in the presence of scale, reflection and rotation transformations and prove the ability to handle noise and articulated structures. In order to increase efficiency, the retrieval process is applied after a coarse scale grouping of objects, without sacrificing effectiveness and allowing indexing into large shape databases.

Size invariant shape recognition by modulated competitive neural circuit

This paper addresses the problem of size invariant shape recognition based on scale transformation within a modulated competition neural layer. In this paper we discuss the advantages of the application of a neural network to pattern recognition and study how the traditional automatic target recognition can fail to recognise a known pattern due to either a size change or background clutter and distortion. The model construction is based on neurophysiology experiments on vision systems. The Neural Circuit Simulation studies undertaken demonstrate the effectiveness of the proposed model in recognising 2D objects in non-ideal visual conditions.

Permutation Coding Technique for Image Recognition Systems

A feature extractor and neural classifier for image recognition systems are proposed. The proposed feature extractor is based on the concept of random local descriptors (RLDs). It is followed by the encoder that is based on the permutation coding technique that allows to take into account not only detected features but also the position of each feature on the image and to make the recognition process invariant to small displacements. The combination of RLDs and permutation coding permits us to obtain a sufficiently general description of the image to be recognized. The code generated by the encoder is used as an input data for the neural classifier. Different types of images were used to test the proposed image recognition system. It was tested in the handwritten digit recognition problem, the face recognition problem, and the microobject shape recognition problem. The results of testing are very promising. The error rate for the Modified National Institute of Standards and Technology (MNIST) database is 0.44% and for the Olivetti Research Laboratory (ORL) database it is 0.1%.

Shape recognition based on neural networks trained by differential evolution algorithm

In this paper a new method for recognition of 2D occluded shapes based on neural networks using generalized differential evolution training algorithm is proposed. Firstly, a generalization strategy of differential evolution algorithm is introduced. And this global optimization algorithm is applied to train the multilayer perceptron neural networks. The proposed algorithms are evaluated through a plant species identification task involving 25 plant species. For this practical problem, a multiscale Fourier descriptors (MFDs) method is applied to the plant images to extract shape features. Finally, the experimental results show that our proposed GDE training method is feasible and efficient for large-scale shape recognition problem. Moreover, the experimental results illustrated that the GDE training algorithm combined with gradient-based training algorithms will achieve better convergence performance.

Shape recognition from three-dimensional point measurements with range and direction uncertainty

We derive a pair of algorithms, one optimal and the other approximate, for recognizing three-dimensional objects from a collection of points chosen from their surface according to some probabilistic mechanism. The measurements are assumed to be noisy, and the mea- sured location of a given point is translated according to a noise prob- ability distribution. Distributions governing surface point selection and measurement noise can take a variety of forms depending upon the particular measurement scenario. At one extreme, each measurement is assumed to yield values restricted to a one-dimensional ray, a special case commonly adopted in the literature. At the other extreme, measured points are chosen uniformly from the object's surface, and the noise distribution is spherically symmetric, a worst-case scenario that involves no prior information about the measurements. We apply these two algo- rithms to shape recognition problems involving simple geometrical ob- jects, and examine their relative behavior using a combination of analyti- cal derivation and Monte Carlo simulation. We show that the approximate algorithm can be far simpler to compute, and its perfor- mance is competitive with the optimal algorithm when noise levels are relatively low. We show the existence of a critical noise level, beyond which the approximate algorithm exhibits catastrophic failure. © 2005 So-

상대거리-곡률 특징 공간을 이용한 형태 기술 및 인식

영상에 회전이나 크기 변형이 가해지면 영상을 구성하는 점들의 좌표값들이 변경되어 형태 기술 및 인식이 어렵게 된다. 그러나 영상을 구성하는 점들 간의 위치관계나 무게중심과의 위치 관계는 변하지 않는다. 따라서 x-y 좌표계로 기술되는 영상 공간의 점들을 회전 및 크기 변형에 불변하는 새로운 좌표계로 사상할 수 있다면, 형태 기술 및 인식의 문제는 보다 수월해진다. 본 논문에서는 영상 공간의 점들을 회전 및 크기 변형에 무관한 새로운 특징 공간으로 사상하여 형태를 기술하는 방법을 제안한다. 특징 공간을 나타내는 새로운 좌표계는 무게중심으로부터의 상대거리와 윤곽선 세그먼트 곡률을 두 축으로 하는 직교 좌표계이다. 상대거리는 윤곽선 상의 임의의 한 점이 무게중심에서 얼마나 멀리 벗어나 있는지를 나타내는 값이고, 윤곽선 세그먼트 곡률은 세그먼트의 굴곡도를 나타내는 값이다. 특징 공간에 사상된 점들의 형태 기술은 메쉬 특징을 통해 이루어진다. 실험을 통해 제안된 형태 기술 방법이 회전 및 크기 변형에 강건함을 확인하였다. Rotation and scale variations make it difficult to solve the problem of shape description and recognition because these variations change the location of points composing the shape. However, some geometric Invariant points and the relations among them are not changed by these variations. Therefore, if points in image space depicted with the r-y coordinates system can be transformed into a new coordinates system that are invariant to rotation and scale, the problem of shape description and recognition becomes easier. This paper presents a shape description method via transformation from the image space into the invariant feature space having two axes: representing relative distance from a centroid and contour segment curvature(CSC). The relative distance describes how far a point departs from the centroid, and the CSC represents the degree of fluctuation in a contour segment. After transformation, mesh features were used to describe the shape mapped onto the feature space. Experimental results show that the proposed method is robust to rotation and scale variations.

Model-guided deformable hand shape recognition without positioning aids

This work addresses the problem of deformable hand shape recognition in biometric systems without any positioning aids. We separate and recognize multiple rigid fingers under Euclidean transformations. An elliptical model is introduced to represent fingers and accelerate the search of optimal alignments of fingers. Unlike other methods, the similarity is measured during alignment search based on finger width measurements defined at nodes by controllable intervals to achieve balanceable recognition accuracy and computational cost. Technically, our method bridges the traditional handcrafted-feature methods and the shape-distance methods. We have tested it using our 108-person-540-sample database with significantly increased positive recognition accuracy.