Color and shape retrieval are two important retrieval methods in content based image database. Recently we developed a new method which use Schwarz representation to match one-dimensional signal. It can obtain closed ...
详细信息
Color and shape retrieval are two important retrieval methods in content based image database. Recently we developed a new method which use Schwarz representation to match one-dimensional signal. It can obtain closed form match function and similarity measure without optimization. Color histogram is a natural one-dimensional signal. Contour can be converted into onedimensional signal using its tangent angle. In this paper, we introduce this method to image retrieval based on the color and shape, respectively. Experimental results show its efficiency and accuracy.
On the basis of discussing General Feed-Forward Networks (GFFN), a Sequential Learning Ahead Masking (SLAM) model and its relevant algorithm for pattern classification are proposed. By adapting this model to the patte...
详细信息
On the basis of discussing General Feed-Forward Networks (GFFN), a Sequential Learning Ahead Masking (SLAM) model and its relevant algorithm for pattern classification are proposed. By adapting this model to the pattern classifier, the computer simulation results show that not only the convergence speed and performance of the network are much better than the existing modified BP algorithms, but also the required network scale is greatly reduced. Moreover, Double- threshold Neuron (DTN) has been applied to SLAM network for pattern classification. the SLAM pattern classifier has been implemented on the domestic micro-neurocomputer CASSANDRA-I and the results are provided as below: For two-class pattern classification problem with 1024 patterns generated randomly in 256-dimensional pattern space, the training time is about 3 hours 20 minutes, and the running time for pattern classification is 0.007 seconds.
Methods based on invariants can be used to recognize objects. In this paper, an invariant of points in stereo vision based on a reference point and two homographies is derived and compared withthe method recently pro...
详细信息
Methods based on invariants can be used to recognize objects. In this paper, an invariant of points in stereo vision based on a reference point and two homographies is derived and compared withthe method recently proposed by Shashua. the other invariant of lines in stereo vision based on one homography and one reference line is also derived. Although basically they still are cross ratios in projective geometry, they are very useful in vision applications.
Introduces a fuzzy representation for isolated character description. this representation maps a character from its original sequence of 2D coordinates into a fuzzy vector space that can thereafter serve as input to a...
详细信息
Human vision has marvelous ability in extracting linear features from images, such as roads, rivers and so on. In this paper we present a new method to simulate this ability. Our method is based on some general groupi...
详细信息
Human vision has marvelous ability in extracting linear features from images, such as roads, rivers and so on. In this paper we present a new method to simulate this ability. Our method is based on some general grouping factors arising at two levels. At the first level, grouping factors are identified as direct bar-bar interaction and orientation interaction. Bar-bar interaction is short-ranged and homogeneous. Orientation interaction is locally oriented and mediated by statistics of local visual context. At the second level, grouping factors are global geometric binding effects which arise from geometric redundancy reduction and thus are global effects. Based on them, we present an energy model. then the extraction of linear features is generally formulated as combinatorial optimization. Since local, global interactions and local context effects are all included in it, the model may capture partially grouping ability of human vision systems. the experiments show that, without selecting any original points, our method can extract linear features from images robustly and quickly.
In literatures, two enhancement algorithms, spatial and spatiotemporal homomorphic filtering (SHF and SthF) have been proposed for enhancement of the far infrared image based upon a far infrared imaging model. Althoug...
详细信息
In literatures, two enhancement algorithms, spatial and spatiotemporal homomorphic filtering (SHF and SthF) have been proposed for enhancement of the far infrared image based upon a far infrared imaging model. Although spatiotemporal homomorphic filtering may reduce the number of iterations greatly in comparison to spatial one for a similar degree of convergence by making explicit use of the additional information provided temporally, the enhanced results from SHF are in general better than those from SthF. In this paper, we design an adaptive spatiotemporal homomorphic filtering (ASthF) which can not only produce enhanced images as good as those from SHF, but keep the number of iterations as low as that of SthF, even lower, for a similar degree of convergence as well.
Deformable models (DMs) have been receiving a growing interest on recognizing non-rigid objects for its shape-varying capability. However, there are still no mechanism embedded in existing DMs to account for highly st...
详细信息
Deformable models (DMs) have been receiving a growing interest on recognizing non-rigid objects for its shape-varying capability. However, there are still no mechanism embedded in existing DMs to account for highly structural patterns and most of them indeed can only model a close or open contour. Also, for this kind of elastic matching approach, many underlying algorithms are indeed adopting pixel-to-pixel matching strategy which induces a high computational cost compared with edge matching one. It leads us to study on a new kind of DM called structural DM to model the patterns together with edge matching strategy to define the image force. then, the elastic matching process is formulated in a Bayesian framework and solved using steepest descent method. A scheme is introduced to ensure the preservation of global structure in terms of relative stroke ordering during minimization, and a multiresolution scheme is also described. the performance of the new model is demonstrated through a small scale chinese character recognition experiment.
the proceedings contain 113 papers. the special focus in this conference is on Structural Matching, Grammatical Inference and recognition of 2D and 3D Objects. the topics include: Error-tolerant graph matching;semanti...
ISBN:
(纸本)3540648585
the proceedings contain 113 papers. the special focus in this conference is on Structural Matching, Grammatical Inference and recognition of 2D and 3D Objects. the topics include: Error-tolerant graph matching;semantic content based image retrieval using object-process diagrams;patternrecognition methods in image and video databases;efficient matching with invariant local descriptors;integrating numerical and syntactic learning models for patternrecognition;synthesis of function-described graphs;marked subgraph isomorphism of ordered graphs;distance evaluation in pattern matching based on frontier topological graph;syntactic interpolation of fractal sequences;minimizing the topological structure of line images;genetic algorithms for structural editing;the noisy subsequence tree recognition problem;object recognition from large structural libraries;acquisition of 2-d shape models from scenes with overlapping objects using string matching;a taxonomy of occlusion in view signature ii representations;a survey of non-thinning based vectorization methods;a benchmark for raster to vector conversion systems;network-based recognition of architectural symbols;recovering image structure by model-based interaction map;an improved scheme to fingerprint classification;character recognition with k-head finite array automata;using semantics in matching cursive chinese handwritten annotations;concavity detection using a binary mask-based approach;structural indexing of line pictures with feature generation models;nonlinear covariance for multi-band image data;a neural network for image smoothing and segmentation;prototyping structural descriptions and neural network based learning of local compatibilities for segment grouping.
Fundamental matrix plays an important role in an un-calibrated stereo vision system. Many researchers designed algorithms to estimate the fundamental matrix under known corresponding points in stereo images. Because t...
详细信息
Fundamental matrix plays an important role in an un-calibrated stereo vision system. Many researchers designed algorithms to estimate the fundamental matrix under known corresponding points in stereo images. Because the fundamental matrix is sensitive to image noise, it is known to be a difficult problem. In this paper, we make use of knowledge of the scene, given planar objects in the scene, a linear and robust method to estimate the fundamental matrix is proposed and implemented.
In this paper, an iterative algorithm for the model-based P3P pose-estimation problem based on the Gauss-Newton method is described. In this system, correspondences between three noncollinear object points on a rigid ...
详细信息
In this paper, an iterative algorithm for the model-based P3P pose-estimation problem based on the Gauss-Newton method is described. In this system, correspondences between three noncollinear object points on a rigid body and their image points after movement under full perspective projection model are used to estimate the new position and orientation of the object. Both real images and synthetic data have been used to verify our method with satisfactory results.
暂无评论