Browsing by Author "Chen, Z"

Now showing 1 - 20 of 23

3-D polyhedral face computation from two perspective views with the aid of a calibration plate
(1997-04-01)
Chiu, JM
;
Chen, Z
;
Wang, CM
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
The 3-D reconstruction of visible polyhedral faces from a pair of general perspective views with the aid of a calibration plate is addressed. A polyhedron is placed on a planar calibration plate and two side views of both the polyhedron and the calibration plate are taken. Through proper arrangements we may assume that in the two views a number of polyhedral edges lying on the calibration plate and the whole calibration plate boundary are visible. We present an on-line camera calibration technique with the aid of the calibration plate and a two-stage process to find the vertex/edge correspondences without encountering the ambiguity problem of the conventional epipolar line technique. Then we give a closed form solution to the 3-D polyhedral vertices visible in both images. We also describe other advantages of using our method for the 3-D polyhedron reconstruction. Experimental results show that the obtained 3-D polyhedral vertex coordinates are rather accurate.
3D shape recovery of complex objects from multiple silhouette images
(2003-06-01)
Fang, YH
;
Chou, HL
;
Chen, Z
;
資訊工程學系
;
Department of Computer Science
A reconstruction method is proposed which represents the object with a line-based geometric model. The method does not need the point correspondence information in recovering the 3D object geometry. It is based on the concept of volume intersection, but it is substantially different from the existing octree-based reconstruction methods in the aspects of data structure, reconstruction process and representation uniqueness under a 2D rigid motion. For visualizing the 3D reconstructed object geometry a conversion from the line-based geometric model to a bounded triangular mesh model is developed. The experimental results show that the method is capable of capturing the different details of the object. And it works fast and requires a relatively low memory space. (C) 2002 Elsevier Science B.V. All rights reserved.
A 3D predictive visual tracker for tracking multiple moving objects with a stereo vision system
(1995)
Hung, YP
;
Tang, CY
;
Shih, SW
;
Chen, Z
;
Lin, WS
This paper presents a 3D feature-based visual tracker for tracking multiple moving objects by using a predictor that first partitions 3D features into different common-motion clusters and then predicts the motion of each cluster with Kalman filters. The 3D features are computed from a sequence of stereo images by combining two 2D temporal matching modules and one stereo correspondence module. To partition the 3D features into different common-motion clusters, we propose a RANSAC-based clustering method by using rigid body consensus which assumes that all the extracted 3D features on a rigid body have the same 3D motion. By using the motion estimates obtained with the RANSAC-based method as the measurements, we are able to use linear Kalman filters to predict the motion of each cluster, and then, to predict the next position of each 3D feature. Preliminary experiments showed that the proposed 3D predictive visual tracker can serve as a robust 3D feature tracker for an active stereo vision system.
A method for rectifying grid junctions in grid-coded images using cross ratio
(1996-08-01)
Chia, TL
;
Chen, Z
;
Yueh, CJ
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
In the grid-coding technique, the accurate grid junction locations are vital to the surface computation. Noise arising from the imaging formation and processing processes will jeopardize the computation accuracy. The objective of this study is to make use of the geometric property of cross ratio to improve the accuracy of grid junction locations in grid-coded images., A method is devised to regenerate grid junctions from the original (i.e., raw) junction data of an image. Through a statistical error analysis, we show that the regenerated junction data are generally more accurate than the old ones in the image taken from a wide range of viewing angles.
A method for rectifying grid junctions in grid-coded images using cross ratio
(1996-08-01)
Chia, TL
;
Chen, Z
;
Yueh, CJ
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
In the grid-coding technique, the accurate grid junction locations are vital to the surface computation. Noise arising from the imaging formation and processing processes will jeopardize the computation accuracy. The objective of this study is to make use of the geometric property of cross ratio to improve the accuracy of grid junction locations in grid-coded images., A method is devised to regenerate grid junctions from the original (i.e., raw) junction data of an image. Through a statistical error analysis, we show that the regenerated junction data are generally more accurate than the old ones in the image taken from a wide range of viewing angles.
A novel 3D planar object reconstruction from multiple uncalibrated images using the plane-induced homographies
(2004-09-01)
Chou, HL
;
Chen, Z
;
資訊工程學系
;
Department of Computer Science
A computer vision method is proposed to determine all the visible 3D planar surfaces in a scene from uncalibrated images and locate them in a single 3D projective space. Most of the existing methods for reconstructing planar objects use point correspondences to estimate the fundamental matrix and derive the compatible projection equations, before they apply the standard triangulation technique to find the 3D points and fit the planes to the 3D points. This type of approaches is generally slow and less accurate because the 3D points are estimated separately, making them vulnerable to image error. We present a plane based reconstruction method to estimate the 3D projective structure using the planar homographies estimated from the plane features in the images. First, we estimate the homography for each visible plane.. and then we use the homographies of two primary planes to compute an epipole. We proceed to represent the epipolar geometry for each image pair using the estimated homography and epipole, together with a specified reference plane coefficient vector. Next, we show that the 3D plane coefficient vector of any plane visible in each image pair can be determined with respect to the reference plane coefficient vector once its planar homography is found. Finally, the reconstruction results obtained in individual projective spaces are integrated within a common projective space. To this end. we use the homography and plane equation information of two planes and the epipole associated to derive the coordinate transformation matrix between two involved projective spaces. To evaluate the performance of our method, we apply our method to the synthetic images and real images. All the results indicate the method works successfully. (C) 2004 Published by Elsevier B.V.
A parallel algorithm for generating chain code of objects in binary images
(2003-02-01)
Chia, TL
;
Wang, KB
;
Chen, LR
;
Chen, Z
;
資訊科學與工程研究所
This paper addresses parallel execution of chain code generation on a linear array architecture. The contours in the proposed algorithm are viewed as a set of edges (or contour segments) that can be traced by a top-down contour tracing method to generate the chain codes for the outer and inner object contours. A parallel algorithm that contains the chain code generating rules and operations needed is also described, and the algorithm is mapped onto a one-dimensional systolic array containing [(1)/(2)(N + 1)] processing elements (PEs) to devise this architecture. The architecture extracts the contours of objects and quickly generates the corresponding chain codes after the image data in all rows are inputted in a linear fashion. The total processing time for generating the chain codes in an N x N image is O(3N). By doing so, the real-time requirement is fulfilled and its execution time is independent of the image content. In addition, a partition method is developed to process an image when the parallel architecture has a fixed number of PEs; say two or more. The total execution time for an N x N image by employing a fixed number of PEs is N(N + 1)/M + 2(M - 1), when M is the fixed number of PEs. (C) 2002 Elsevier Science Inc. All rights reserved.
A simple recursive method for converting a chain code into a quadtree with a lookup table
(2001-05-01)
Chen, Z
;
Chen, IP
;
資訊工程學系
;
Department of Computer Science
We present a simple recursive method for converting a chain code into a quadtree. We generate the quadtree black nodes level by level, starting from the finest resolution (i.e. the bottom) level toward the coarsest resolution (i.e. the top) level. Meanwhile, at each resolution level a new object border is revealed after the removal of the black nodes. The chain code elements for this new object border can be easily generated. Thus, the quadtree black nodes and the chain code elements of the new object border are generated side by side and their generations constitute a basic cycle of the conversion process. We show the generations can be done with the aid of a lookup table. Finally, we compare our method with the existing two-phase methods exemplified by Samet's method in terms of the total numbers of major operations including node allocations, node color filling, and node pointer linking. It indicates that our method is conceptually simpler and runs faster. (C) 2001 Elsevier Science B.V. All rights reserved.
A study on the dual vanishing point property
(1999-12-01)
Huang, JB
;
Chen, Z
;
Lin, JY
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
Vanishing points and vanishing lines are useful information in computer vision. In this study, an interesting dual property of vanishing point is first introduced. Next, we point out that there also exists a dual property of vanishing line. With the dual vanishing point and vanishing line properties, some 3D intersection inference can be made based on their image lines. Two applications are given to illustrate the usage of the new results. The first one is to derive the 3D pose determination of a circle using two parallel image lines. The second one uses six specially designed 3D lines to adjust the cameras with respect to a fixture in a binocular vision system such that the resultant camera coordinate axes become parallel. (C) 1999 Pattern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.
A vision-based method for the circle pose determination with a direct geometric interpretation
(1999-12-01)
Chen, Z
;
Huang, JB
;
資訊工程學系
;
Department of Computer Science
A novel vision-based method for the circle pose determination is addressed. This method is based on two particular projected chords of a circle image. The first one is the projection of a circle chord which subtends the largest apex angle of the viewing cone for the circle image and the second one is the projection of a circle diameter whose backprojection plane bisects the above largest apex angle. This method is conceptually simple, since the circle center is the center point of the diameter chord and the circle orientation is given by the cross product of these two (directed) chords. We present theorems on the geometry of both the viewing cone and a reprojected circle image which are essential to the pose determination. We then give a pose determination method which is applicable under all viewing conditions. Experimental results illustrate the good performance of the method, when compared with other existing methods.
Automatic detection and tracking of human heads using an active stereo vision system
(2000-03-01)
Tang, CY
;
Chen, Z
;
Hung, YP
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
A new head tracking algorithm for automatically detecting and tracking human heads in complex backgrounds is proposed. By using an elliptical model for the human head, our Maximum Likelihood (ML) head detector can reliably locate human heads in images having complex backgrounds and is relatively insensitive to illumination and rotation of the human heads. Our head detector consists of two channels: the horizontal and the vertical channels. Each channel is implemented by multiscale template matching. Using a hierarchical structure in implementing our head detector, the execution time for detecting the human heads in a 512 x 512 image is about 0.02 second in a Spare 20 workstation (not including the time for image acquisition). Based on the ellipse-based ML head detector, we have developed a head tracking method that can monitor the entrance of a person, detect and track the person's head, and then control the stereo cameras to focus their gaze on this person's head. In this method, the ML head detector and the mutually-supported constraint are used to extract the corresponding ellipses in a stereo image pair. To implement a practical and reliable face detection and tracking system, further verification using facial features, such as eyes, mouth and nostrils, may be essential. The 3D position computed from the centers of the two corresponding ellipses is then used for fixation. An active stereo head has been used to perform the experiments and has demonstrated that the proposed approach is feasible and promising for practical uses.
Clustering by comparing regions of different density
(1996-01-01)
Lin, JC
;
Lin, JY
;
Chen, Z
;
資訊工程學系
;
Department of Computer Science
This paper presents a histogram-based clustering method that automatically determines the number of clusters in a set of data points. Input data are first partitioned into several rectangular blocks. The number of points in each block is determined, and the thirty percent of the blocks with the most points are marked to obtain a feature. Next, the forty percent of the blocks with the most points are marked to obtain a second feature. These two features are then compared to determine the number of clusters in the input data. The proposed clustering method is fast, and the data to be clustered do not need to be linearly separable. Experimental results are included.
Determination of feature correspondences in stereo images using a calibration polygon
(1997-09-01)
Chiu, JM
;
Chen, Z
;
Chuang, JH
;
Chia, TL
;
資訊科學與工程研究所
In this paper, we propose a novel approach to solving the vertex/edge correspondence problem for stereo images. Assume an object is placed on a calibration plate (C-plate) and two perspective views of them are given. The C-plate vertex correspondence, determined by cross-ratios, is used to reduce the search space for determining object vertex/edge correspondence. First, the correspondence of object edges lying on the C-plate (called the object base edges) is considered. This is because the subpolygons obtained from the division of the C-plate by the extended line of each of these base edges are viewpoint invariant, i.e. the cross-ratio of each of their vertices in different images will have equal value. Assume that at least one of the base edges is visible in each image. Since a visible object base edge has to be an object boundary edge in an image, only object boundary edges will be considered for the cross-ratios check for the associated subpolygons as well as other geometric constraint checks in determining the object base vertex/edge correspondence. These geometric constraints include (i) the position of the edges along the boundary of object faces, and (ii) the division of the C-plate vertices by the extended lines of these edges. Based on one of the determined corresponding base edge pairs, the correspondence of all other edges can be determined by constraints similar to (i). Finally, some additional viewpoint invariant measures about the object faces and the C-plates are introduced to resolve the ambiguity problem which may arise for specific viewpoints and some special object shapes. The proposed approach only needs 2-D image data and has no constraint on the number of vertices of an object face. Experimental results are presented for some polyhedral as well as curved objects. (C) 1997 Pattern Recognition Society. Published by Elsevier Science Ltd.
Fast octree construction endowed with an error bound controlled subdivision scheme
(2006-05-01)
Chou, HL
;
Chen, Z
;
資訊工程學系
;
Department of Computer Science
In the conventional octree construction method any grey octant is subdivided recursively until all its descendant octants are no longer grey. The subdivision process is executed no matter how small the white portion in the grey octant is. When there are many grey octants, each containing a fairly small white portion, then the huge increase in the resulting descendant nodes due to the subdivisions may cause the construction process to terminate due to an insufficient amount of memory. In such a case the subdivisions made are not worthwhile. In this paper, we shall make effective use of octant subdivision to improve the overall system performance. A new octree construction method is proposed with a novel subdivision strategy such that only those octants with a projection error exceeding a pre-specified error bound will be subdivided. Furthermore, we also present a fast way to compute the 2D projection of octant vertices and a new intersection test to reduce overall processing time. Computer simulations are conducted which show that the new method performs better than the conventional method in terms of memory space and computation time. Moreover, a theoretical analysis of the performance of the new method is included.
Identification of business forms using relationships between adjacent frames
(1996)
Lin, JY
;
Lee, CW
;
Chen, Z
;
交大名義發表
;
資訊科學與工程研究所
This paper proposes a method of using relationships between adjacent frames to identify business forms. First, we use the relations between adjacent frames in the business form to represent each form as two graphs. We then convert these graphs to two 1-D strings, which are used to perform matching. This approach enables forms to be identified quickly, and it places no restrictions on the layout of the business forms.
Obtaining base edge correspondence in stereo images via quantitative measures along C-diagonals
(1997-01-01)
Chuang, JH
;
Chiu, JM
;
Chen, Z
;
資訊工程學系
;
Department of Computer Science
We propose a novel approach to solving the object edge correspondence problem for stereo images. For an object placed on a calibration plate (C-plate), the proposed approach first obtains the correspondences of the object edges lying on the C-plate (the base edges) via quantitative measures of locations of the intersections of the extended line of the edges and the diagonals of the C-plate (C-diagonals) using cross ratios. The measures an viewpoint invariant for an object base edge, and are expressed in the number of image pixels. Special cases which need only coarse calculations, as well as those which require extra measures for additional point features, are also considered, The proposed approach requires 2-D image data only, and is robust in the presence of errors in the feature detection. Experimental results are presented for polyhedral and curved objects to demonstrate the effectiveness of the proposed approach. (C) 1997 Elsevier Science B.V.
Parallel distance transforms on a linear array architecture
(2002-04-30)
Chia, TL
;
Wang, KB
;
Chen, Z
;
Lou, DC
;
資訊科學與工程研究所
Distance transformation (DT) has been widely used for image matching and shape analysis. In this paper, a parallel algorithm for computing distance transformation is presented. First, it is shown that the algorithm has an execution time of 6N - 4 cycles, for an N x N image using a parallel architecture that requires [N/2] parallel processors. By doing so, the real time requirement is fulfilled and its execution time is independent of the image contents. In addition, a partition method is developed to process an image when the parallel architecture has a fixed number of processing elements (PEs); say two or more. The total execution time for an N x N image by employing a fixed number of PEs is 2[N-2/M + 2(M - 1)], when M is the fixed number of PEs. (C) 2002 Elsevier Science B.V. All rights reserved.
Parallel execution of a connected component labeling operation on a linear array architecture
(2003-03-01)
Wang, KB
;
Chia, TL
;
Chen, Z
;
Lou, DC
;
資訊科學與工程研究所
This work presents a novel parallel algorithm and architecture for finding connected components in an image. Simulation results indicate that the proposed algorithm has an execution time of N-2 +6N-4 cycles for an NxN image using an architecture containing 4 parallel processors. The proposed hardware can process a 128 x 128 image in 0.8574 ms and uses only 4 processors, compared to 0.85 ms and 128 processors for the work of Ranganathan et al. [14], and 94.6 ms and 16384 processors for the MPP [22]. Among the advantages of the novel architecture are modularity, expandability, regular data flow, and simple hardware. These properties are extremely desirable for VLSI implementations. Additionally, the execution time of the algorithm is independent of its image content; thus, it is quite flexible.
Pose determination of a cylinder using reprojection transformation
(1996-09-02)
Huang, JB
;
Chen, Z
;
Chia, TL
;
資訊科學與工程研究所
;
Institute of Computer Science and Engineering
Extracting useful and robust feature points from a cylinder image for 3D pose determination is a relatively difficult task, since there are no feature points or lines on the cylinder. In this study, a key vector from a cylinder image is found to yield a transformation. The cylinder is then ''reprojected'' from the original image to obtain a new image using the transformation. The new image reveals some important features capable of determining the 3D cylinder position robustly. The proposed method can yield an exact and analytical solution without the requirement of using any special marks, extra equipment or assumptions. Additionally, a problem in which both circles on the cylinder are not completely visible due to the viewing angle can be resolved by this method. Finally, experimental results are provided to verify the theory and show the robustness of our method.
Recognition of radicals in handwritten Chinese characters by means of problem reduction and knowledge guidance
(1996-09-01)
Cheng, RH
;
Lee, CW
;
Chen, Z
;
資訊工程學系
;
Department of Computer Science
Recognizing handwritten Chinese characters is a complex problem. We break the problem down into a series of sub-problems and concentrate on the radical recognition problem. The sub-problems are linked in a hierarchy of three layers: radical, radical sub-unit, and salient stroke. This problem reduction technique enables us to solve complex recognition problems effectively. We describe how to analyze and choose radical sub-units and salient strokes in order to recognize a set of radicals. We also construct two knowledge bases in the form of decision trees to guide the hypothesis and test procedures used to solve the recognition sub-problems. The problem reduction is done in a top-down fashion, while the problem solving process proceeds in a bottom-up fashion. The advantages of our method are described. The representation and organization of the two knowledge bases are explicitly described. The method is tested on two sets of handprinted characters using an TBM PC (486-33). The recognition rate is over 95.7% and the computer time needed to recognize a radical averages about 0.07 s. The experimental results indicate that our method effectively copes with a wide range of ordinary variations encountered in handwritten Chinese characters.