The main purpose of imageprocessing and analysis is, often, to segment the image into objects in order to analyze the geometrical properties and the structure of the objects and recognize them. The analysis of the ge...
详细信息
ISBN:
(纸本)0818692162
The main purpose of imageprocessing and analysis is, often, to segment the image into objects in order to analyze the geometrical properties and the structure of the objects and recognize them. The analysis of the geometric objects must be quantitative, because only such an analysis and description of the geometric objects can provide a coherent mathematical framework for describing the spatial organization. The quantitative description of geometrical structures is the objective of mathematical morphology. So far, the use of such framework has allowed the development of a class of algorithms to deal with binary and grayscale images. Multi-channel processing has been of growing interest recently, especially in color imageprocessing. In this paper, the extension of mathematical morphology to color images by treating multi-channel data as vectors are presented. An overview of several approaches to this extension using different ordering relations is also provided. Basic vector morphological algorithms are introduced based on reduced ordering. We also present experimental results and show that these algorithms can be easily mapped into the generic-programming framework.
Reduced pyramids, i.e pyramids without analysis filters are known to produce excellent results when used for lossless signal and image compression. The present paper introduces a novel method for the construction of r...
详细信息
ISBN:
(纸本)0818692162
Reduced pyramids, i.e pyramids without analysis filters are known to produce excellent results when used for lossless signal and image compression. The present paper introduces a novel method for the construction of reduced pyramids with synthesis filters adapted to the orientation of local image features. It is shown that this construction produces more accurate interpolation than conventional reduced pyramids. It also has the added advantage of producing lossy replicas of the original, which, at lower resolutions lack undesirable artifacts near edges. Experimental evaluation shows that the resulting method produces results comparable and often superior to those of the most efficient known algorithms for progressive lossless image coding.
This paper describes a method to correct lens distortion for digital cameras. A simple but effective distortion model using local spatial transformation will be used. In addition, the effect of effective focal length ...
详细信息
ISBN:
(纸本)0818692162
This paper describes a method to correct lens distortion for digital cameras. A simple but effective distortion model using local spatial transformation will be used. In addition, the effect of effective focal length on the distortion model will be considered, and we will discuss how to introduce this factor to enhance the model. Finally we have some theoretical analysis and experiments demonstrating that the correction improves the accuracy of image stitching.
This paper investigates computer vision techniques for three-dimensional reconstruction having as source of information non-professional images captured from cameras of arbitrary position and orientation. The main pur...
详细信息
ISBN:
(纸本)0818692162
This paper investigates computer vision techniques for three-dimensional reconstruction having as source of information non-professional images captured from cameras of arbitrary position and orientation. The main purpose is to compose a methodology for 3D reconstruction, aiming specially at its use in the documentation of historic buildings and monuments through their three-dimensional modeling, For being particularly suitable, the technique of shape from stereo is then more closely analysed. Laboratory experiments, with controlled environment and objects are carried out in order to study the influence of some elements in the quality of reconstruction. A practical application with actual photographs of old buildings is presented, and final conclusions are drawn.
This paper describes Interface, a real-time facial animation system. The system uses a simple set of pre-modeled facial expressions to create a wide range of emotions, mouth positions and complete head movements. The ...
详细信息
ISBN:
(纸本)0818692162
This paper describes Interface, a real-time facial animation system. The system uses a simple set of pre-modeled facial expressions to create a wide range of emotions, mouth positions and complete head movements. The animation is done through layered groups of actions that combined gives the virtual actor complete freedom to perform different actions independently and simultaneously, without the intervention of the animator. Due to its software structure, Interface can be used as a "render plug-in" for an artificial intelligence program or a script-based animation system. The system was successfully implemented using the languages Java and VRML, and can be executed through the Internet in PCs or workstations with a Java-enable Web browser. Applications that can benefit from this system are interface agents, virtual reality systems and animation software, among others.
This paper describes a technique developed to analyze and recognize paper based electronic circuit diagram images. One of the key aspects of this recognition system relies on the identification of feature points used ...
详细信息
ISBN:
(纸本)0818692162
This paper describes a technique developed to analyze and recognize paper based electronic circuit diagram images. One of the key aspects of this recognition system relies on the identification of feature points used to guide a new line vectorization method developed for circuit diagram images. The feature points are also used as a discriminate element in the symbol classification process. Once the lines are vectorized, the open symbols are first identified by a syntax-based classifier, followed by the identification of the loop-based symbols using a supervised parametric statistical classifier. We have demonstrated the successful of the proposed system by implementing and testing it on various real images of circuit diagrams.
The reconstruction of a 3D model from range images can be conveniently split into two stages. The first stage consists basically in the extraction of geometrical information, e.g. the depth and the orientation of imag...
详细信息
ISBN:
(纸本)0818692162
The reconstruction of a 3D model from range images can be conveniently split into two stages. The first stage consists basically in the extraction of geometrical information, e.g. the depth and the orientation of image points, and the second stage concentrates on the transformation of these data to a 3D model. In this paper we focus on the second stage. Our approach is based on a deformable model, which offers more flexibility to mold a 3D model according to several geometrical constraints. Moreover, instead of looking for a global functional that carries out the deformation, we opted for establishing discrete local correspondences between 3D surface point data extracted from images and 3D model points.
This work reports the development of a versatile framework allowing the characterization and analysis of computer vision techniques as well as their applications to biological shapes, with attention focused on neural ...
详细信息
ISBN:
(纸本)0818692162
This work reports the development of a versatile framework allowing the characterization and analysis of computer vision techniques as well as their applications to biological shapes, with attention focused on neural cells. The proposed framework has been implemented within the Sigma ynergos system, a powerful imaging laboratory that includes, among other features, tools for performance assessment of computer vision techniques, image databases, real-time processing by using distributed systems and interface with the Internet. The motivations for the development of such a framework: (i) the importance of biological shape analysis;(ii) its potential as an effective tool for the systematic assessment of imageprocessing and analysis techniques;and (iii) the possibility of conducting extensive characterizations of biological shapes. The paper describes an experiment to assess multiscale shape features for complexity characterization, which have been adopted for the classification of two types of ganglion neural cells (cat), namely alpha and beta. This experiment involves: (1) a training stage where the k-means clustering algorithm learns the prototypes of each class from the database;(2) the neurons in the database are classified;(3) the classification results are compared to the original classes;and (4) the number of misclassifications is determined. The genetic algorithm is used as a means of effectively investigating the N-dimensional spaces defined by the parameter configurations.
The problem of matching two images can be posed as finding a displacement field which assigns each point of the reference image to a point in the test image. In this paper we present an iterative algorithm to estimate...
详细信息
ISBN:
(纸本)0818692162
The problem of matching two images can be posed as finding a displacement field which assigns each point of the reference image to a point in the test image. In this paper we present an iterative algorithm to estimate the probability density function relating the intensity distribution of two MR scanners, based on the topological constraints embedded in the elastic matching model. The set of images used as input for the algorithm is the Harvard Atlas. The density estimation resulting from this method is compared with two other algorithms which do not assume any prior information about the media being imaged. The results show that the density estimation obtained with the elastic matching approach produce more realistic deformed images and is suitable to represent MR sensor models.
This study presents a technique for quantitative enhancement of both linear attenuation coefficients and visual features of maps obtained from multiple energies using tomographic systems. The quality of images obtaine...
详细信息
ISBN:
(纸本)0818692162
This study presents a technique for quantitative enhancement of both linear attenuation coefficients and visual features of maps obtained from multiple energies using tomographic systems. The quality of images obtained by computerized X-ray tomography in multiple energies is limited due to the number of noise pixels in the images. The method developed here uses enhancement techniques with adaptative filtering that aims for results comparable to those obtained by single energy computerized tomography. The improvement achieved in the processed images confirms that this method is suitable for analyses of multiple energies with loss of pixel to pixel information, where information on density and material humidity assayed tomographically can be obtained accurately in the acceptable error range of single energy tomography.
暂无评论