We present methods for synthesizing 3D shape features on subdivision surfaces using multiscale procedural techniques. Multiscale synthesis is a powerful approach for creating surfaces with different levels of detail. ...
详细信息
ISBN:
(纸本)0769513301
We present methods for synthesizing 3D shape features on subdivision surfaces using multiscale procedural techniques. Multiscale synthesis is a powerful approach for creating surfaces with different levels of detail. Our methods can also blend multiple example multiresolution surfaces, including procedurally-defined surfaces as well as captured models.
this paper suggests an hierarchical region based representation to implement detection of shapes in gray-scale images. the Max-Tree structure has demonstrated to be useful for this purpose, offering a semantic objects...
详细信息
ISBN:
(纸本)0769522270
this paper suggests an hierarchical region based representation to implement detection of shapes in gray-scale images. the Max-Tree structure has demonstrated to be useful for this purpose, offering a semantic objects vision of the image, therefore, reducing the number of elements to process in relation to the pixel based representation. In this way, a particular thresholding information, associated with a specific shape analysis, can be used to determine the segmentation or filtering of regions of interest with lesser computational effort. For to demonstrate this technique, a generic application of detection of lines is proposed using the distance transform approach on each connected component from the depth searching in the mentioned tree.
this paper presents a new method for structuring and tracking of objects in video sequences. Our approach is based on image moments and the bsp-tree data structure. We use invariant properties of these moments to cons...
详细信息
ISBN:
(纸本)076951846X
this paper presents a new method for structuring and tracking of objects in video sequences. Our approach is based on image moments and the bsp-tree data structure. We use invariant properties of these moments to construct a bsp-tree and determine an ellipsis that approximates the object's shape. then, we employ this information to track objects frame by frame through the image sequence. the method works well for segmented images with a single object and we assume that the motion will not change abruptly.
this paper describes the development of a vision-based, nighttime navigation support system for automotive vehicles driving in semi-structured environments. the focus is on automatic reduction of glare produced by hig...
详细信息
ISBN:
(纸本)0769522270
this paper describes the development of a vision-based, nighttime navigation support system for automotive vehicles driving in semi-structured environments. the focus is on automatic reduction of glare produced by high beams from incoming vehicles and on generation of virtual lanes that are overlaid on the image where actual road markings should exist. To accomplish these goals, we develop video-based robust techniques that produce enhanced road images in real-time. the proposed techniques are validated by experiments with real video streams, which demonstrates the system's performance and accuracy under real conditions.
Active contours approach is a very used family of techniques in image analysis. this work presents a comparative study between two active contour approaches for image segmentation. Level sets method and deformable con...
详细信息
ISBN:
(纸本)0769522270
Active contours approach is a very used family of techniques in image analysis. this work presents a comparative study between two active contour approaches for image segmentation. Level sets method and deformable contours under B-spline representation are compared these image segmentation methods have different features and are difficult to compare in terms of performance, accuracy and initialization conditions. Both are implemented and a way to calculate the approximation error is developed As a conclusion of this work a theoretical comparison and a comparative characterization of the approximation error for each method are presented.
An image analysis method has been developed to segment yeast cells. Yeasts belong to the taxonomic group fungi and have been used on fuel and food industry, for example. the method is capable of segmenting yeast cells...
详细信息
An image analysis method has been developed to segment yeast cells. Yeasts belong to the taxonomic group fungi and have been used on fuel and food industry, for example. the method is capable of segmenting yeast cells based on Watershed Transform and space-scale analysis of the Tree of Critical Lakes. We analise hierarchical, geometric and gray-scale properties of the Tree of Critical Lakes. We show experimental results for one group of yeast images obtained from the School ofFood Engineering at Unicamp, Brazil. Comparison shows that the proposed method provides cells with area 10% lower than traditional approach. Moreover, this approach preserves the cells contour, an important feature because of the performance of bioreactors and other chemical processes are greatly influenced by their morphological character. (c) 2006 Elsevier B.V. All rights reserved.
this work presents a simple method to obtain fingerprint classification via directional map smoothing. the program was developed in C++ language, using statistical techniques. that strategy allows to extract attribute...
详细信息
ISBN:
(纸本)076951846X
this work presents a simple method to obtain fingerprint classification via directional map smoothing. the program was developed in C++ language, using statistical techniques. that strategy allows to extract attributes of an image and to increase the quality of the process due reduction of the noise, if compared withthe result obtained by mean other method exposed in literature. the experimental results, obtained by this technique, demonstrate that a better directional map is arranged.
this work exploits the resemblance between content-based image retrieval and image analysis with respect to the design of image descriptors and their effectiveness. In this context, two shape descriptors are proposed:...
详细信息
this work exploits the resemblance between content-based image retrieval and image analysis with respect to the design of image descriptors and their effectiveness. In this context, two shape descriptors are proposed: contour saliences and segment saliences. Contour saliences revisits its original definition. where the location of concave points was a problem, and provides a robust approach to incorporate concave saliences. Segment saliences introduces salience values for contour segments, making it possible to use an optimal matching algorithm as distance function. the proposed descriptors are compared with convex contour saliences, curvature scale space, and beam angle statistics using a fish database with 11,000 images organized in 1100 distinct classes. the results indicate segment saliences as the most effective descriptor for this particular application and confirm the improvement of the contour salience descriptor in comparison with convex contour saliences. (c) 2006 Elsevier B.V. All rights reserved.
this paper presents a novel and more successful learning based approach to extracting low level features in a retina-like (log-polar) image representation. the low level features (edges, bars, blobs and ends) are base...
详细信息
ISBN:
(纸本)0769513301
this paper presents a novel and more successful learning based approach to extracting low level features in a retina-like (log-polar) image representation. the low level features (edges, bars, blobs and ends) are based on Marr's primal sketch hypothesis for the human visual system [10]. the feature extraction process used a neural network that learns examples of the features in a window of receptive fields of the image representation. An architecture designed to encode the feature's class, position, orientation and contrast has been proposed and tested. Success depended on the incorporation of a function to normalise the feature's orientation and a PCA pre-processing module to produce better separation in the feature space.
Many intrinsically 2-dimensional visual signals can be effectively encoded in a I-D form. this simpler representation is well-suited to both pattern recognition and image retrieval tasks. this poster deals with contou...
详细信息
ISBN:
(纸本)076951846X
Many intrinsically 2-dimensional visual signals can be effectively encoded in a I-D form. this simpler representation is well-suited to both pattern recognition and image retrieval tasks. this poster deals with contour and texture, combined together in order to obtain an effective technique for content based image indexing. the data used for experimentally assessing CONTEXT were contours and textures from various application domains. the experiments reveal a high discriminating power which in turn yields a high perceived quality of the retrieval results.
暂无评论