The proceedings contain 54 papers. The special focus in this conference is on Oral Presentations and Long Posters. The topics include: Blur and disorder;applications of locally orderless images;scalespace technique f...
ISBN:
(纸本)354066498X
The proceedings contain 54 papers. The special focus in this conference is on Oral Presentations and Long Posters. The topics include: Blur and disorder;applications of locally orderless images;scalespace technique for word segmentation in handwritten documents;fast geodesic active contours;morphing active contours;unfolding the cerebral cortex using level set methods;reconciling distance functions and level sets;computation of ridges via pullback metrics from scalespace;the maximal scale ridge - incorporating scale into the ridge definition;detection of critical structures in scalespace;qualitative multi-scale feature hierarchies for object tracking;riemannian drums, anisotropic curve evolution and segmentation;an active contour model without edges;a compact and multiscale image model based on level sets;morphological scalespace and mathematical morphology;scale-space from a level lines tree;morphological scale-space representation with levelings;numerical solution schemes for continuous-scale morphology;scale-space properties of regularization methods;an adaptive finite element method for large scale image processing;a scale-space approach to nonlocal optical flow calculations;scales in natural images and a consequence on their BV norm;anisotropic smoothing using local image statistics;the hausdorff dimension andscale-space normalisation of natural images;lattice boltzmann models for nonlinear diffusion filtering;geometric-variational approach for color image enhancement and segmentation;a level set model for image classification;calculations on critical points under gaussian blurring;region tracking on surfaces deforming via level-sets methods;geometric multiscale representation of numerical images and multiscale morphological segmentations based on watershed, flooding, and eikonal PDE.
Besides small vessel target detection in maritime images, other computer-vision-related challenges also remain-primarily associated with limited data concerning small targets as they are often obscured by larger objec...
详细信息
ISBN:
(纸本)9798400711732
Besides small vessel target detection in maritime images, other computer-vision-related challenges also remain-primarily associated with limited data concerning small targets as they are often obscured by larger objects or background noise. Therefore, Forest-Transformer-based models tend to struggle with the size of specific datasets impacting on an effective response to targets of various orientation andscales. Hence, some models need to work with more agility and less cumbersome to address these issues. In this respect, this paper proposes exciting two methods improving ocean craft detection and segmentation significantly. The first is that the lightweight YOLO v9 architecture is integrated into the SliceBoost framework using programmable gradient information to minimize loss of critical information during feature extraction. The second is the vision Mamba model that further augments spatial awareness by embedding position information into a newly constructed bidirectional state-space model that enhances contextual modeling abilities. To adept at the modeling of small vessel detection in infrared maritime images, it combines the local feature processing efficiency of convolution-only architectures with the global knowledge aberration capabilities of Transformer-based architectures. Experimental results show significant improvements in both detection and processing efficiency making real-time small-target detection possible within varying maritime applications. Furthermore, the paper discusses possible future maritime target recognition applications of these two techniques. The source code is available at the following link: https://***/CaiGuoHui123/SPSDet.
In this work we describe the possible transitions for the hierarchical structure that describes air image in Gaussian scalespace. Until now, this tree structure has only been used for topological segmentation. In ord...
详细信息
ISBN:
(纸本)9783642022555
In this work we describe the possible transitions for the hierarchical structure that describes air image in Gaussian scalespace. Until now, this tree structure has only been used for topological segmentation. In order to perform image matching and retrieval tasks based on this structure, one needs to know which transitions are allowed when the structure is changed under influence of one control parameter. We present a list of such transitions, enabling tree edit distance operations.
We investigate the maximum likelihood metameres of local pure 2(nd) order structure in natural images. Using the shape index, we re-parameterise the 2(nd) order structure and gain a one-parameter index which offers a ...
详细信息
ISBN:
(纸本)9783540728221
We investigate the maximum likelihood metameres of local pure 2(nd) order structure in natural images. Using the shape index, we re-parameterise the 2(nd) order structure and gain a one-parameter index which offers a qualitative description of local pure 2(nd) order image structure. Inspired by Koenderink and previous work within Geometric Texton Theory the maximum likelihood metameres are calculated for a quantised version of the shape index. Results are presented and discussed for natural images, Gaussian noise images, and Brownian or pink noise images. Furthermore, we present statistics for the shape index, principal direction, and curvedness of natural images. Finally, the results are discussed in the terms of their applicability to Geometric Texton Theory.
This paper shows a potential use of scalespace for statistical validation of watershed regions of a greyscale image. The watershed segmentation has difficulty in distinguishing valid watershed regions associated with...
详细信息
ISBN:
(纸本)9783642022555
This paper shows a potential use of scalespace for statistical validation of watershed regions of a greyscale image. The watershed segmentation has difficulty in distinguishing valid watershed regions associated with real structures of the image from invalid random regions due to background noise. In this paper, a hierarchy of watershed regions is established by following merging process of the regions in a Gaussian scalespace. The distribution of annihilation scales (lives) of the regional minima is investigated to statistically judge the regions as being valid or not. Recursive validation using the hierarchy prevents oversegmentation due to the randomness.
In this paper, a multi-scale vectorial total variation model for color image restoration is introduced. The model utilizes a spatially dependent regularization parameter in order to preserve the details during noise r...
详细信息
ISBN:
(纸本)9783642022555
In this paper, a multi-scale vectorial total variation model for color image restoration is introduced. The model utilizes a spatially dependent regularization parameter in order to preserve the details during noise removal. The automated adjustment strategy of the regularization parameter is based on local variance estimators combined with a confidence interval technique. Numerical results on images are presented to demonstrate the efficiency of the method.
We propose several very fast;algorithms to restore jittered digital video frames (their rows are shifted) in one iteration. The restored row shifts minimize non-smooth and possibly non-convex local criteria applied on...
详细信息
ISBN:
(纸本)9783642022555
We propose several very fast;algorithms to restore jittered digital video frames (their rows are shifted) in one iteration. The restored row shifts minimize non-smooth and possibly non-convex local criteria applied on the second-order differences between consecutive rows. We introduce specific error measures to assess the quality of dejittering. Our algorithms, are designed for gray-value, color and noisy images. Some of them can be considered as parameter-free. They outperform by far the existing algorithms both in quality and in speed. They are a crucial step towards real-time dejittering of digital video.
We present a generative model based method for recovering both the shape and the reflectance of the surface(s) of a scene from multiple images, assuming that illumination conditions are known in advance. Based on a va...
详细信息
ISBN:
(纸本)9783642022555
We present a generative model based method for recovering both the shape and the reflectance of the surface(s) of a scene from multiple images, assuming that illumination conditions are known in advance. Based on a variational framework and via gradient descents, the algorithm minimizes simultaneously and consistently a global cost functional with respect to both shape and reflectance. Contrary to previous works which consider specific individual scenarios, Our method applies to a number of scenarios - mutiview stereovision, multiview photometric stereo, and multiview shape from shading. In addition, our approach naturally combines stereo, silhouette and shading cues in a single framework and, unlike most previous methods dealing with only Lambertian surfaces, them proposed method considers general dichromatic surfaces.
variational approaches to correspondence problems such as stereo or optic flow have now been studied for more than 20 years. Nevertheless, only little attention has been paid to a subtle numerical approximation of der...
详细信息
ISBN:
(纸本)9783642022555
variational approaches to correspondence problems such as stereo or optic flow have now been studied for more than 20 years. Nevertheless, only little attention has been paid to a subtle numerical approximation of derivatives. In the area of numerics for hyperbolic partial differential equations (HDEs) it is, however, well-known that such issues can be crucial for obtaining favourable results. In this paper we show that the use of hyperbolic numerics for variational approaches can lead to a significant quality gain in computational results. This improvement can be of the same order as obtained by introducing better models. Applying our novel scheme within existing variational models for stereo reconstruction and optic flow, we show that this approach can be beneficial for all variational approaches to correspondence problems.
We analyze the rate in which image details are suppressed as a function of the regularization parameter, using first order Tikhonov regularization, Linear Gaussian scalespace and Total Variation image decomposition. ...
详细信息
ISBN:
(纸本)9783642022555
We analyze the rate in which image details are suppressed as a function of the regularization parameter, using first order Tikhonov regularization, Linear Gaussian scalespace and Total Variation image decomposition. The squared L-2-norm of the regularized solution and the residual are Studied as a function of the regularization parameter. For first order Tikhonov regularization it is shown that the norm of the regularized solution is a convex function, while the norm of the residual is not a concave function. The same result holds for Gaussian scalespace when the parameter is the variance of the Gaussian, but may fail when the parameter is the standard deviation. Essentially this imply that the norm of regularized solution can not be used for global scale selection because it does not contain enough information. An empirical study based on synthetic images as well as a database of natural images confirms that the squared residual norms contain important scale information.
暂无评论