We present a method allowing a significant speed-up of the eigen-detection method (detection based on principle component analysis). We derive a formula for an upper bound on the class-conditional probability (or equi...
详细信息
We present a method allowing a significant speed-up of the eigen-detection method (detection based on principle component analysis). We derive a formula for an upper bound on the class-conditional probability (or equivalently a lower bound on the Mahalanobis distance) on which detection is based. Often, the lower bound of Mahalanobis distance (MD) reaches a preset threshold after computation of only a few eigen-projections. In this case the computation of MD can be immediately terminated. Regardless of the precise value of MD, the detection hypothesis (object from class /spl Omega/ is detected) can be rejected. While provably obtaining results identical to the standard technique, we achieved a two- to three-fold speed-up in face detection experiments on images from the CMU database.
Most algorithms for segmenting connected handwritten digit strings are based on the analysis of the foreground pixel distributions and the features on the upper/lower contours of the image. A new approach is presented...
详细信息
ISBN:
(纸本)0780344286
Most algorithms for segmenting connected handwritten digit strings are based on the analysis of the foreground pixel distributions and the features on the upper/lower contours of the image. A new approach is presented to segment connected handwritten two-digit strings based on the thinning of background regions. The algorithm first locates several feature points on the background skeleton of the digit image. Possible segmentation paths are then constructed by matching these feature points. With geometric property measures, these segmentation paths are ranked using fuzzy rules generated from a decision-tree approach. Finally, the top ranked segmentation paths are tested one by one by an optimized nearest neighbor classifier until one of these candidates is accepted based on an acceptance criterion. Experimental results on NIST special database 3 show that our approach can achieve a correct classification rate of 92.4% with only 4.7% of digit strings rejected, which compares favorably with the other techniques tested.
We study the problem of representing images within a multimedia Database Management System (DBMS), in order to support fast retrieval operations without compromising storage efficiency. To achieve this goal, we propos...
We study the problem of representing images within a multimedia Database Management System (DBMS), in order to support fast retrieval operations without compromising storage efficiency. To achieve this goal, we propose new image coding techniques which combine a wavelet representation, embedded coding of the wavelet coefficients, and segmentation of image-domain regions in the wavelet domain. A bitstream is generated in which each image region is encoded independently of other regions, without having to explicitly store information describing the regions. Simulation results show that our proposed algorithms achieve coding performance which compares favorably, both perceptually and objectively, to that achieved using state-of-the-art image/video coding techniques while additionally providing region-based support.
Smooth interpolants defined over tetrahedra are currently being developed for they have many applications in geography, solid modeling, finite element analysis, etc. In this paper, we will characterize a certain class...
详细信息
Smooth interpolants defined over tetrahedra are currently being developed for they have many applications in geography, solid modeling, finite element analysis, etc. In this paper, we will characterize a certain class of C-1 discrete tetrahedral interpolants with only C-1 data required. As special cases of the class characterized, we give two C-1 discrete tetrahedral interpolants which have concise expressions.
A central problem in stereo matching using correlation techniques lies in selecting the size of the search window. Small windows contain only a small number of data points, and thus are very sensitive to noise and the...
详细信息
Stereo computes the distance of objects, "their depth", from two images of two cameras using the triangulation principle. Points of imaged objects are mapped in different locations in the two stereo images. ...
详细信息
This paper is concerned with algorithms for removal additive noise from images. The proposed (α,ß)-trimmed mean filtering is suitable for application to real images corrupted by Gaussian, uniform and impulsive n...
详细信息
This paper is concerned with algorithms for removal additive noise from images. The proposed (α,ß)-trimmed mean filtering is suitable for application to real images corrupted by Gaussian, uniform and impulsive noise. The developed technique is a generalization of α-trimmed mean filter and have the same basic properties as rank-order filters. The actual performance of proposed technique was compared with that of average, median and midpoint filters and evaluated on noisy images by the error of restoration. The illustrative examples are given.
Stereo computation is just one of the vision problems where the presence of outliers cannot be neglected. Most standard algorithms make unrealistic assumptions about noise distributions, which leads to erroneous resul...
详细信息
Concept generalization under incomplete domain theory is a very important research aspect in artificial intelligence. Current multilayer perceptron and EBL (explanation based learning) approaches cannot deal with it e...
详细信息
ISBN:
(纸本)0780342534
Concept generalization under incomplete domain theory is a very important research aspect in artificial intelligence. Current multilayer perceptron and EBL (explanation based learning) approaches cannot deal with it effectively. We present a new method, hybrid multilayer perceptron/EBL approach for concept generalization, which can deal with concept generalization more effectively.
The paper presents an image registration method based on a two-dimensional Hopfield neural network, where the problem of image matching is treated with the minimization of the energy function of the Hopfield neural ne...
详细信息
ISBN:
(纸本)0818682035
The paper presents an image registration method based on a two-dimensional Hopfield neural network, where the problem of image matching is treated with the minimization of the energy function of the Hopfield neural network. The input data used for registration are the locations of the corner points extracted from the images. In order to improve and expedite the matching process, a fast block-based algorithm is put forward, together with the laboratory results obtained, which show the effectiveness of the algorithm.
暂无评论