image fidelity (inferred by the ability to discriminate between two images) and image quality (inferred by the preference for one image over another) are often assumed to be directly related. We investigated the relat...
详细信息
image fidelity (inferred by the ability to discriminate between two images) and image quality (inferred by the preference for one image over another) are often assumed to be directly related. We investigated the relationship between the perceived image fidelity and image quality of halftone textures. Subjects were asked to rank order a set of printed halftone swatches on the basis of smoothness. They were then asked to reduce the contrast of each pattern until it was at threshold, thus providing an estimate of the pattern's perceptual strength and its discriminability from a non-textured swatch. We found only a moderate correlation between image fidelity and image quality.
We present a parallel divide-and-conquer approach for detecting line segments on coarse grained architectures in general, and a particular solution on a hyper-pyramid structure. We use a method which extract directly ...
详细信息
We present a parallel divide-and-conquer approach for detecting line segments on coarse grained architectures in general, and a particular solution on a hyper-pyramid structure. We use a method which extract directly the segments from the image without quantification such as in the classical Hough transform. The proposed solution is implemented on an environment on top of PVM.
We introduce spherical triangular B-splines for closed shape representation, and discuss the shape reconstruction using our new model "winged B-snakes", which are deformable surfaces coupled with active edge...
详细信息
We introduce spherical triangular B-splines for closed shape representation, and discuss the shape reconstruction using our new model "winged B-snakes", which are deformable surfaces coupled with active edges and junctions. We show the results of using the spherical winged B-snakes for simultaneous surface reconstruction and feature detection from range images or scattered 3D data.
We present a method to obtain a partition of binary edge maps. The edge maps are partitioned into three subsets of active elements, according to the following classification: matched edge elements, smeared edge elemen...
详细信息
We present a method to obtain a partition of binary edge maps. The edge maps are partitioned into three subsets of active elements, according to the following classification: matched edge elements, smeared edge elements, and false edge elements. Based on this partition, figures of merit for the quantitative evaluation of edge maps can be derived.
This paper describes the problems and issues involved in developing artificial visual agents that watch moving objects and their interactions in real world situations. One objective of this work is to form conceptual ...
详细信息
This paper describes the problems and issues involved in developing artificial visual agents that watch moving objects and their interactions in real world situations. One objective of this work is to form conceptual descriptions that capture the behaviours of objects. To do this we use the dynamic scene context but here we go further and incorporate task context and simple learning of behavioural cues. The computational approach uses results from the VIEWS project. The issues concerned with extending computational vision to learn behavioural models and use attention are described for a surveillance system with "task-level control". This means that the visual processing is guided by both the current scene knowledge and the current surveillance task control policy.
This paper considers the characterization of three different image stabilization algorithms when used as a preprocessor for a computervision application. These algorithms vary in computational complexity, accuracy, a...
详细信息
This paper considers the characterization of three different image stabilization algorithms when used as a preprocessor for a computervision application. These algorithms vary in computational complexity, accuracy, and performance. The first algorithm (developed by the Army Research Laboratory (ARL)) is capable of image alignment to an accuracy of one pixel. Algorithms 2 and 3 (developed by the University of Maryland) are capable of full subpixel stabilization with respect to translation, rotation, and scale. The evaluation tools incorporated include mean square error of the output data set and the overall performance of an automatic target acquisition (ATA) system (developed at ARL) that uses the algorithms as a front-end preprocessor. We show that for the ARL ATA application, extremely accurate subpixel stabilization is a requirement for proper operation. Based on experiments, we conclude that algorithm 3 performs significantly better than the other two algorithms.
The growth of networked multimedia systems has magnified the need for image copyright protection. One approach used to address this problem is to add an invisible structure to an image that can be used to seal or mark...
详细信息
The growth of networked multimedia systems has magnified the need for image copyright protection. One approach used to address this problem is to add an invisible structure to an image that can be used to seal or mark it. These structures are known as digital watermarks. We describe two techniques for the invisible marking of images. We analyze the robustness of the watermarks with respect to linear and nonlinear filtering, and JPEG compression. The results show that our watermarks detect all but the most minute changes to the image.
Mosaic techniques have been used to combine two or more signals into a new one with an invisible seam, and with as little distortion of each signal as possible. Multiresolution representation is an effective method fo...
详细信息
ISBN:
(纸本)0780332598
Mosaic techniques have been used to combine two or more signals into a new one with an invisible seam, and with as little distortion of each signal as possible. Multiresolution representation is an effective method for analyzing the information content of signals, and it also fits a wide spectrum of visual signal processing and visual communication applications. The wavelet transform is one kind of multiresolution representations, and has found a wide variety of application in many aspects, including signal analysis, image coding, imageprocessing, computervision and etc. Due to its characteristic of multiresolution signal decomposition, the wavelet transform is used for the image mosaic by choosing the width of the mosaic transition zone proportional to the frequency represented in the band. Both 1-D and 2-D signal mosaics are described, and some factors which affect the mosaics are discussed.
We introduce a fully adaptive active contour model in which no parameters have to be set a priori or tuned by the user. It is based on elliptic Fourier contour description and on the minimum description length (MDL) p...
详细信息
We introduce a fully adaptive active contour model in which no parameters have to be set a priori or tuned by the user. It is based on elliptic Fourier contour description and on the minimum description length (MDL) principle. The proposed technique estimates all the observation model parameters (e.g., noise variances), the order of the contour description (number of Fourier coefficients), and the contour itself.
The ability to recognize edges of an object fast and accurately is a fundamental goal in computervision and imageprocessing. The facility is important in automatic manufacturing environment in industry. This paper d...
详细信息
The ability to recognize edges of an object fast and accurately is a fundamental goal in computervision and imageprocessing. The facility is important in automatic manufacturing environment in industry. This paper defines and investigates an efficient method of edge extraction of image using directional tracing algorithm. Based on the approach of linear feature extraction presented by R. Nevatia and K.R. Babu (1980), our algorithm is developed for the general situation of edge extraction. The technique employs the basic and intuitive principle that an edge pixel should possess local maximal gradient but no more domain information or knowledge is required. An effective and robust tracing strategy is proposed. Our algorithm has been tested on different kinds of images, and the results are satisfactory.
暂无评论