This study presents a suboptimal boundary estimation algorithm for noisy images which is based upon an optimal maximum likelihood problem formulation. Both the maximum likelihood formulation and the resulting algorith...
详细信息
This study presents a suboptimal boundary estimation algorithm for noisy images which is based upon an optimal maximum likelihood problem formulation. Both the maximum likelihood formulation and the resulting algorithm are described in detail, and computational results are given. In addition, the potential power of the likelihood formulation is demonstrated through the presentation of three simple but insightful analyses of algorithm performance.
A syntactic method for the analysis of time-varying (TV) images is proposed. The problem is first analyzed through the use of pattern deformation model. The TV phenomena is then interpreted in terms of a language tran...
详细信息
A syntactic method for the analysis of time-varying (TV) images is proposed. The problem is first analyzed through the use of pattern deformation model. The TV phenomena is then interpreted in terms of a language translation schema. Furthermore, theorems are proved that some TV patterns can be formulated as a Context-Free Programmed Language (CFPL). Thus the existing literature on error-correcting syntax analysis can be applied to the TV pattern analysis problem. Tree grammar translation is proposed to handle more complex TV patterns.
The algorithm presented is a direct approach to the problem of determining a linear discriminant function which minimizes total classification error in a two-class training set. Discriminant function coefficients are ...
详细信息
The algorithm presented is a direct approach to the problem of determining a linear discriminant function which minimizes total classification error in a two-class training set. Discriminant function coefficients are determined by a search technique which is based on two theorems derived from convexity theory. The number of iterative steps in the search sequence has a theoretical upper bound which is strictly less than the number required for all enumeration, and applications of the algorithm to experimental data have established its tractability under nontrivial conditions. Although the training set patterns are initially assumed to be in general position, this restriction is essentially eliminated by an extension of the algorithm which is briefly discussed.
PASM, a multimicroprocessor system being designed at Purdue Univeristy for imageprocessing and patternrecognition, is described. This system can be dynamically reconfigured to operate as one or more independent SIMD...
详细信息
PASM, a multimicroprocessor system being designed at Purdue Univeristy for imageprocessing and patternrecognition, is described. This system can be dynamically reconfigured to operate as one or more independent SIMD and/or MIMD machines. The functions that the PASM operating system will perform are discussed, demonstrating how it will handle a variety of types of imageprocessing tasks. Examples of how PASM will improve computational speeds in comparison to conventional computers are presented. In particular, smoothing, histogram, and two-dimensional FFT algorithms are analyzed.
A description is given of a technique for generating a skeleton of a ribbon-like object using sequential data for all or part of the boundary. It shows how one may use local geometric information derived from the cont...
详细信息
A description is given of a technique for generating a skeleton of a ribbon-like object using sequential data for all or part of the boundary. It shows how one may use local geometric information derived from the contour to aid in the generation of a skeleton. For contours or curves of length n, this may be accomplished with a computation time of order n, while previous algorithms generally require order n**2 and require a two-dimensional matrix for their working representation.
Many applications in picture processing require accurate slope estimation. A brief survey of these applications is followed by discussion of the Sobel edge detection operator. The Sobel operator is very fast and simpl...
详细信息
Many applications in picture processing require accurate slope estimation. A brief survey of these applications is followed by discussion of the Sobel edge detection operator. The Sobel operator is very fast and simple;however, it is inherently biased as a function of true slope value and displacement of the true edge from center of the processing window. An iterative version of the operator is presented which reduces this bias with only a moderate increase in required processing.
A segmentation method for pap smear images is proposed to locate cellular boundary based on the stability of cellular boundary, and to locate nucleus based on the threshold deciding from the modified histogram. It emp...
详细信息
A segmentation method for pap smear images is proposed to locate cellular boundary based on the stability of cellular boundary, and to locate nucleus based on the threshold deciding from the modified histogram. It emphasizes the selection of the thresholds for nuclear and cytoplasmic boundary tracking. The authors also look into the problem of touching cells by using shape information of cellular boundary and texture information in the touching area. The proposed method, as applied to 11 scenes which contain 19 cells of different classes, works satisfactorily without any human interactions.
Segmentation of monochrome images to obtain boundaries of the object is an important problem in scene analysis. An algorithm is described for locating object boundaries from an image of objects. The algorithm, called ...
详细信息
Segmentation of monochrome images to obtain boundaries of the object is an important problem in scene analysis. An algorithm is described for locating object boundaries from an image of objects. The algorithm, called Global-Local-Edge-Coincidence (GLEC) uses both local and global edge information to select a stable set of object boundaries. Significantly improved results are shown in several examples including blocks, building and aerial photograph. The significance of this algorithm is that the boundaries of objects can often be located from a single image.
A method is described that generates 3D descriptions of arbitrarily-curved surfaces of physical objects by imageprocessing and photogrammetric techniques. A surface that is to be described is illuminated by a pattern...
详细信息
A method is described that generates 3D descriptions of arbitrarily-curved surfaces of physical objects by imageprocessing and photogrammetric techniques. A surface that is to be described is illuminated by a pattern of lines. The 3D shape of the surface causes the pattern to be distorted. The illuminated parts of the surface are photographed from two or more vantage points and the resulting images are digitized. The 2D perspective projection of the surface-distorted line pattern is extracted from each digitized image. Camera calibration marks are also extracted and the position of the camera is determined for each image. Two or more projections of the line pattern are then matched and the line pattern projected on the surface is reconstructed into 3D.
A theory is proposed for the structure of edges in a two dimensional visual field. This theory consists of models of ideal edges, and of distortion criteria for the evaluation of the accuracy with which a visual field...
详细信息
A theory is proposed for the structure of edges in a two dimensional visual field. This theory consists of models of ideal edges, and of distortion criteria for the evaluation of the accuracy with which a visual field fits this model. In this formulation the edge is specified by a curve described by a pair of parametric equations, and the line or edge detection problem is solved by minimizing a cost function over the set of curves.
暂无评论