The problem in document analysis of the conversion of technical drawings into descriptions that are appropriate to be manipulated by CAD systems is addressed. Ideas are presented for the construction of 3-D-geometric ...
详细信息
The problem in document analysis of the conversion of technical drawings into descriptions that are appropriate to be manipulated by CAD systems is addressed. Ideas are presented for the construction of 3-D-geometric models from noisy and inaccurate images of simple workpiece drawings. The main ideas are the search for instances of primitive solids in the image, the incremental construction of partial descriptions based on these primitives, and their manipulation to achieve consistency with the image data. The implementation of the method is described, limitations are discussed, and the possible future development of the approach is outlined.
Two kinds of Fourier shape descriptors (FDs) are considered in this article: ZR (defined by Zahn and Roskies) and G (defined by Granlund). The first section of this article studies ZR descriptors. Four modifications o...
详细信息
Two kinds of Fourier shape descriptors (FDs) are considered in this article: ZR (defined by Zahn and Roskies) and G (defined by Granlund). The first section of this article studies ZR descriptors. Four modifications of ZR descriptors are proposed. These new descriptors are based on the step signature, smoothed signature, linearized smoothed signature, and the curvature function, respectively. The amplitudes of FDs are shown to be invariant under rotations, translations, changes in size, mirror reflections, and shifts in the starting point. In all cases the reconstruction accuracy in terms of the number of FDs is studied, resulting in approximation error bounds. An efficient reconstruction formula based on the composite rectangle integration method is proposed. This method also provides polygonal approximation for arbitrary contours. In the second section of this article, the theoretical results are illustrated with examples and verified in numerical experiments involving digitized patterns. In the same experiments the performances of ZR and G descriptors are compared.
An important impediment to to the use of machinevision systems in industrial applications is the volume of image data to be processed in relation to the execution speed of the processor. An approach to this problem i...
详细信息
An important impediment to to the use of machinevision systems in industrial applications is the volume of image data to be processed in relation to the execution speed of the processor. An approach to this problem is to reduce the data flow and increase the information density by hardware pipelined preprocessing so reducing the final processing requirements to manageable proportions. This paper outlines the design and implementation of an image pre-processor previously constructed in Fast Fairchild Shottky TTL MSI components and gives a brief account of the design on field programmable gate array (FPGA). Preliminary results of the real-time application of this system to the laboratory inspection of large numbers of machined parts are described.
In this paper, we describe how the recently developed technique of polynomial transforms can be applied in image coding. Two related algorithms, based on two-dimensional (2D) and one-dimensional (1D) polynomial approx...
详细信息
ISBN:
(纸本)0819402389
In this paper, we describe how the recently developed technique of polynomial transforms can be applied in image coding. Two related algorithms, based on two-dimensional (2D) and one-dimensional (1D) polynomial approximations respectively, are presented. It is subsequently shown how the latter coding algorithm could be improved by including ideas from the field of computer vision.
This paper describes the performance testing of a real-time image pre-processor designed for high-speed edge detection but working in a typical real-time environment with standard vidicon camera input and LSI circuit ...
详细信息
This paper describes the performance testing of a real-time image pre-processor designed for high-speed edge detection but working in a typical real-time environment with standard vidicon camera input and LSI circuit technology. Performance is assessed by real-time measurement of noise behaviour compared to processing output of images containing well defined image information content. The extraction of useful image data is achieved by a non-linear thresholding technique which has been show to yield useful practical results in a number of different applications through an ad-hoc experimental approach to the establishment of the appropriate thresholding algorithm. This paper illustrates the systematic establishment of the non-linear algorithm resulting real-time performance of the optimal threshold function for an input image with a simple sample object. The use of the method for inspection applications is discussed briefly.
A description is given of an automatic microorganism recognition and identification method, using imageprocessing based on an expert system, for application in a micro-manipulator system. Since microorganisms under a...
详细信息
A description is given of an automatic microorganism recognition and identification method, using imageprocessing based on an expert system, for application in a micro-manipulator system. Since microorganisms under a microscope all appear different, are moving, and have time-dependent shapes, the conventional pattern-matching method used for industrial products can be applied. The proposed method is based on segmentation of the microorganisms, taking advantage of characteristic segment features that are independent of individual size and length. Complicated shapes are divided into simple-shaped segments such as lines, circles, ovals, etc. The relation between segments is described in the database of the expert system. Microorganisms can then be expressed simply by a set of segments and so that their individual differences can be avoided. Tracking the movement of a microorganism by its image is also shown.
This paper presents the theory and realisation of the Multiple Description Segment (MDS) technique used in a commercially available integrated vision environment. It was applied to a sequence of scenes of an office co...
详细信息
ISBN:
(纸本)0819402389
This paper presents the theory and realisation of the Multiple Description Segment (MDS) technique used in a commercially available integrated vision environment. It was applied to a sequence of scenes of an office corridor. The preliminary results of the classification have shown confident decision by the classifier on well defined objects and objects with high complexity and noise interference. Objects that are very similar (but different in functionality) were also recognised correctly. The merits and pitfalls of the technique and the future direction of development will be discussed.
The use of depth and motion for three-dimensional scene analysis requires a system that can accurately and reliably measure image motion. Such measurements can be obtained by tracking the position of edge lines in a c...
详细信息
The use of depth and motion for three-dimensional scene analysis requires a system that can accurately and reliably measure image motion. Such measurements can be obtained by tracking the position of edge lines in a closely spaced monocular sequence of images. The reliability of such a technique is greatly enhanced by maintaining a model of image flow composed of the position and velocity of tokens constructed from edge lines. Algorithmic aspects of such a system are presented, and it is shown how the complexity can be reduced in order to optimize hardware. Results on real data are provided, and architectural aspects are discussed.
A method of boundary representation based on equation transformation codes is described. Features derived from the encoded boundary are used to recognize objects in a model-based scheme. Results indicate that the algo...
详细信息
A method of boundary representation based on equation transformation codes is described. Features derived from the encoded boundary are used to recognize objects in a model-based scheme. Results indicate that the algorithm works well even when there is occlusion.
The three-dimensional shape of a polyhedron is reconstructed from its perspectively projected image by finding rectangular corners. If a corner is known to be rectangular, its three-dimensional configuration is analyt...
详细信息
The three-dimensional shape of a polyhedron is reconstructed from its perspectively projected image by finding rectangular corners. If a corner is known to be rectangular, its three-dimensional configuration is analytically computed, and hence the surface gradients of the faces incident to it are also determined. In order to decide which corners are rectangular, the rectangularity heuristics are introduced, i.e., corners are assumed to be rectangular as long as no inconsistency arises as a result. An optimization technique is presented in order to cope with the problem of insufficiency of information and inconsistency due to error and noise.
暂无评论