Showing papers on "Edge detection published in 1983"

Finding Edges and Lines in Images

Abstract: : The problem of detecting intensity changes in images is canonical in vision. Edge detection operators are typically designed to optimally estimate first or second derivative over some (usually small) support. Other criteria such as output signal to noise ratio or bandwidth have also been been argued for. This thesis is an attempt to formulate a set of edge detection criteria that capture as directly as possible the desirable properties of an edge operator. Variational techniques are used to find a solution over the space of all linear shift invariant operators. The first criterion is that the detector have low probability of error i.e. failing to mark edges or falsely marking non-edges. The second is that the marked points should b The third criterion is that there should be low probability of more than one response to a single edge. The technique is used to find optimal operators for step edges and for extended impulse profiles (ridges or valleys in two dimensions). The extension of the one dimensional operators to two dimensions is then discussed. The result is a set of operators of varying width, length and orientation. The problem of combining these outputs into a single description is discussed, and a set of heuristics for the integration are given. (Author)

On Edge Detection of X-Ray Images Using Fuzzy Sets

Abstract: The effectiveness of the theory of fuzzy sets in detecting different regional boundaries of X-ray images is demonstrated. The algorithm includes a prior enhancement of the contrast among the regions (having small change in gray levels) using the contrast intensification (INT) operation along with smoothing in the fuzzy property plane before detecting its edges. The property plane is extracted from the spatial domain using S, ? and (1 ?) functions and the fuzzifiers. Final edge detection is achieved using max or min operator. The system performance for different parameter conditions is illustrated by application to an image of a radiograph of the wrist.

Image sequence processing and dynamic scene analysis

Abstract: I Overview.- Overview on Image Sequence Analysis.- Dynamic Scene Analysis.- II Image Sequence Coding.- Recursive Motion Compensation: A Review.- The Differential Method for Image Motion Estimation.- Edges in Visual Scenes and Sequences: Application to Filtering, Sampling and Adaptive DPCM Coding.- Movement-Compensated Interframe Prediction for NTSC Color TV Signals.- Coding of Colour TV Signals with 34 MBit/s Transmission Rate.- Analysis of Different Displacement Estimation Algorithms for Digital Television Signals.- An Adaptive Gradient Approach to Displacement Estimation.- Motion Parameter Estimation in TV-Pictures.- Image Sequence Coding Using Scene Analysis and Spatio-Temporal Interpolation.- Two Motion Adaptive Interframe Coding Techniques for Air to Ground Video Signals.- Motion Estimation in a Sequence of Television Pictures.- Comparative Study Between Intra- and Inter-Frame Prediction Schemes.- A Narrow-Band Video Communication System for the Transmission of Sign Language Over Ordinary Telephone Lines.- Classification and Block Coding of the Frame Difference Signal.- Histograms of Image Sequence Spectra.- III Scene Analysis and Industrial Applications.- Determining 3-D Motion and Structure of a Rigid Body Using Straight Line Correspondences.- Comparison of Feature Operators for use in Matching Image Pairs.- Displacement Estimation for Objects on Moving Backgroud.- Linear Filtering in Image Sequences.- Photometric Stereo For Moving Objects.- On the Selection of Critical Points and Local Curvature Extrema of Region Boundaries for Interframe Matching.- Image Segmentation Considering Properties of the Human Visual System.- A Fast Edge Detection Algorithm Matching Visual Contour Perception.- Image Sequence Analysis for Target Tracking.- Track Acquisition of Sub-Pixel Targets.- A Pre-Processor for the Real-Time Interpretation of Dynamic Scenes.- Control of an Unstable Plant by Computer Vision.- Real-time Processing of Rasterscan Images.- 3-D Kalman Filtering of Image Sequences.- Atmospheric Disturbances Tracking in Satellite Images.- Aspects of Dynamic Scene Analysis in Meteorology.- IV Biomedical Applications.- Processing and Analysis of Radiographic Image Sequences.- Image Sequence Processing and Pattern Recognition of Biomedical Pictures.- A Rule-Based System for Characterizing Blood Cell Motion.- Three Dimensional Imaging from Computed Tomograms.- Model Based Analysis of Scintigraphic Image Sequences of the Human Heart.

Edge And Line Detection in Multidimensional Noisy Imagery Data

Abstract: Detection of edges and lines in multidimensional data is an important operation in a number of image processing applications. The multidimensional picture function is a sampling of the underlying reflectance function of the objects in the scene with the noise added to the true function values. Edges and lines refer to places in the image where there are jumps in the values of the function or its derivatives. The multidimensional greytone surface is expanded as a weighted sum of basis functions. Using multidimensional orthogonal polynomial basis functions, expressions are developed for the coefficients of the fitted quadrautic and cubic surfaces. The parameters of the fitted surfaces are obtained when there is a rotation in the coordinate system. Assuming the noise is Gaussian, statistical tests are devised for the detection of significant edges and lines. Direction isotropic properties of the fitted surfaces are described. For computational efficiency, recursive relations are obtained between the parameters of the fitted surfaces of successive neighborhoods. Furthermore, experimental results are presented by applying the developed theory to multiband Landsat-Imagery Data.

Detection of Edges Using Range Information

Abstract: Range data provide an important source of 3-D shape information. This information can be used to extract jump boundaries which correspond to occluding boundaries of objects in a scene and ``edges'' which correspond to points lying between significantly different regions on the surface of objects. We are mainly interested in range data obtained from sensors such as lasers. The main problem with this type of range finder is the fact that the accuracy of the measurements depends on the power of the signal that reaches the receiver. This study describes how a range edge detection procedure can be designed that has low sensitivity to noise and imbeds all the knowledge available on the range measurement accuracy.

Monocular depth perception from optical flow by space time signal processing

Abstract: A theory of monocular depth determination is presented. The effect of finite temporal resolution is incorporated by generalizing the MarrHildreth edge detected operator - V2G(r) where V2 is the Laplacian and G(r) is a two-dimensional Gaussian. The constraint that the edge detection operator in space-time should produce zero-crossings at the same place in different channels, i.e. at different resolutions of the Gaussian, led to the conclusion that the Marr-Hildreth operator should be replaced by - LI2G(r, t) where E2 is the d'Alembertian V2- (1/u2)(a2/at2) and G(r, t) is a Gaussian in space-time. To ensure that the locations of the zerocrossings are independent of the channel width, G(r, t) has to be isotropic in the sense that the velocity u appearing in the defintion of the d'Alembertian must also be used to relate the scales of length and time in G. However, the new operatior - LI2G(r, t) produces two types of zero-crossing for each isolated edge feature in the image I(r, t). One of these, termed the 'static edge', corresponds to the position of the image edge at time t as defined by V2I(r, t) = 0; the other, called a 'depth zero', depends only on the relative motion of the observer and object and is usually found only in the periphery of the field of view. When an edge feature is itself in the periphery of the visual field and these zeros coincide, there is an additional cross-over effect. It is shown how these zero-crossings may be used to infer the depth of an object when the observer and object are in relative motion. If an edge feature is near the centre of the image (i.e. near the focus of expansion), the spatial and temporal slopes of the zeros crossing at the static edge may be used to infer the depth, but, if the edge feature is in the periphery of the image, the cross-over effect enables the depth to be obtained immediately. While the former utilizes sharp spatial and temporal resolution to give detailed three-dimensional information, the cross-over effect relies on longer integration times to give a direct measure of the time-to-contact. We propose that both mechanisms could be used to extract depth information in computer vision systems and speculate on how our theory could be used to model depth perception in early visual processing in humans where there is evidence of both monocular perception of the environment in depth and of looming detection in the periphery of the field of view. In addition it is shown how a number of previous models are included in our theory, in particular the directional sensor proposed by Marr & Ullman and a method of depth determination proposed by Prazdny.

Detection of moving edges

Abstract: A technique requiring only two frames is presented for finding time-varying edges in a dynamic scene. The frames need not be contiguous. The proposed operator picks up moving edge points which are not easily detected with simple static edge operators. The operator includes both change and edge detection in a way that improves overall performance. Also presented are arguments against the use of three-dimensional operators for scenes where the third dimension is temporal rather than spatial.

A variational approach to edge detection

Abstract: The problem of detecting intensity changes in images is canonical in vision. Edge detection operators are typically designed to optimally estimate first or second derivative over some (usually small) support. Other criteria such as output signal to noise ratio or bandwidth have also been argued for. This paper describes an attempt to formulate a set of edge detection criteria that capture as directly as possible the desirable properties of the detector. Variational techniques are used to find a solution over the space of all possible functions. The first criterion is that the detector have low probability of error i.e. failing to mark edges or falsely marking non-edges. The second is that the marked points should be as close as possible to the centre of the true edge. The third criterion is that there should be low probability of more than one response to a single edge. The third criterion is claimed to be new, and it became necessary when an operator designed using the first two criteria was found to have excessive multiple responses. The edge model that will be considered here is a one-dimensional step edge in white Gaussian noise although the same technique has been applied to an extended impulse or ridge profile. The result is a one dimensional operator that approximates the first derivative of a Gaussian. Its extension to two dimensions is also discussed.

Fast Contour Detection Algorithm for High Precision Quantitative CT

Abstract: An algorithm for the automatic contour detection of objects in CT images is presented. It requires little memory space, allows easy incorporation of different types of local digital filters, and is fast. Typical computation time with a minicomputer is less than one second. An implementation of the algorithm for the detection of bone contours from low dose CT images is given as an example.

Monocular depth perception from optical flow by space time signal processing.

Abstract: A theory of monocular depth determination is presented. The effect of finite temporal resolution is incorporated by generalizing the Marr-Hildreth edge detected operator -$ abla ^{2}$G(r) where $ abla ^{2}$ is the Laplacian and G(r) is a two-dimensional Gaussian. The constraint that the edge detection operator in space-time should produce zero-crossings at the same place in different channels, i.e. at different resolutions of the Gaussian, led to the conclusion that the Marr-Hildreth operator should be replaced by -$\square ^{2}$G(r,t) where $\square ^{2}$ is the d9Alembertian $ abla ^{2}-(1/u^{2})(\partial ^{2}/\partial t^{2})$ and G(r, t) is a Gaussian in space--time. To ensure that the locations of the zero-crossings are independent of the channel width, G(r, t) has to be isotropic in the sense that the velocity u appearing in the defintion of the d9Alembertian must also be used to relate the scales of length and time in G. However, the new operatior -$\square ^{2}$G(r,t) produces two types of zero-crossing for each isolated edge feature in the image I(r, t). One of these, termed the `static edge9, corresponds to the position of the image edge at time t as defined by $ abla ^{2}$I(r,t) = 0; the other, called a `depth zero9, depends only on the relative motion of the observer and object and is usually found only in the periphery of the field of view. When an edge feature is itself in the periphery of the visual field and these zeros coincide, there is an additional cross-over effect. It is shown how these zero-crossings may be used to infer the depth of an object when the observer and object are in relative motion. If an edge feature is near the centre of the image (i.e. near the focus of expansion), the spatial and temporal slopes of the zeros crossing at the static edge may be used to infer the depth, but, if the edge feature is in the periphery of the image, the cross-over effect enables the depth to be obtained immediately. While the former utilizes sharp spatial and temporal resolution to give detailed three-dimensional information, the cross-over effect relies on longer integration times to give a direct measure of the time-to-contact. We propose that both mechanisms could be used to extract depth information in computer vision systems and speculate on how our theory could be used to model depth perception in early visual processing in humans where there is evidence of both monocular perception of the environment in depth and of looming detection in the periphery of the field of view. In addition it is shown how a number of previous models are included in our theory, in particular the directional sensor proposed by Marr & Ullman and a method of depth determination proposed by Prazdny.

Edge detection using sliding statistical tests

Abstract: The problem discussed is that of detecting edges in one-dimensional image profiles, with particular attention being paid to the case in which there is prior information available regarding the scene. Five sliding statistical tests are presented which often display well-defined turning points in the vicinity of edges, thereby enabling edges to be located. Some of the tests are applicable to the case of detecting edges between objects of different intensity, and others are applicable to the case of textures where there may be no difference in average intensity on each side of the edge.

Transforming images into block stationary behavior

Abstract: The statistical behavior of images is inherently nonstationary. Unfortunately, most image processing algorithms assume stationary image models. Spatially adaptive algorithms have been developed which take into account local image statistics. In this paper we derive radiometric and geometric transforms which generate nearly stationary (block stationary) images in the first and second moments. We show that true stationarity is impossible to realize. The aim of these transformations is to enhance the performance of nonadaptive processing techniques, in particular data compression.

Nonlinear local image preprocessing using coherent optical techniques.

Abstract: Two coherent optical systems are described that can realize local nonlinear preprocessing operators such as the Sobel edge-enhancement function in parallel on a 2-D input image. By local, we refer to the size of the image region rather than a nonstationary process. Realization of such operators using a multiple-exposure matched spatial filter and a computer-generated hologram is discussed. Experimental results using these techniques for 3 × 3 and 5 × 5 Sobel operators, respectively, are presented. Our techniques can be extended to larger window sizes and other edge-enhancement operators. These new operators are achieved by novel coherent systems using complex arithmetic with magnitude evaluation of the output pattern.

A multiple model algorithm for the adaptive restoration of images

Abstract: We present a multiple model image restoration technique with on-line edge detection over noisy and blurred images. Four edge models corresponding to major correlation directions and an isotropic model are used to represent the image. Unlike previous results, a decision-directed approach is used to adaptively estimate the edge orientation which then defines the appropriate deconvolution model. The overall effect is space-variant deconvolution implemented by efficient reduced update filters. Processed images are shown as examples.

Elimination of redundant operations for a fast Sobel operator

Abstract: Automatic image segmentation by edge detection techniques often involves recursive operations that incorporate redundant computations. Two methods for implementation of a fast Sobel operator are proposed, which eliminate a considerable amount of redundancy at the expense of increased storage capacity. Comparisons between the standard method and the proposed fast methods are included.

A yariational approach to edge detection

Abstract: l’tlt: probIc:rn of dztc~ctin~ iri:,c:il::ity c!~arlges in images is canonical in \ isIon. lXi;e del,ectlou opc.1 :ltc,rs arc typ]Lally designed to optimally estimate first or second derival.ivr over some (usually small) support. Other criteria such as output signal to noise ratio or bandwidth have also been argued for. This paper describes an attempt to formulate a set of edge detection criteria that capture as directly as possible the desirable properties of the detector. Variational techniques are used to find 2 solution over the space of all possible functions. The first criterion is that the detector have low probability of error i.e. failing to mark edges or falsely marking non-edges. The second is that the marked points should be as close as possible to the centre of the true edge. The third criterion is that there should be low probability of more than one response to a single edge. The third criterion is claimed to be new, and it became necessary when an operator designed using the first two criteria was found to have excessive multiple responses. The edge model that will be considered here is 2 one-dlmensional step edge in white Gaussian noise although the same technique has been applied to an extended impulse or ridge profile. The result is a one dimensional operator that approximates the first derivative of a Gaussian. Its extension to two dimensions is also discussed.

VAP - A Video Array Processor Using Cascaded Look-Up Tables And Its Applications In Biomedicine

Abstract: VAP is a novel digital video array processor for local image transforms working at TV pixel rates in conjunction with an image memory having simultaneous but independent read/ write capability. A programmable system of delay lines is used to access 16 selectable pixels within a local neighbourhood. This information is reduced in a pipeline structure similar to a binary tree, the nodes of which are look-up tables with two inputs and one output. Transformations on whole images may thus be iterated at the TV frame rate. Applications include image preprocessing to allow segmentation by texture, edge detection, etc. but also binary image transformations such as erosion, dilatation, thinning and others.© (1983) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Efficient algorithm for image enhancement

Abstract: The problem of smoothing, point, line and edge detection in image processing is considered. It has been shown that computation in each case can be speeded up considerably by looking at the redundancy. The methods are simple and can be implemented easily in software or hardware.

Differential Operators for Edge Detection

Abstract: We present several results characterizing two differential operators used for edge detection: the Laplacian and the second directional derivative along the gradient. In particular, * (a) we give conditions for coincidence of the zeros of the two operators, and * (b) we show that the second derivative along the gradient has the same zeros of the normal curvature in the gradient direction. Biological implications are also discussed. An experiment is suggested to test which of the two operators may be used by the human visual system. A.I. Laboratory Working Papers are produced for internal circulation, and may contain information that is, for example, too preliminary or too detailed for formal publication. It is not intended that they should be considered papers to which reference can be made in the literature.

A Fast Edge Detection Algorithm Matching Visual Contour Perception

Abstract: For several applications it is very useful to have an edge detection algorithm matching human visual contour perception