Showing papers on "Image gradient published in 1986"

A Computational Approach to Edge Detection

Abstract: This paper describes a computational approach to edge detection. The success of the approach depends on the definition of a comprehensive set of goals for the computation of edge points. These goals must be precise enough to delimit the desired behavior of the detector while making minimal assumptions about the form of the solution. We define detection and localization criteria for a class of edges, and present mathematical forms for these criteria as functionals on the operator impulse response. A third criterion is then added to ensure that the detector has only one response to a single edge. We use the criteria in numerical optimization to derive detectors for several common image features, including step edges. On specializing the analysis to step edges, we find that there is a natural uncertainty principle between detection and localization performance, which are the two main goals. With this principle we derive a single operator shape which is optimal at any scale. The optimal detector has a simple approximate implementation in which edges are marked at maxima in gradient magnitude of a Gaussian-smoothed image. We extend this simple detector using operators of several widths to cope with different signal-to-noise ratios in the image. We present a general method, called feature synthesis, for the fine-to-coarse integration of information from operators at different scales. Finally we show that step edge detector performance improves considerably as the operator point spread function is extended along the edge.

Method of extracting an image of a moving object

Abstract: A method of extracting an image of a moving object comprises the steps of extracting edge-enhanced input image from an input image taken by an image pick-up element moving in a plane perpendicular to an optical axis, recollecting a background edge-enhanced image from the input edge image, transfer in parallel horizontally and vertically the background edge-enhanced edge image so to correspond to an input edge-enhanced image forwarded successively; extracting an edge image of a moving object by extracting the background edge-enhanced image from the input edge-enhanced image, and renewing the background edge enhanced image based on weighted mean of the background edge enhanced image and the input edge image.

Method for eliminating noise in regions of a color image exhibiting a specific color

Abstract: A spatial image processing method for noise elimination, including the steps of subjecting an image signal representing a color image and including a density data signal component and a color data signal component to processing for extracting color data for each picture element of the image, discriminating regions of the color image exhibiting a specific color on the basis of the extracted color data, and subjecting the image signal to spatial image processing for elimination of noise under different processing conditions for regions of the image exhibiting the specific color and the remaining regions of the image not exhibiting the specific color.

Image apparatus having a color separation function

Abstract: An image forming apparatus wherein an image exposure device, a whole surface exposure device by light of specific colors and a developing device, are disposed facing a photosensitive member having a surface insulating layer and provided with a color separation function in its surface. In the developing device, the operating condition is changed according to image forming modes. By the combination of the whole surface exposure and development, from an original image in specific colors a visible image in other colors can be provided.

Method for superimposing different images

Abstract: Method for superimposing the difference image of a reference image and current image on the reference image or the current image of a changing process, the reference image or current image being displayed in black/white and the difference image, which can be previously generated and amplified separately or directly during the superimposition, being displayed in colour, or conversely.

Method for detecting/processing image information

Abstract: In photographic printing systems, it is necessary to detect image information of an original film to determine printing exposure amount for optimum prints. When an image sensor is used as a detector or for detecting image information, the detection area in the image sensor should correspond exactly with the detected area on the film. Particularly, when images are measured by separation into three colors of RGB (Red, Green, Blue), color registration among RGB should be attained. Image information can be automatically and accurately detected and processed without needing mechanical positional adjustment of the image sensor(s).

Edge detection with image enhancement via dynamic programming

Abstract: A dynamic programming technique is proposed for locally finding edges in gray level digital images. A class of characteristic functions is proposed from which the best “edge function” (according to an optimality criterion) is chosen for each window by the dynamic programming technique. The windows are selected by first decomposing the entire image into equal square regions and then breaking into four equal subsquares any region where the quality of fit of the best edge function (as measured by the optimality criterion) falls below a fixed tolerance, and so on. The procedure is summarized in a conceptual algorithm, and the technique is first illustrated by application to a digital image of an artificial design. Second, for comparison, the dynamic programming method and a split-and-merge method are used for image enhancement on a noisy version of a muscle cell culture image. Then the dynamic programming method and a thresholded gradient operator technique are used for edge detection on both the original and the noisy version of this image.

Nonlinear point operators for edge detection

Abstract: An edge detection scheme is proposed. This scheme employs a nonlinear point operator, followed by a usual edge detector, e.g. the Sobel edge detector. Three point operator functions are defined and their effects on edge detection are analyzed. The performance of the Sobel edge detector and the nonlinear point operators in the presence of noise is evaluated. Examples of simulated and actual images are given, and the results show that a suitable point operator can improve the edge detection process.

Automatic classification of zooplancton by image analysis

Abstract: Automatic identification and counting of zooplankton is becoming very useful in marine ecology to avoid long and tedious tasks for biologists. We propose a method of image analysis to detect edges, filter the binary image to reduce spots and appendices such as legs or antennas, and measure parameters, such as the length of the abdomen which are finally used in the classification. Some results of classification on 320 specimens are then presented.

Compensation of motion related blur in CCD color endoscope images

Abstract: Color CCD endoscopes with sequential three color illuminations have the advantage of a high resolution but are sensitive to object motion. A digital method for the restoration of color-shifted images is proposed. The method makes use of phase-only matching followed by a resampling operation to compensate for a space-variant shift between different primary color images. Restored images are suitable for further image processing such as color emphasis and structure enhancement.

Storing Feature Descriptions as 2-D Trees

Abstract: Many methods have been proposed which produce lowlevel features from digital images, e. g., the raw primal sketch or intrinsic images. However, in some cases the features occur sparsely in the image, and a more efficient storage scheme can be used than a registered array of feature images. Edges constitute one of the most useful sorts of information for scene analysis. Even though edge responses usually occur sparsely throughout an image, the output from an edge detector in most image analysis systems is itself an image of the same dimensions (but possibly multichannel) as the original intensity image. Appreciable savings in space and time can be achieved if the full edge descriptions (orientation, radius, and likelihood information) are stored as a 2-D tree. This is a binary tree which uses the (x, y) locations of the pixels as keys and splits the data at the median along the key with greatest spread (i. e., this is a k-d tree for k = 2).

Device and method for the optically coherent correlation of two images

Abstract: The invention relates to a device for the optically coherent correlation of two images, namely of an original image and of a reference image representing an object to be found, with the aid of at least two optically coherent Fourier transforms. To be able to carry out all correlations, even triple invariant and multi-spectral correlations, it is proposed to connect an optical Fourier system (22) for optical Fourier transformation with an electronic digital image and transformation system (23). The core of the image and transformation system (23) is a digital computer (31) followed by a digital transformation circuit (34). The computer can be supplied with the image data of the original image, the image data of the reference image, the image data of the image transformed in the optical Fourier system (22) and, if necessary, with further external image or correction data. The digital computer controls the digital transformation circuit, in which image coordinate transformations are carried out for achieving a position, scale and rotation invariance of the correlation.

A True 2D Edge Detector

Abstract: Line finding is a very basic and important step in the low level vision process. Lines are important because they represent the border between two regions and thus help to define and distinguish the region. Lines may also represent physical objects in themselves (at low resolution a tank barrel looks like a ridge line). In the past, ad hoc approaches to line finding have been most prevalent. Recently, Canny has taken a more rigorous approach to edge detection. He has developed an optimal edgel (i.e., individual edge pixel) detector. This detector is optimal under the assumption of additive Gaussian noise and with the constraint that multiple responses from the same edge should be minimized. For a step edge in white Gaussian noise this operator can be closely approximated by convolving the image with a Gaussian mask and then calculating the gradient. A non-maximal suppression operation (in the direction of the gradient at each pixel) is applied to the image. The resulting image is then thresholded. The width of the smoothing Gaussian and the threshold are determined by performance constraints: probability of detection; probability of false alarm; and localization error of the edge. If one accepts the edge model and performance criteria given by Canny, this operator is optimal for detecting and estimating the amplitude and direction of individual edgels (i.e., it is optimal for detecting the 1D edge profile). Unfortunately, Canny's method of grouping individual edgels into lines is not optimal and is in fact quite ad hoc. We have taken a more rigorous approach to extending the Canny 1D edgel detector into a optimal 2D edge detector. Our method applies optimal detection and estimation techniques to the 2D problem. Optimality is determined with respect to the universally most powerful (UNIP) 2 detector for 2D edge. We have been able to develop the optimal detector for a number of edge models. A detector for a constant but unknown amplitude, straight edge model has been implemented. In the implementation we closely approximate the UNIP detector. This paper describes our approach to the problem.

Edge chain analysis for object verification

Abstract: A technique for verifying an object hypothesis by comparing hypothesized edge patterns to detected edge patterns is presented. Positive evidence for a hypothesis is identified as those portions of the model edges that are found in the image, either as geometrical features (such as straight lines or circular arcs} or as linked edges of arbitrary shape. Occluded portions of a model edge are detected by analyzing junctions of image and model edge chains. Negative evidence is identified as those portions of the model edges that are not found and are not occluded. We describe the types of junctions that can occur between image and model edges and discuss a technique for identifying occluded regions based on these junctions. Preliminary results of these techniques are also presented.

Image enhancement via transform amplitude sharpening

Abstract: A new frequency-domain based image edge enhancement technique is described. In the proposed method, the magnitude of the transform of an image is modified using a novel amplitude change function with the phase kept invariant, which is followed by an inverse transform resulting in the crispening of the edge. The new method is somewhat similar to the alpha-rooting method but does not suffer from certain objectionable artifacts associated with the latter, and also exhibits less degradation due to noise. A post-processing of the enhanced image generated using the proposed transform amplitude sharpening method via histogram modification has been found to improve the image quality. Images enhanced with the proposed method are included to illustrate the method.