US 4,908,872 AGranted

Method and apparatus for extracting pattern contours in image processing

AssigneeFujitsu Limited

InventorsTakashi Toriu, Hiromichi Iwase

Life of this patent

Estimating the term…

Loading patent…

Abstract

This invention relates to a method and apparatus for extracting contour lines from a pattern obtained in image processing. The disclosed method and apparatus makes it possible to obtain the contour lines in real-time and, moreover, can be easily incorporated into hardware. To achieve these objects, a invention applies the simplified discrimination method for selecting pixels, which form the pattern contour, wherein each pixel is examined within groups of pixel, i.e. X-axis and Y-axis groups. The magnitude of a gray level gradient of each pixel is compared with those of neighboring pixels located in either direction of X-axis group or Y-axis. The pixel which has the maximum magnitude of gray level gradient among those of adjacent pixels is discriminated from each group as contour date. This discrimination method is simple and remarkably improved. When contour lines for two groups are combined, the pattern contours can be easily obtained.

Claims8

Dependent claims

A method of extracting pattern contours from an image formed by a plurality of pixels arranged in x and y coordinates, each pixel having a gray level graident, said method comprising the steps of:

(a) calculating and outputting a magnitude and direction of the gray level gradient of each pixel among the plurality of pixels;
(b) discriminating the direction of the gray level gradient from within predetermined ranges, and outputting the magnitude of the discriminated gray level gradient for a first group among the plurality of pixels, the discriminated direction of the gray level gradient being within a specified angle range relative to the X-axis and outputting a null signal for pixels not within the specified angle range;
(c) discriminating the direction of the gray level gradient from within predetermined ranges, and outputting the magnitude of the discriminated gray level gradient for a second group among the plurality of pixels, the discriminated direction being within a specified angle range relative to the Y-axis and outputting a null signal for pixels not within the specified angle range;
(d) discriminating each pixel from among the first group, a pixel being discriminated when having a magnitude of gray level gradient greater than or equal to the magnitude of gray level gradient of a specified number of adjacent pixels located on both sides of the pixel in the X-direction, outputting a first level logic signal for said discriminated pixel, and outputting a second level logic signal when the pixel is not discriminated;
(e) discriminating each pixel from among the second group, a pixel being discriminated when having a magnitude of gray level gradient greater than or equal to the magnitude of gray level gradient of a specified number of adjacent pixels located on both sides of the pixel in the Y-direction, outputting a first level logic signal for said discriminated pixel, and outputting a second level logic signal when the pixel is not discriminated;
and
(f) forming the pattern contours by providing a logical sum of the outputted logic signals from said steps (d) and (e).

4 more elements

An apparatus for extracting pattern contours from an image formed by a plurality of pixels arranged in X and Y coordinates, each pixel having a gray level gradient with a magnitude and direction, said apparatus comprising:

first means for calculating the magnitude and direction of gray level gradient for a pixel;
second means for discriminating the direction of the gray level gradient from among a specified angle range with respect to the X-axis, and outputting said discriminated magnitude of the gray level gradient for a first group of pixels, said second means outputting a null signal for other pixels not discriminated from within the specified angle range with respect to the X-axis;
third means for discriminating the direction of the gray level gradient from among a specified angle range with respect to the Y-axis, and outputting said discriminated magnitude of the gray level gradient for a second group of pixels, said third means outputting a null signal for other pixels not discriminated from within the specified angle range with respect to the Y-axis;
fourth means for discriminating each pixel from within the first group, a pixel being discriminated when having the magnitude of gray level gradient greater than or equal to a maximum magnitude of adjacent pixels of a specified number located on both sides of the pixel in the X-direction, said fourth means outputting a first level logic signal for said discriminated pixel, and outputting a second level logic signal when the pixel is not discriminated;
fifth means for discriminating each pixel from within the second group, a pixel being discriminated when having the magnitude of gray level gradient greater than or equal to the magnitude of adjacent pixels of the specified number located on both sides of the pixel in the Y-direction, said fifth means outputting a first level logic signal for said discriminated pixel, and outputting a second level logic signal when the pixel is not discriminated;
and
a logic sum circuit for summing the outputted logic signals from said fourth and fifth means.

4 more elements

Dependencies are inferred from the claim text.

Facts

Application: 07/153,426Filed 5 Feb 1988
Priority: JP 62-257116 Feb 1987
Granted: 13 Mar 1990
Term ends: Not estimatedThe source records don’t support an estimate.

Classification6 codes

G06T 7/12
Segmentation; Edge detection: edge-based segmentation
G06V 10/36
Image preprocessing: applying a local operator, i.e. means to operate on image points situated in the vicinity of a given point; Non-linear local…
G06V 10/443
Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections;…: by…
G06T 2200/28
Indexing scheme for image data processing or generation, in general: involving image processing hardware