Expandable grayscale morphological erosion operator
AssigneeLucent Technologies Inc.
InventorReitseng Lin
Life of this patent
Estimating the term…
Abstract
A grayscale morphological erosion operator image processor which operates on images of any size and dimension with a structuring element of any dimension and shape. The present invention performs all 1D erosion operations in the domain by first decomposing an image and structuring element into a series of 1D slices, and then further segmenting the structuring element into a plurality of shorter 1D segments. These segments are processed in parallel using an array of parallel processors. Once the erosion operations are complete, the resulting one dimensional slices are composed into a final eroded image.
Claims15
A method for producing an N-dimensional eroded grayscale image, comprising the steps of:
- (a) defining an N-dimensional structuring element for an N-dimensional uneroded image, wherein both the structuring element and the image are comprised of a plurality of grayscale pixels;
- (b) decomposing both the N-dimensional structuring element and the uneroded image into a plurality of one-dimensional slices of pixels;
- (c) segmenting each one of the plurality of one-dimensional slices of the structuring element into at least two segments;
- (d) producing a plurality of one-dimensional slices of an eroded grayscale image by operating on selected ones of the plurality of one-dimensional slices of the uneroded image with the plurality of one-dimensional slices of the structuring element;
- (e) combining the plurality of one-dimensional slices of the uneroded image to attain an N-dimensional eroded grayscale image of the uneroded image, wherein the at least two segments of each one of the plurality of one-dimensional slices of the structuring element operate in parallel to reducing the computational complexity of the erosion operation and to increase the speed of the erosion operation.
A method for producing an N-dimensional eroded image from an N-dimensional uneroded image, comprising the steps of:
- (a) defining an N-dimensional structuring element for an N-dimensional uneroded image, wherein both the structuring element and the uneroded image are comprised of a plurality of grayscale pixels;
- (b) decomposing both the N-dimensional structuring element and the N-dimensional uneroded image into a plurality of one-dimensional slices of pixels;
- (c) segmenting each one of the plurality of one-dimensional slices of the structuring element into at least two segments;
- (d) employing the at least two segments of a first one of the plurality of one-dimensional slices of the structuring element to operate in parallel on a first one of the plurality of one-dimensional slices of the uneroded image;
- (e) shifting the result of the parallel operation of one of the at least two segments of the first one of the plurality of one-dimensional slices or the structuring element;
- (f) performing a first minimum operation on the shifted result to generate a plurality of one-dimensional erosion arrays, wherein each one of the different pixel locations in the plurality of one-dimensional erosion arrays is comprised of a pixel having the smallest grayscale value of all of the possible pixels available to fill each one or the different pixel locations;
- and
- (g) performing a second minimum operation on the plurality of one-dimensional erosion arrays to generate a single one-dimensional slice of the erosion of the N-dimensional uneroded image, wherein each one of the different pixel locations in the single one-dimensional slice of the erosion of the N-dimensional uneroded image is comprised of a pixel having the smallest grayscale value of all of the possible pixels available to fill each one of the different pixel locations.
An image processing system for processing an image comprised of an array of grayscale pixels, comprising:
- an N-dimensional structuring element comprised of an array of pixels;
- a decomposing unit for decomposing both an N-dimensional image comprised of an array of grayscale pixels and the N-dimensional structuring element into a plurality of one-dimensional slices of pixels;
- segmenting means for segmenting each one of the plurality of one-dimensional slices of the structuring element into at least two segments;
- operating means by which a first one of the at least two segments of the plurality of one-dimensional slices of the structuring element operate in parallel on a first one of the plurality of one-dimensional slices of the image;
- means for shifting the result of the operating means;
- a plurality of intermediate minimum operators coupled to the operating means and to the shifting means for generating a plurality of one-dimensional erosion arrays, wherein each one of the different pixel locations in the plurality of one-dimensional erosion arrays is comprised of a pixel having the smallest grayscale value of all of the possible pixels available to fill each one of the different pixel locations;
- and
- a final minimum operator coupled to the intermediate minimum operators for generating a single one-dimensional slice of the erosion of the image, wherein each one of the different pixel locations in the single one-dimensional slice of the erosion of the uneroded image is comprised of a pixel having the smallest grayscale value of all of the possible pixels available to fill each one of the different pixel locations.
Dependencies are inferred from the claim text.