Identification and segmentation of finely textured and solid regions of binary images

Claim 2

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that irreversibly maps the textured and solid regions to solid regions in a resulting image, and reversibly maps features containing ON pixels in the other regions to proportionately thicker features in the resulting image; and subjecting the resulting image to a second set of operations that maps the solid regions to solid regions in a further resulting image and maps the proportionately thickened features in a manner that they are substantially absent from the further resulting image; whereupon the further resulting image contains solid regions corresponding to the textured and solid regions of the input image.

Claim 3

The improvement of claim 1 wherein at least some of said first and second sets of operations are performed in a programmed digital computer.

Claim 4

The improvement of claim 1 wherein the resulting image is smaller than the input binary image.

Claim 5

The improvement of claim 1, and further comprising the steps of: determining the number of solid regions in the further resulting image; and in the event of determining the number of solid regions to be greater than a predetermined maximum, removing the smallest of the solid regions.

Claim 6

The improvement of claim 4 wherein said step of removing the smallest of the solid regions comprises the steps of: initializing a structuring element size; determining the remaining number of solid regions and exiting if the number of remaining solid regions is below the predetermined maximum; increasing the structuring element size; opening the image with the structuring element as currently sized; and repeating said determining, increasing, and opening steps.

Claim 7

The improvement of claim 1, and further comprising the step of: filling holes, if any, within the solid regions of the further resulting image.

Claim 8

The improvement of claim 6 wherein said step of filling holes within the solid regions of the further resulting image comprises filling each eight-connected region of ON pixels to the smallest possible enclosing rectangle.

Claim 9

The improvement of claim 6 wherein said step of filling holes within the solid regions comprises filling each four-connected region of ON pixels to the smallest possible enclosing rectangle.

Claim 10

The improvement of claim 1 wherein: said first set of operations includes, in the order stated at least one reduction operation with a low threshold level and a close operation using a solid structuring element; and said second set of operations includes, in the order stated, at least one reduction operation with a high threshold level and an open operation using a solid structuring element.

Claim 11

The improvement of claim 1 wherein: at least one of said first and second sets of operations entails reducing the image size; and the improvement further comprises the step of expanding the further resulting image to form a mask at the same scale as the input image.

Claim 12

The improvement of claim 10, and further comprising the steps of logically combining the mask and the original binary image.

Claim 13

The improvement of claim 1, and further comprising the step of extracting information regarding the boundaries surrounding the solid regions in the further resulting image.

Claim 14

The improvement of claim 1 wherein the further resulting image is smaller than the resulting image.

Claim 15

The improvement of claim 1 wherein said first set of operations comprises at least one reduction operation with a low threshold level.

Claim 16

The improvement of claim 1 wherein said first set of operations comprises at least one subsampling operation.

Claim 17

The improvement of claim 1 wherein said second set of operations comprises at least one reduction operation with a high threshold level.

Claim 18

The improvement of claim 1 wherein said second set of operations comprises at least one subsampling operation.

Claim 19

The improvement of claim 1 wherein: said first set of operations includes at least one reduction operation with a low threshold level; and said second set of operations includes at least one reduction operation with a high threshold level.

Claim 20

The improvement of claim 1 wherein said first set of operations comprises the steps of: dilating the input image using a solid structuring element; and subsampling the dilated version of the input image to form a reduced image.

Claim 21

The improvement of claim 1 wherein said second set of operations comprises the steps of: eroding the resulting image using a solid structuring element; and subsampling the eroded version of the resulting image to form a reduced image.

Claim 22

The improvement of claim 1 wherein: both of said first and second sets of operations entail reducing the image size; and the improvement further comprises the step of expanding the further resulting image to form a mask at the same scale as the input image.

Claim 23

The improvement of claim 21, and further comprising the steps of logically combining the mask and the original binary image.

Claim 24

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that eliminates OFF pixels that are near ON pixels to produce a resulting image with solid regions corresponding to textured and solid regions in the input image; and subjecting the resulting image to a second set of operations that eliminates ON pixels that are near OFF pixels to produce a further resulting image with solid regions corresponding to textured and solid regions in the input image and regions substantially devoid of ON pixels corresponding to other regions in the input image; whereupon textured and solid regions in the input image are selectively converted to solid ON pixels and ON pixels within other regions in the input image are selectively eliminated.

Claim 25

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that eliminates OFF pixels that are near ON pixels to produce a resulting image with solid regions corresponding to textured and solid regions in the input image; and subjecting the resulting image to a second set of operations that eliminates ON pixels that are near OFF pixels to produce a further resulting image with solid regions corresponding to textured and solid regions in the input image and regions substantially devoid of ON pixels corresponding to other regions in the input image; whereupon textured and solid regions in the input image are selectively converted to solid ON pixels and ON pixels within other regions in the input image are selectively eliminated.

Claim 26

The improvement of claim 23 wherein said first set of operations comprises: at least one reduction operation with a low threshold level; and a close operation using a solid structuring element.

Claim 27

The improvement of claim 23 wherein said first set of operations comprises the steps of: dilating the input image using a solid structuring element; and subsampling the dilated version of the input image to form a reduced image.

Claim 28

The improvement of claim 23 wherein said second set of operations comprises: at least one reduction operation with a high threshold level; and an open operation using a solid structuring element.

Claim 29

The improvement of claim 23 wherein said second set of operations comprises the steps of: eroding the resulting image using a solid structuring element; and subsampling the eroded version of the resulting image to form a reduced image.

Claim 30

The improvement of claim 23 wherein the resulting image is smaller than the input image.

Claim 31

The improvement of claim 23 wherein the further resulting image is smaller than the resulting image.

Claim 32

The improvement of claim 23 wherein said first set of operations comprises at least one reduction operation with a low threshold level.

Claim 33

The improvement of claim 23 wherein said first set of operations comprises at least one subsampling operation.

Claim 34

The improvement of claim 23 wherein said second set of operations comprises at least one reduction operation with a high threshold level.

Claim 35

The improvement of claim 23 wherein said second set of operations comprises at least one subsampling operation.

Claim 36

The improvement of claim 23 wherein: said first set of operations includes at least one reduction operation with a low threshold level; and said second set of operations includes at least one reduction operation with a high threshold level.

Claim 37

The improvement of claim 23 wherein: at least one of said first and second sets of operations entails reducing the image size; and the improvement further comprises the step of expanding the further resulting image to form a mask at the same scale as the input image.

Claim 38

The improvement of claim 35, and further comprising the steps of logically combining the mask and the input image.

Claim 39

The improvement of claim 23, and further comprising the step of extracting information regarding the boundaries surrounding the solid regions is the further resulting image.

Claim 40

In a method for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first transformation that leaves the solid regions solid, converts the textured regions to solid regions, and thickens features in the other regions; and subjecting the resulting image to a second transformation that eliminates ON pixels in the other regions without substantially affecting the solid regions.

Claim 41

In a method for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first transformation that leaves the solid regions solid, converts the textured regions to solid regions, and thickens features in the other regions; and subjecting the resulting image to a second transformation that eliminates ON pixels in the other regions without substantially affecting the solid regions.

Claim 42

The improvement of claim 38 wherein said first-mentioned and second-mentioned subjecting steps together comprise the following steps, carried out in the order stated: performing at least one reduction operation with a low threshold level; performing a close operation using a solid structuring element; performing at least one reduction operation with a high threshold level; and performing an open operation using a solid structuring element.

Claim 43

The improvement of claim 23 wherein said first-mentioned and second-mentioned subjecting steps together comprise the following steps, carried out in the order stated: dilating the input image using a solid structuring element; subsampling the dilated image to form a reduced image; eroding the image using a solid structuring element; and subsampling the eroded image to form a reduced image.

Claim 44

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that irreversibly maps the textured and solid regions to solid regions in a resulting image of reduced size, and reversibly maps features containing ON pixels in the other regions to proportionately thicker features in the resulting image; and subjecting the resulting image to a second set of operations that maps the solid regions to solid regions in a further resulting image of further reduced size and maps the proportionately thickened features in a manner that they are substantially absent from the further resulting image; whereupon the further resulting image contains solid regions corresponding to the textured and solid regions of the input image.

Claim 45

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that irreversibly maps the textured and solid regions to solid regions in a resulting image of reduced size, and reversibly maps features containing ON pixels in the other regions to proportionately thicker features in the resulting image; and subjecting the resulting image to a second set of operations that maps the solid regions to solid regions in a further resulting image of further reduced size and maps the proportionately thickened features in a manner that they are substantially absent from the further resulting image; whereupon the further resulting image contains solid regions corresponding to the textured and solid regions of the input image.

Claim 46

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that eliminates OFF pixels that are near ON pixels to produce a resulting image of reduced size with solid regions corresponding to textured and solid regions in the input image; and subjecting the resulting image to a second set of operations that eliminates ON pixels that are near OFF pixels to produce a further resulting image of further reduced size with solid regions corresponding to textured and solid regions in the input image and regions substantially devoid of ON pixels corresponding to other regions in the input image; whereupon textured and solid regions in the input image are selectively converted to solid ON pixels and ON pixels within other regions in the input image are selectively eliminated.

Claim 47

In a method, performed in a digital processor, for separating textured and solid regions, if any, in a binary input image from other regions, if any, the improvement comprising the steps of: subjecting the input image to a first set of operations that eliminates OFF pixels that are near ON pixels to produce a resulting image of reduced size with solid regions corresponding to textured and solid regions in the input image; and subjecting the resulting image to a second set of operations that eliminates ON pixels that are near OFF pixels to produce a further resulting image of further reduced size with solid regions corresponding to textured and solid regions in the input image and regions substantially devoid of ON pixels corresponding to other regions in the input image; whereupon textured and solid regions in the input image are selectively converted to solid ON pixels and ON pixels within other regions in the input image are selectively eliminated.