KR101920998B1 - apparatus and device for extracting contour by connected component labeling in gray images - Google Patents

Abstract

The present invention relates to a device and a method to extract a contour by labeling connection components. The device includes: a scanning part performing raster scanning on each pixel constituting a grayscale image inputted from the outside; an edge type determining part determining the edge type of an X pixel, which is a pixel being currently scanned, as one type among a strong edge, a weak edge, and a background based on high and low thresholds which are predetermined; a determination tree selecting part selecting one among two preset determination trees based on an edge type of an S pixel, which is a pixel scanned right before the X pixel; a provisional label assigning part assigning a provisional label for the X pixel based on the selected determination tree; and a label flattening part flattening the provisional label based on Vs, which is a memory component storing whether the provisional label includes a strong edge or not.

Description

BACKGROUND OF THE INVENTION 1. Field of the Invention [0001] The present invention relates to an apparatus and method for extracting contours from a gray image,

TECHNICAL FIELD The present invention relates to a contour extracting apparatus and method in a gray image, and more particularly, to a contour extracting apparatus and method through connected component labeling in a gray image.

Thresholding is the most basic low-level operation in many image processing application operations. The thresholding is performed on the input image to obtain a binary image containing foreground and background pixels that can be processed to extract the object.

Hysteresis Thresholding is a more effective thresholding technique in the presence of noise. The hysteresis thresholding includes a high threshold and a low threshold, pixels with a size larger than a high threshold are denoted by strong edge pixels, and pixels with a size lower than the threshold are denoted by background pixels .

Also, pixels with a size lower than the high threshold but higher than the low threshold are initially identified as weak edge pixels. Weak edge pixels are considered strong edge pixels in the future only if they are connected to a strong edge pixel (e.g., a pixel having a magnitude greater than a high threshold).

Conventional contour extraction methods use connection component analysis or contour tracing to extract attributes for binary images. In the conventional method, connected component labeling is performed first, and then attributes are extracted from each labeled object. Further, in hysteresis-thresholded images, an operation of classifying weak edge pixels into strong edge or background pixels is performed.

The conventional method described above had to first apply a connection component labeling to an edge map after passing a gray image through an edge detector to extract a binary edge map.

However, since the use of the edge map requires a large amount of memory, there is a problem in that there is a lot of time for contour extraction.

Korean Patent No. 10-1626538, " System, Method, and Medium for Extracting Image Objects and Contour Feature " Korean Patent No. 10-1679968, " Track Image Counting Method Using Cannier Edge Detection "

The present invention provides a contour extracting apparatus and method for providing connected component labeling directly in a hysteresis thresholding step using two decision tree algorithms proposed by the present invention and a memory element Vs I want to.

It is also intended to provide a contour extracting apparatus and method capable of minimizing time for memory space and contour extraction using a memory element Vs instead of an edge map.

The contour extracting apparatus according to an embodiment of the present invention includes a scanning unit for performing raster scanning on each pixel constituting a gray-scale image input from the outside, Pixel X pixels The edge type of the current pixel, X pixels, is defined as a strong edge, a weak edge, and a background (i.e., a high edge and a low edge) based on a predetermined high threshold and a low threshold and an edge type of the S pixel which is the immediately preceding pixel scanned immediately before the X pixel which is a current pixel of the X pixel which is the current pixel, A decision tree selecting unit for selecting one decision tree, and assigning a provisional label of an X pixel which is a current pixel based on the selected decision tree. Label assignment unit, and based on whether or not strong edge Vs memory element storing the inclusion of the temporary labels assigned to the label includes a platen unit that performs the platen (flatten) for the work of the temporary label the assigned.

According to one aspect, the memory element Vs is represented by '1' if the edge type of the pixel is strong edge, '0' if the edge is not strong edge, and a Boolean vector).

According to one aspect of the present invention, when the edge type of the S pixel which is the immediately preceding pixel is background, the decision tree selection unit selects the X pixel which is the current pixel and the X pixel which is the same as the current pixel, , A P-pixel scanned immediately before the Q-pixel, and a temporary tree of X-pixels that are the current pixel based on the edge type of the R-pixel scanned immediately after the Q-pixel .

According to one aspect, if the edge type of the S pixel which is the immediately preceding pixel is not background, the decision tree selection unit determines that the X pixel, the Q pixel, and the X pixel which is the current pixel based on the edge type of the R pixel You can select a decision tree to assign a temporary label.

According to one aspect, the temporary label assigning unit assigns the temporary label of the Q pixel to an X pixel which is a current pixel if the edge type of the X pixel as the current pixel is strong edge and the edge type of the Q pixel is not background, Changes the memory element Vs corresponding to the temporary label to " 1 ", and if the edge type of the X pixel as the current pixel is a weak edge and the edge type of the Q pixel is not background, Can be assigned to pixels.

According to one aspect, if the edge type of the S pixel which is the immediately preceding pixel is the background, the edge type of the X pixel which is the current pixel is not the background, and the edge type of the Q pixel is the background, Assigning a temporary label of an X pixel which is a current pixel based on an edge type of a pixel and determining whether the edge type of the immediately preceding pixel is not the background and the edge type of the X pixel being the current pixel is not background, If the type is background, it may assign a temporary label of X pixels which is the current pixel based on the edge type of the R pixel.

According to one aspect, the temporary label assigning unit assigns the edge type of the S pixel which is the immediately preceding pixel as the background and the edge type of the X pixel as the current pixel as the background, and the edge type of the Q pixel is the background and the P pixel and the R pixel If the edge type of the previous pixel is not background, perform a label equivalence operation on the temporary label of the P pixel and the R pixel, assign the integrated temporary label to the X pixel which is the current pixel, The edge type of the current pixel is not the background and the edge type of the X pixel is not the background and the edge type of the Q pixel is the background and the edge type of the R pixel is not the background, A label equivalence operation may be performed on the temporary label of the pixel and the integrated temporary label may be assigned to the X pixel which is the current pixel.

)을 수행하고, 상기 현재 픽셀인 X픽셀의 에지 유형이 약한 에지이고 복수개의 통합 대상 픽셀들에 할당된 임시 라벨들 중 어느 하나의 임시 라벨에 대응되는 메모리 요소 Vs가 '1'이면 나머지 하나의 임시 라벨에 대응되는 메모리 요소 Vs도 '1'로 변경하고, 상기 복수개의 통합 대상 픽셀들에 할당된 임시 라벨들을 통합하는 약한 에지 라벨 통합 동작(

)을 수행할 수 있다.According to one aspect, in the label integration operation, if the edge type of the X pixel which is the current pixel is a strong edge, the memory element Vs corresponding to the temporary labels assigned to the plurality of integration target pixels is changed to '1' Strong edge label integration operation < RTI ID = 0.0 >(< / RTI >

), And if the memory element Vs corresponding to the temporary label of any temporary label assigned to the plurality of integration target pixels is '1', the edge type of the X pixel which is the current pixel is a weak edge, A weak edge label integration operation (" 1 ") in which the memory element Vs corresponding to the temporary label is also changed to ' 1 ' and integrates the temporary labels assigned to the plurality of integration target pixels

) Can be performed.

The contour extracting method according to an exemplary embodiment of the present invention includes performing raster scanning on each pixel constituting a gray-scale image input from the outside, A weak edge, and a background, based on a predetermined high threshold and a low threshold, based on the edge type of the X pixels, Selecting one of two predetermined decision trees based on the edge type of the S pixel which is the immediately preceding pixel scanned immediately before the X pixel which is the current pixel; Assigning a provisional label of an X pixel which is a current pixel based on a decision tree of the current pixel and determining whether there is a pixel to be scanned after the X pixel which is the current pixel Determining a completion of the scanning according to the determined temporary labels, and performing a flattening operation on the allocated temporary labels based on the memory element Vs storing the strong edges of the allocated temporary labels after completion of the scanning .

According to one aspect, the memory element Vs is represented by '1' if the edge type of the pixel is strong edge, '0' if the edge is not strong edge, and a Boolean vector).

According to one aspect of the present invention, the selecting of the decision tree comprises: if the edge type of the S pixel which is the immediately preceding pixel is background, the X pixel which is the current pixel and the X pixel which is the same as the current pixel y axis), a P pixel scanned immediately before the Q pixel, and a temporary tree of X pixels that are the current pixel based on the edge type of the R pixel scanned immediately after the Q pixel You can choose.

According to one aspect, the step of selecting the decision tree comprises: if the edge type of the S pixel, which is the immediately preceding pixel, is not background, the current pixel is determined based on the X pixel, the Q pixel, and the edge type of the R pixel You can select a decision tree to assign a temporary label of X pixels.

According to the present invention, connection element labeling can be performed directly in the hysteresis thresholding step using the two decision tree algorithms proposed in the present invention and the memory element Vs.

In addition, using the memory element Vs instead of the edge map, the time for memory space and contour extraction can be minimized.

1 is a configuration diagram showing a contour extracting apparatus according to an embodiment of the present invention.
Figure 2 shows the structure of a pixel on which raster scanning is performed.
FIG. 3A shows the decision tree A. FIG.
FIG. 3B shows the decision tree B. FIG.
4 is a flowchart showing a contour extraction method according to an embodiment of the present invention.

Hereinafter, various embodiments of the present document will be described with reference to the accompanying drawings.

It is to be understood that the embodiments and terminologies used herein are not intended to limit the invention to the particular embodiments described, but to include various modifications, equivalents, and / or alternatives of the embodiments.

In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail.

The following terms are defined in consideration of functions in various embodiments and may vary depending on the intention of a user, an operator, or the like. Therefore, the definition should be based on the contents throughout this specification.

In connection with the description of the drawings, like reference numerals may be used for similar components.

The singular expressions may include plural expressions unless the context clearly dictates otherwise.

In this document, the expressions " A or B " or " at least one of A and / or B " and the like may include all possible combinations of the items listed together.

Expressions such as " first, " " second, " " first, " or " second, " But is not limited to those components.

When it is mentioned that some (e.g., first) component is "(functionally or communicatively) connected" or "connected" to another (second) component, May be connected directly to the component, or may be connected through another component (e.g., a third component).

As used herein, the term " configured to " is intended to encompass all types of information, including, but not limited to, " , "" Made to "," can do ", or" designed to ".

In some situations, the expression " a device configured to " may mean that the device can " do " with other devices or components.

For example, a processor configured (or configured) to perform the phrases " A, B, and C " may be implemented by executing one or more software programs stored in a memory device or a dedicated processor (e.g., an embedded processor) , And a general purpose processor (e.g., a CPU or an application processor) capable of performing the corresponding operations.

Also, the term 'or' implies an inclusive or 'inclusive' rather than an exclusive or 'exclusive'.

That is, unless expressly stated otherwise or clear from the context, the expression 'x uses a or b' means any of the natural inclusive permutations.

In the above-described specific embodiments, elements included in the invention have been expressed singular or plural in accordance with the specific embodiments shown.

It should be understood, however, that the singular or plural representations are selected appropriately for the sake of convenience of description and that the above-described embodiments are not limited to the singular or plural constituent elements, , And may be composed of a plurality of elements even if they are represented by a single number.

While the invention has been shown and described with reference to certain exemplary embodiments thereof, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention as defined by the appended claims.

Therefore, the scope of the present invention should not be limited by the illustrated embodiments, but should be determined by the scope of the appended claims, as well as the appended claims.

FIG. 1 is a configuration diagram showing a contour extracting apparatus according to an embodiment of the present invention, and FIG. 2 shows a structure of a pixel on which raster scanning is performed.

FIG. 2 is a schematic diagram showing an example of a pixel in which raster scanning is in progress; X pixel which is the current pixel, S pixel which is the immediately preceding pixel scanned just before X pixel, Q pixel which is located above the same column (y axis) The P pixel scanned just before the Q pixel and the R pixel scanned immediately after the Q pixel.

1 and 2, a contour extracting apparatus 100 according to an embodiment of the present invention performs raster scanning on all the pixels constituting a gray image, and performs a raster scanning based on a preset threshold value Determines the edge type of the X pixel which is the current pixel under scanning, selects one of the two determined decision trees based on the edge type of the S pixel which is the immediately preceding pixel scanned immediately before the X pixel, Assigns a provisional label of the current pixel based on the selected decision tree, assigns a provisional label of the current pixel to the provisional labels assigned based on the memory element Vs (not shown) that stores whether or not the strong edges of the assigned temporary labels are included, Perform a flattening operation on the surface.

The contour extracting apparatus 100 according to an embodiment of the present invention includes a scanning unit 110, an edge type detecting unit 120, a decision tree selecting unit 130, a temporary label assigning unit 140, (150).

According to an embodiment, the memory element Vs may be represented as '1' if the edge type of the pixel is strong, '0' if it is not a strong edge, and a boolean vector of length one longer each time a new temporary label is assigned Boolean vector).

The scanning unit 110 performs raster scanning on each pixel constituting a gray-scale image input from the outside.

For example, the scanning unit 110 may receive a gray image from a Canny Edge Detector.

The edge type determination unit 120 determines the edge type of the X pixel which is the current pixel in the progress of scanning based on a strong threshold and a strong edge based on a predetermined high threshold and a low threshold weak edge), and background (background).

For example, when the gradient magnitude value of the X pixel which is the current pixel is greater than the high threshold value, the edge type determination unit 120 determines that the edge is strong edge, and if the gradient magnitude value is included in the range between the high threshold value and the low threshold value It is determined as a weak edge. If it is smaller than a low threshold value, it can be determined as a background.

The decision tree selection unit 130 selects any one of the two decision trees preset based on the edge type of the S pixel which is the immediately preceding pixel scanned immediately before the current pixel X pixels.

According to an embodiment, if the edge type of the S pixel which is the immediately preceding pixel is the background, the decision tree selection unit 130 selects the X pixel which is the current pixel and the X pixel which is the current pixel among the scanned peripheral pixels ), A P-pixel scanned just before the Q-pixel, and a temporary tree of X-pixels being the current pixel based on the edge type of the R-pixel scanned immediately after the Q-pixel .

According to the embodiment, if the edge type of the S pixel which is the previous pixel is not background, the decision tree selection unit 130 selects the edge pixel of the X pixel, the Q pixel, and the R pixel, A decision tree B for assigning a temporary label can be selected.

The temporary label assigning unit 140 assigns a provisional label of the current pixel X pixels based on one selected decision tree.

Hereinafter, the detailed operation of the contour extracting apparatus for assigning temporary labels of X pixels, which is the current pixel, based on any one decision tree selected with reference to FIGS. 3A, 3B, [Table 1] and [Table 2] do.

Fig. 3A shows a decision tree A, and Fig. 3B shows a decision tree B. Fig.

The method of labeling operation based on decision tree A or decision tree B is shown in Table 1 below.

[Table 1]

The method of the Vs Modifying Operation of the memory element Vs based on the decision tree A or the decision tree B is shown in Table 2 below.

[Table 2]

On the other hand, in decision tree A and decision tree B, We represents a branch corresponding to a weak edge, and Se indicates a branch corresponding to a strong edge. And b represents a branch corresponding to the background.

Referring to FIGS. 3A, 3B, and 1 and 2, if the edge type of the X pixel which is the current pixel is a strong edge and the edge type of the Q pixel is not the background, the temporary label allocation unit 140 assigns labeling Operation and the correcting operation of Vs can be determined as " 4, gamma ".

Accordingly, the temporary label assigning unit 140 assigns the temporary label of the Q pixel, which is the operation corresponding to Number 4 in [Table 1] and the operation corresponding to? In Table 2, to the X pixel which is the current pixel, The memory element Vs corresponding to the temporary label can be changed to '1'.

On the other hand, if the edge type of the current pixel X is a weak edge and the edge type of the Q pixel is not background, the temporary label assigning unit 140 determines the labeling operation as Number 4 in Table 1, To the X pixel which is the current pixel.

According to the embodiment, the temporary label assigning unit 140 assigns the P pixel and the R pixel of the current pixel to the P pixel and the R pixel, respectively, if the edge type of the S pixel as the previous pixel is the background and the edge type of the X pixel as the current pixel is not the background, And assign a temporary label of X pixels which is the current pixel based on the edge type of the pixel.

On the other hand, if the edge type of the X pixel, which is the background pixel, is not the background and the edge type of the Q pixel is the background, the temporary label assigning unit 140 assigns the edge type of the S pixel To assign a temporary label of X pixels which is the current pixel.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the strong edge, the edge type of the Q pixel is the background, And the edge type of the R pixel is background, the labeling operation and the correction operation of Vs can be determined as " 3, gamma ".

Accordingly, the temporary label assigning unit 140 assigns the temporary label of the P pixel, which is the operation corresponding to Number 3 in Table 1 and the operation corresponding to? In Table 2, to the X pixel which is the current pixel, The memory element Vs corresponding to the temporary label can be changed to '1'.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the strong edge, the edge type of the Q pixel is the background, And if the edge type of the R pixel is not background, the labeling operation and the correcting operation of Vs can be determined as " 5, y ".

Accordingly, the temporary label assigning unit 140 assigns the temporary label of the R pixel, which is the operation corresponding to Number 5 in Table 1 and the operation corresponding to? In Table 2, to the X pixel which is the current pixel, The memory element Vs corresponding to the temporary label can be changed to '1'.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the strong edge, the edge type of the Q pixel is the background, And the edge type of the R pixel is background, the labeling operation and the correction operation of Vs can be determined as " 2, "

Accordingly, the temporary label assigning unit 140 generates a new temporary label for an X pixel which is a current pixel, which is an operation corresponding to Number 2 in [Table 1] and an operation corresponding to? In [Table 2] The memory element Vs corresponding to the label can be assigned '1'.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the weak edge, the edge type of the Q pixel is the background, And the labeling operation is set to " 3 " if the edge type of the R pixel is background.

 Therefore, the temporary label assigning unit 140 can assign the temporary label of the P pixel, which is an operation corresponding to Number 3 in Table 1, to the X pixel which is the current pixel.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the weak edge, the edge type of the Q pixel is the background, And the labeling operation is determined to be " 5 " if the edge type of the R pixel is not background.

Accordingly, the temporary label assigning unit 140 can assign the temporary label of the R pixel, which is the operation corresponding to the Number 5 in Table 1, to the X pixel which is the current pixel.

"로 결정할 수 있다. For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the weak edge, the edge type of the Q pixel is the background, And the edge type of the R pixel is background, the labeling operation and the correction operation of Vs are set to " 2,

"Can be determined.

에 해당하는 동작인, 현재 픽셀인 X픽셀에 대한 새로운 임시 라벨을 생성하고, 새로운 임시 라벨에 대응되는 메모리 요소 Vs를 '0'으로 할당할 수 있다.Therefore, the temporary label assigning unit 140 assigns the operations corresponding to Number 2 in [Table 1]

A new temporary label for the X pixel which is the current pixel, and assigns the memory element Vs corresponding to the new temporary label to '0'.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the previous pixel, to the current pixel, that is, the edge type of the X pixel whose edge type is strong, the edge type of the Q pixel is the background, If it is background, the labeling operation and the correction operation of Vs can be determined as " 6, gamma ".

Accordingly, the temporary label assigning unit 140 assigns the temporary label of the S pixel which is the immediately preceding pixel, which is the operation corresponding to Number 6 in [Table 1] and the operation corresponding to? In Table 2, to the X pixel which is the current pixel , And the memory element Vs corresponding to the assigned temporary label can be changed to '1'.

For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the edge of the X pixel whose edge type is not the background, the edge type of the Q pixel is the background, If the background is the background, the labeling operation can be determined to be " 6 ".

Therefore, the temporary label assigning unit 140 can assign the temporary label of the S pixel which is the immediately preceding pixel, which is the operation corresponding to Number 6 in [Table 1], to the X pixel which is the current pixel.

The temporary label assigning unit 140 determines the labeling operation to be " 1 " if the edge type of the X pixel as the current pixel is the background, and sets the labeling operation to the X pixel which is the current pixel It may not perform the labeling operation.

According to the embodiment, the temporary label assigning unit 140 assigns the edge type of the S pixel which is the immediately preceding pixel as background and the edge type of the X pixel whose current pixel is not the background, the edge type of the Q pixel as the background, A label equalence operation may be performed on the temporary label of the P pixel and the R pixel, and the integrated temporary label may be assigned to the X pixel which is the current pixel.

On the other hand, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, not the background but the edge type of the X pixel where the current pixel is not the background, the edge type of the Q pixel is the background and the edge type of the R pixel is the background It is possible to perform a label equivalence operation on the temporary label of the S pixel and the R pixel, which are the immediately preceding pixels, and assign the integrated temporary label to the X pixel which is the current pixel.

)을 수행할 수 있다. Depending on the embodiment, the label integration operation may change the memory element Vs corresponding to the temporary labels assigned to the plurality of integration target pixels to '1' if the edge type of the X pixel which is the current pixel is strong edge, A strong edge label integration operation that incorporates temporary labels assigned to the target pixels of integration

) Can be performed.

)을 수행할 수 있다.On the other hand, if the memory element Vs corresponding to the temporary label of one of the temporary labels assigned to the plurality of integration target pixels is '1', the edge integration type of the X pixel, which is the current pixel, A weak edge label integration operation (" 1 ") that changes the memory element Vs corresponding to the temporary label to ' 1 ' and integrates temporary labels assigned to a plurality of integration target pixels

) Can be performed.

"로 결정할 수 있다.For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background and the edge type of the X pixel which is the current pixel is the strong edge, the edge type of the Q pixel is the background, If the type is not background, the labeling operation and the corrective action of Vs are " 7,

"Can be determined.

에 해당하는 동작인, P픽셀 및 R픽셀의 임시 라벨들에 대하여 강한 에지 라벨 통합 동작(

)을 수행하고 통합된 임시 라벨을 현재 픽셀인 X픽셀에 할당할 수 있다.Accordingly, the temporary label assigning unit 140 assigns the operation corresponding to the number 7 in Table 1 and the operation

A strong edge label integration operation (see < RTI ID = 0.0 >

) And assign the integrated temporary label to the current pixel, X pixels.

"으로 결정할 수 있다. For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the immediately preceding pixel, to the background, and the edge type of the X pixel whose current pixel is the weak edge, and the edge type of the Q pixel is the background, If the type is not background, the labeling operation and the corrective action of Vs are " 7,

"Can be determined.

에 해당하는 동작인, P픽셀 및 R픽셀의 임시 라벨들에 대하여 약한 에지 라벨 통합 동작(

)을 수행하고 통합된 임시 라벨을 현재 픽셀인 X픽셀에 할당할 수 있다.Accordingly, the temporary label assigning unit 140 assigns the operation corresponding to the number 7 in Table 1 and the operation

A weak edge label integration operation (see < RTI ID = 0.0 >

) And assign the integrated temporary label to the current pixel, X pixels.

"로 결정할 수 있다. For example, the temporary label assigning unit 140 assigns the edge type of the S pixel, which is the previous pixel, to the current pixel, that is, the edge type of the X pixel whose edge type is strong, the edge type of the Q pixel is the background, If the background is not the labeling operation and the correction operation of Vs is " 8,

"Can be determined.

에 해당하는 동작인, 직전 픽셀인 S픽셀 및 R픽셀의 임시 라벨들에 대하여 강한 에지 라벨 통합 동작(

)을 수행하고 통합된 임시 라벨을 현재 픽셀인 X픽셀에 할당할 수 있다.Accordingly, the temporary label assigning unit 140 assigns the operation corresponding to the Number 8 in Table 1 and the operation corresponding to [Table 2]

A strong edge label integration operation (see < RTI ID = 0.0 >

) And assign the integrated temporary label to the current pixel, X pixels.

"로 결정할 수 있다. For example, the temporary label assigning unit 140 assigns the edge type of the S pixel of the immediately preceding pixel to a weak edge (We) of the X pixel whose background pixel is not the background, If the edge type is not background, the labeling operation and the correction operation of Vs are " 8,

"Can be determined.

에 해당하는 동작인, 직전 픽셀인 S픽셀 및 R픽셀의 임시 라벨들은 약한 에지 라벨 통합 동작(

)을 수행하고 통합된 임시 라벨을 현재 픽셀인 X픽셀에 할당할 수 있다.Accordingly, the temporary label assigning unit 140 assigns the operation corresponding to the Number 8 in Table 1 and the operation corresponding to [Table 2]

The temporary labels of the S pixel and the R pixel, which are the immediately preceding pixels, are the weak edge label integration operation

) And assign the integrated temporary label to the current pixel, X pixels.

)은 하기의 알고리즘1에 의하여 수행될 수 있다.A strong edge label integration operation performed when the edge type of the current pixel X pixels is strong edge

) Can be performed by the following algorithm 1.

[Algorithm 1]

)은 임시 라벨

및

의 대표 라벨(representative label)들(

및

)이 강한 에지로 간주되어 대표 라벨들(

및

)에 대응되는 메모리 요소 Vs가 '1'로 변경되고, 대표 라벨

이 대표 라벨

보다 큰 경우에, 대표 라벨

를 대표 라벨

로 변경함으로써, 임시 라벨

및

의 라벨 통합 동작이 수행될 수 있다.Strong edge label integration operation by Algorithm 1 (

) Is a temporary label

And

Of representative labels (

And

) Are regarded as strong edges, and the representative labels (

And

) Is changed to " 1 ", and the representative label

This representative label

If greater, the representative label

Representative label

, The temporary label

And

Can be performed.

)은 하기의 알고리즘2에 의하여 수행될 수 있다.A weak edge label integration operation performed when the edge type of the current pixel X pixels is a weak edge

) Can be performed by the following algorithm 2.

[Algorithm 2]

)은 임시 라벨

및

의 대표 라벨(representative label)들(

및

) 중 어느 하나의 임시 라벨에 대응되는 메모리 요소 Vs가 1인 경우에, 나머지 하나의 임시 라벨에 대응되는 메모리 요소 Vs도 '1'로 변경되고, 대표 라벨

이 대표 라벨

보다 큰 경우에, 대표 라벨

를 대표 라벨

로 변경함으로써, 임시 라벨

및

의 라벨 통합 동작이 수행될 수 있다.Weak Edge Label Integration Operation by Algorithm 2 (

) Is a temporary label

And

Of representative labels (

And

) Is 1, the memory element Vs corresponding to the remaining one temporary label is also changed to " 1 ", and the representative label

This representative label

If greater, the representative label

Representative label

, The temporary label

And

Can be performed.

Referring again to FIG. 1, the label platen part 150 of the contour extracting apparatus 100 according to an embodiment of the present invention determines whether a strong edge of the assigned temporary labels is included, Perform flattening of labels.

For example, when the raster scan operation is completed once for all the pixels, the label platen unit 150 determines that the scanning is completed and performs the flattening operation.

The flattening operation based on the memory element Vs can be performed by the following algorithm 3.

[Algorithm 3]

By algorithm 3, the label platen section 150 can extract contours by removing temporary labels with memory element Vs '0', leaving temporary labels with memory element Vs '1'.

4 is a flowchart showing a contour extraction method according to an embodiment of the present invention.

The method shown in Fig. 4 can be performed by the contour extracting apparatus 100 according to an embodiment of the present invention.

Referring to FIG. 4, in step S410, the apparatus performs raster scanning on each pixel constituting a gray-scale image input from the outside.

For example, in step S410, the device may receive a gray image from a Canny Edge Detector.

In step S420, the apparatus determines whether the edge type of the X pixel, which is a current pixel in the scanning process, is a strong edge, a weak edge (low edge) based on a predetermined high threshold and a low threshold ), And background (background).

For example, in step S420, the apparatus determines a strong edge if the gradient magnitude value of the X pixel, which is the current pixel, is greater than the high threshold value, and if the gradient magnitude value is included in the range between the high threshold value and the low threshold value, And if it is smaller than the lower threshold value, it can be determined as a background.

In step S430, the device selects any one of two preset decision trees based on the edge type of the S pixel which is the immediately preceding pixel scanned just before the current pixel X pixels.

According to an embodiment, in step S430 of selecting a decision tree, the apparatus determines whether the edge type of the S pixel, which is the immediately preceding pixel, is the background, the X pixel which is the current pixel and the X pixel which is the current pixel of the scanned peripheral pixels (y-axis), a P-pixel scanned just before the Q-pixel, and a decision tree that allocates a temporary label of the X-pixel which is the current pixel based on the edge type of the R-pixel scanned immediately after the Q-pixel .

Depending on the embodiment, in step S430 of selecting a decision tree, the device determines whether the edge type of the S pixel, which is the immediately preceding pixel, is background, based on the edge type of X pixels, Q pixels, You can select a decision tree that assigns a temporary label of X pixels.

In step S440, the device assigns a provisional label of an X pixel which is a current pixel based on a selected decision tree.

In step S450, the device determines the completion of scanning according to the presence or absence of a pixel to be scanned after the current pixel X pixels.

For example, in step S450, the apparatus can determine that scanning has been completed once the raster scan operation has been completed for all the pixels constituting the gray image.

For example, if it is determined in step S451 that the laster scanning is not completed, the apparatus can perform a scanning operation on the next pixel.

In step S460, the device performs a flattening operation on the assigned temporary labels based on the memory element Vs storing the strong edges of the temporary labels allocated after the completion of the scanning.

Depending on the embodiment, the memory element Vs may be represented as '1' if the edge type of the pixel is strong, '0' if it is not a strong edge, and as a Boolean vector Boolean vector).

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the apparatus and components described in the embodiments may be implemented within a computer system, such as, for example, a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable array (FPA) A programmable logic unit (PLU), a microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to execution of the software. For ease of understanding, the processing apparatus may be described as being used singly, but those skilled in the art will recognize that the processing apparatus may have a plurality of processing elements and / As shown in FIG. For example, the processing unit may comprise a plurality of processors or one processor and one controller. Other processing configurations are also possible, such as a parallel processor.

The software may include a computer program, code, instructions, or a combination of one or more of the foregoing, and may be configured to configure the processing device to operate as desired or to process it collectively or collectively Device can be commanded. The software and / or data may be in the form of any type of machine, component, physical device, virtual equipment, computer storage media, or device , Or may be permanently or temporarily embodied in a transmitted signal wave.

The software may be distributed over a networked computer system and stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to an embodiment may be implemented in the form of a program command that can be executed through various computer means and recorded in a computer-readable medium. The computer-readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions to be recorded on the medium may be those specially designed and configured for the embodiments or may be available to those skilled in the art of computer software. Examples of computer-readable media include magnetic media such as hard disks, floppy disks and magnetic tape; optical media such as CD-ROMs and DVDs; magnetic media such as floppy disks; Magneto-optical media, and hardware devices specifically configured to store and execute program instructions such as ROM, RAM, flash memory, and the like. Examples of program instructions include machine language code such as those produced by a compiler, as well as high-level language code that can be executed by a computer using an interpreter or the like. The hardware devices described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

100: contour extracting device
110: Scanning section
120: edge type determination unit
130: decision tree selector
140: Temporary label assigning unit
150: Label platen
Se: strong edge
We: Weak Edge
b: Background

Claims (12)

A scanning unit for performing raster scanning on each pixel constituting a gray-scale image input from the outside;
The edge type of the X pixel which is the current pixel in which the scanning is in progress is set to a strong edge, a weak edge, and a background (i.e., a high edge and a low edge) based on a predetermined high threshold and a low threshold an edge type determining unit determining the type of the edge type;
A decision tree selection unit for selecting any decision tree from among two previously determined decision trees based on the edge type of the S pixel which is the immediately preceding pixel scanned immediately before the current pixel X pixels;
A temporary label assigning unit for assigning a provisional label of an X pixel which is a current pixel based on the selected decision tree; And
A label platen part for performing a flattening operation on the allocated temporary labels based on the memory element Vs storing the strong edge of the allocated temporary labels,
Lt; / RTI >
The memory element Vs
If the edge type of the pixel is '1' for strong edges and '0' for non-strong edges, it is a Boolean vector with one longer length each time a new temporary label is assigned.
The decision tree selection unit
If the edge type of the S pixel which is the immediately preceding pixel is the background, the Q pixel located on the same column (y axis) as the current pixel X pixel among the scanned neighboring pixels, Selects a decision tree that allocates a temporary label of the X pixel that is the current pixel based on the immediately previous scanned P pixel and the edge type of the R pixel scanned immediately after the Q pixel
Contour extraction device.
delete delete The method according to claim 1,
The decision tree selection unit
If the edge type of the S pixel which is the immediately preceding pixel is not background, a decision tree for assigning the temporary label of the X pixel which is the current pixel based on the X pixel, the Q pixel, and the edge type of the R pixel, To choose
Contour extraction device.
5. The method of claim 4,
The temporary label assigning unit
If the edge type of the current pixel X pixel is a strong edge and the edge type of the Q pixel is not background, a temporary label of the Q pixel is assigned to the X pixel which is the current pixel, and a memory element Vs To '1', and if the edge type of the current pixel X is a weak edge and the edge type of the Q pixel is not background, the temporary label of the Q pixel is assigned to the current pixel X pixel
Contour extraction device.
5. The method of claim 4,
The temporary label assigning unit
Wherein if the edge type of the S pixel which is the immediately preceding pixel is the background and the edge type of the X pixel which is the current pixel is not the background and the edge type of the Q pixel is the background, based on the edge type of the P pixel and the R pixel, If the edge type of the S pixel which is the immediately preceding pixel is not the background and the edge type of the X pixel which is the current pixel is not the background and the edge type of the Q pixel is the background, Based on the edge type of the current pixel
Contour extraction device.
The method according to claim 6,
The temporary label assigning unit
If the edge type of the S pixel which is the immediately preceding pixel is the background and the edge type of the X pixel which is the current pixel is not the background and the edge type of the Q pixel is the background and the edge type of the P pixel and the R pixel is not the background, Pixel and a temporary label of the R pixel, assigns the integrated temporary label to the X pixel which is the current pixel, and if the edge type of the S pixel which is the immediately preceding pixel is not the background, If the edge type of the current pixel X pixel is not background and the edge type of the Q pixel is the background and the edge type of the R pixel is not background, the label integration of the S pixel immediately before and the temporary label of the R pixel performs a label equivalence operation and allocates the integrated temporary label to the X pixel which is the current pixel
Contour extraction device.
8. The method of claim 7,
The label integration operation
If the edge type of the X pixel as the current pixel is a strong edge, changing the memory element Vs corresponding to temporary labels assigned to a plurality of integration target pixels to '1' Strong edge label integration behavior that integrates labels (

), And if the memory element Vs corresponding to the temporary label of any temporary label assigned to the plurality of integration target pixels is '1', the edge type of the X pixel which is the current pixel is a weak edge, A weak edge label integration operation (" 1 ") in which the memory element Vs corresponding to the temporary label is also changed to ' 1 ' and integrates the temporary labels assigned to the plurality of integration target pixels

) To perform
Contour extraction device.
Performing raster scanning on each pixel constituting a gray-scale image input from the outside;
The edge type of the X pixel which is the current pixel in which the scanning is in progress is set to a strong edge, a weak edge, and a background (i.e., a high edge and a low edge) based on a predetermined high threshold and a low threshold background);
Selecting any of the two decision trees from among the two previously determined decision trees based on the edge type of the S pixel which is the immediately preceding pixel scanned immediately before the current pixel is X pixels;
Assigning a provisional label of an X pixel which is a current pixel based on the selected decision tree;
Determining completion of scanning according to presence or absence of a pixel to be scanned after X pixels as the current pixel; And
Performing a flatten operation on the assigned temporary labels based on a memory element Vs storing the strong edges of the allocated temporary labels after completion of the scanning
Lt; / RTI >
The memory element Vs
If the edge type of the pixel is '1' for strong edges and '0' for non-strong edges, it is a Boolean vector with one longer length each time a new temporary label is assigned.
The step of selecting the decision tree
If the edge type of the S pixel which is the immediately preceding pixel is the background, the Q pixel located on the same column (y axis) as the current pixel X pixel among the scanned neighboring pixels, Selects a decision tree that allocates a temporary label of the X pixel that is the current pixel based on the immediately previous scanned P pixel and the edge type of the R pixel scanned immediately after the Q pixel
Contour extraction method.
delete delete 10. The method of claim 9,
The step of selecting the decision tree
If the edge type of the S pixel which is the immediately preceding pixel is not background, a decision tree for assigning the temporary label of the X pixel which is the current pixel based on the X pixel, the Q pixel, and the edge type of the R pixel, To choose
Contour extraction method.

Images