KR102340303B1 - Method for acquiring binary edge image and computer readable recording medium having computer program recorded therefor - Google Patents

Abstract

A method for obtaining a binary edge image according to an embodiment of the present invention includes the steps of generating a gray scale image, selecting N pixels in the order of increasing edgeness for each region from a plurality of regions in the gray scale image , calculating a lower limit value based on the average of the edge maps of the plurality of selected pixels, based on the average of the edge maps of the remaining pixels except for pixels having an edge map of the plurality of selected pixels equal to or less than the lower limit value calculating an upper limit value, selecting a binary edge threshold value between the upper limit value and the lower limit value, and performing binary processing of the grayscale image based on the selected binary edge threshold value

Description

Method for acquiring binary edge image and a computer readable recording medium in which a program for executing the same is recorded

The present invention relates to a method for obtaining a binary edge image and a computer-readable recording medium on which a program for executing the same is recorded, and more particularly, to a method for obtaining a binary edge image obtained by binarizing an edge portion of a shape in an image and the same It relates to a computer-readable recording medium in which a program for executing a method is recorded.

In a chip mounter that mounts various types and sizes of electronic components such as integrated circuits, diodes, capacitors, resistors, and transistors on a printed circuit board at high speed, a vision algorithm is essential for accurate mounting of electronic components.

The vision algorithm for electronic parts uses a camera to identify the types and distortions of electronic parts through images taken, and reflects them so that the electronic parts are mounted in a required position and in a required direction.

In order to use the electronic component image in the vision algorithm, clear outline information of the electronic component image is required. To this end, the vision algorithm extracts a binary edge image from the electronic component image taken by the camera. this is required

The conventional vision algorithm is configured to input a binary edge threshold by a user, but due to the diversity of electronic components and various lighting environments in which electronic components are photographed, the binary edge threshold cannot be uniformly determined and used. .

For example, even if the same binary edge threshold is used, the quality of the binary edge image varies according to the type of electronic component or the lighting environment in which the electronic component is photographed.

An object of the present invention is to provide a binary edge image acquisition method for automatically acquiring a binary edge image, and a computer-readable recording medium in which a program for executing the same is recorded.

The problems of the present invention are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood by those skilled in the art from the following description.

The method for obtaining a binary edge image according to an embodiment of the present invention for solving the above problems includes the steps of generating a gray scale image, and in a plurality of regions in the gray scale image, N in the order of increasing edgeness for each region selecting pixels, calculating a lower limit value based on an average of edge degrees of the plurality of selected pixels, and edges of pixels other than a pixel having an edge degree of less than or equal to the lower limit among the selected plurality of pixels calculating an upper limit value based on the average of the figures, selecting a binary edge threshold value between the upper limit value and the lower limit value, and binary processing the gray scale image based on the selected binary edge threshold value .

Other specific details of the invention are included in the detailed description and drawings.

According to the embodiments of the present invention, there are at least the following effects.

By automatically calculating the binary edge threshold, it is easier to obtain a more accurate binary edge image.

The effect according to the present invention is not limited by the contents exemplified above, and more various effects are included in the present specification.

1 is a flowchart illustrating a binary edge image acquisition method according to an embodiment of the present invention.
FIG. 2 schematically illustrates an example of a component image input in step S10 of FIG. 1 .
FIG. 3 shows a gray scale image generated by the Sobel operation in step S20 of FIG. 1 with respect to the part image of FIG. 2 .
FIG. 4 is a graph showing the distribution of pixels selected in step S30 of FIG. 1 with respect to the gray scale image of step S20 of FIG. 1 .
FIG. 5 is a primary histogram H1 generated in step S40 of FIG. 1 .
6 is a histogram of all pixels in the gray scale image of step S20 of FIG. 1 .
FIG. 7 is a histogram representing the lower limit of the binary edge threshold calculated in step S50 of FIG. 1 on the primary histogram H1 of FIG. 5 .
8 is a histogram expressing the secondary histogram H2 generated in step S60 of FIG. 1 and the upper limit value of the binary edge threshold calculated in step S70 of FIG. 1 .
FIG. 9 illustrates a binary edge image finally generated in step S90 of FIG. 1 .

Advantages and features of the present invention and methods of achieving them will become apparent with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but may be implemented in various different forms, and only these embodiments allow the disclosure of the present invention to be complete, and common knowledge in the art to which the present invention pertains It is provided to fully inform those who have the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numerals refer to like elements throughout.

Further, the embodiments described herein will be described with reference to cross-sectional and/or schematic diagrams that are ideal illustrative views of the present invention. Accordingly, the shape of the illustrative drawing may be modified due to manufacturing technology and/or tolerance. In addition, in each of the drawings shown in the present invention, each component may be enlarged or reduced to some extent in consideration of convenience of description. Like reference numerals refer to like elements throughout.

Hereinafter, the present invention will be described with reference to drawings for explaining a method for obtaining a binary edge image according to an embodiment of the present invention.

1 is a flowchart illustrating a binary edge image acquisition method according to an embodiment of the present invention.

As shown in Fig. 1, the binary edge image acquisition method according to an embodiment of the present invention includes a step of inputting a part image (S10), a step of generating a gray scale image (S20), and selecting N pixels for each area. (S30), generating a primary histogram (H1) (S40), calculating a lower limit value of the binary edge threshold value (S50), generating a secondary histogram (H2) (S60), binary edge Calculating the upper limit of the threshold value (S70), selecting a binary edge threshold value (S80), and generating a binary edge image (S90) are included.

The step S10 of inputting the component image is a step of inputting an image of a component to be mounted on a printed circuit board or already mounted on the printed circuit board. FIG. 2 shows an example of a component image input in step S10 of FIG. 1 .

As shown in FIG. 2 , the input part image I1 may be an image in which the part and the background exist together.

The step (S10) of inputting the part image may be configured such that the part image photographed by the camera is input in real time. To this end, the apparatus using the binary edge image acquisition method according to an embodiment of the present invention may include a camera for photographing parts or may be configured to receive image data photographed by the camera.

Alternatively, the step of inputting the part image ( S10 ) may be configured to load the part image stored in a separate storage medium. To this end, the apparatus using the binary edge image acquisition method according to an embodiment of the present invention may include a storage medium in which part images are stored as data or may be configured to receive image data from the storage medium.

The step of generating the gray scale image ( S20 ) is a step of processing the part image input in step S10 and converting it into a gray scale image. A gray scale image in the method for obtaining a binary edge image according to an embodiment of the present invention is an image expressed in shades of gray including white and black, and refers to an image that is not composed of only black and white. The gray scale image may be an image in which the gray scale is divided into 256 levels from 0 to 255, and 0 and 255 may mean black and white, respectively.

In the step of generating the gray scale image ( S20 ), a process of converting a color image into a gray scale image may be performed. When the part image input in step S10 is a gray scale image, the input image may be used as a gray scale image of the present embodiment as it is.

In addition, for more effective binary edge processing, in the step S20 of generating the gray scale image, edge processing for making the contour portion of the shape in the image stand out may be performed on the part image input in step S10 .

As a method for edge processing, there are edge processing methods using a matrix operator as shown in the following table.

operator name f(x) f(y) Roberts

Prewitt

Sobel

In the step (S20) of generating a grayscale image according to the present embodiment, an example in which a Sobel operation capable of extracting horizontal, vertical edges, and diagonal edges and a strong noise-resistant Sobel operation is used will be described. FIG. 3 shows a gray scale image generated by the Sobel operation in step S20 of FIG. 1 with respect to the part image shown in FIG. 2 .

Meanwhile, in the step of selecting N pixels for each area ( S30 ), N pixels are selected for each area from a plurality of areas in the gray scale image generated in step S20 .

A plurality of regions are set by dividing the gray scale image into equal parts. For example, the plurality of regions may be set by equally dividing the gray scale image I2 into a plurality of columns and/or a plurality of rows. Alternatively, the plurality of regions may be set by dividing the gray scale image I2 into each pixel column and/or each pixel row. In this embodiment, a case in which a plurality of regions is set by dividing the gray scale image I2 into respective pixel columns will be described as a reference.

The N pixels selected for each region are selected in order of highest edgeness among pixels constituting each region. For example, when a plurality of regions is set by dividing the gray scale image I2 into each pixel column, N pixels are selected in an order of increasing edgeness among a plurality of pixels constituting each pixel column. .

N may be an integer corresponding to about 10% of the number of pixels constituting one pixel column. For example, when one pixel column consists of 1024 pixels, 102 pixels among 1024 pixels may be selected. This describes an example in which N is selected, and N may be determined as a relative number such as a top few percent of the number of pixels constituting one pixel column, or may be determined as an absolute number such as 50 or 100.

The edgeness may be a gray scale value expressed by being divided into 256 steps in a gray scale image. Alternatively, it may be a numerical value of each pixel through edge processing such as a Sobel operation.

FIG. 4 is a graph showing the distribution of pixels selected in step S30 of FIG. 1 with respect to the gray scale image of step S20 of FIG. 1 . The horizontal direction of the graph shown in FIG. 4 corresponds to the horizontal direction of the gray scale image as a distribution of each pixel column, and the vertical direction of the graph shown in FIG. 4 is edgeness.

As shown in FIG. 4 , as N pixels are selected in the order of high edgeness for each pixel column, a noise edge in the background and a noise edge in part region) and a valid edge in part region are roughly distinguished.

Most of the part image is with the part in the center and the background is formed on both sides of the image. In addition, pixels corresponding to the background in the gray scale image of the part image are mostly close to black and are not easily distinguished from adjacent pixels, and thus have low edgeness.

Accordingly, as shown in FIG. 4 , most of the N pixels selected from both pixel columns of the graph corresponding to the background portion are concentrated at the bottom of the graph, and these correspond to the noise edge in the background. can be easily identified.

On the other hand, in the part image, the part is mostly located in the center, and pixels corresponding to the part in the gray scale image of the part image are expressed brighter than the background.

Accordingly, as shown in FIG. 4 , each of the N pixels selected from the pixel columns in the center of the graph where the component is located is located above the noise edge in the background.

And, as shown in FIG. 4 , the pixels corresponding to the part have a valid edge in part region positioned at the upper part of the graph, and a noise edge in the part region positioned lower than that. part region).

A valid edge in a part region corresponds to pixels corresponding to an actual part edge. In FIG. 3 , an actual part edge is expressed in the brightest color, and the edge is expressed close to a line to be clearly distinguished from an adjacent pixel, and thus has high edgeness. Accordingly, in the graph of FIG. 4, these pixels are located at the top. As shown in FIG. 3 , since the edge of the part expressed in the brightest color is located on both sides, the valid edge in the part region is expressed in two groups in the graph of FIG. 4 as well.

Meanwhile, a noise edge in part region corresponds to pixels that are not actual part edges. In the gray scale image, the part corresponding to the non-edge part of the part is expressed brighter than the background, but darker than the part's edge. It has a lower edgeness than Accordingly, in the graph of FIG. 4 , these pixels are positioned between a valid edge in part region and a noise edge in background.

As expressed by the edgeness distribution for N pixels selected for each pixel column in FIG. 4 , a valid edge in part region is defined as a noise edge in the background. and a noise edge in part region, valid edges can be extracted.

To find a binary edge threshold that can distinguish a valid edge in part region from Noise edge in background and Noise edge in part region, steps can be performed.

In the step S40 of generating the first histogram H1, the first histogram H1 is generated using the N pixels selected for each pixel column in step S30.

5 is a primary histogram generated in step S40 of FIG. 1 .

As shown in FIG. 5 , the primary histogram H1 is generated such that the horizontal axis represents the edgeness of each pixel and the vertical axis represents the number of pixels.

As shown in FIG. 5 , pixels in the primary histogram H1 largely form three groups A1 , A2 , and A3 according to edgeness. The first group A1 includes pixels corresponding to a noise edge in the background of FIG. 4 , and the second group A2 is a noise edge in part region of FIG. 4 . ), and the third group A3 includes pixels corresponding to a valid edge in a part region.

Meanwhile, FIG. 6 is a histogram of all pixels in the gray scale image of step S20.

When the histogram H0 is generated for all pixels in the image of the gray scale image of step S20 without going through step S30, as shown in FIG. 6, a valid edge in part region is Of course, a noise edge in part region and a noise edge in background are not distinguished from each other.

In step S50 of calculating the lower limit of the binary edge threshold, the lower limit T _L of the binary edge threshold is calculated based on the primary histogram H1 generated in step S40.

FIG. 7 is a histogram representing the lower limit of the binary edge threshold calculated in step S50 of FIG. 1 on the primary histogram of FIG. 5 .

_{The lower limit value T L} of the binary edge threshold value may be calculated as an average of the primary histogram H1 . That is, the average of edgeness of each pixel expressed in the primary histogram H1 may be _{the lower limit value T L of the binary edge threshold value.}

As shown in FIG. 7 , since a significant number of pixels expressed in the primary histogram H1 are located in the first group A1, the lower limit value of the binary edge threshold value T _L is in the first group A1. It may be located adjacent to, between the first group A1 and the second group A2 or within the second group A2.

Alternatively, the step of calculating the lower limit of the binary edge threshold (S50) may be performed after the step (S30) of selecting N pixels for each area without going through the step (S40) of generating the first histogram H1. have. _{In this case, without separately generating the primary histogram H1, the lower limit value T L} of the binary edge threshold value can be calculated by summing the edgeness of the pixels selected in step S30 and dividing by the number of pixels.

In the step S60 of generating the secondary histogram H2, among the pixels selected in step S30, pixels having an edgeness less than or less than _{the lower limit value T L of the binary edge threshold calculated in step S50 are selected.} A second histogram H2 is generated with respect to the remaining pixels.

FIG. 8 is a histogram expressing the secondary histogram generated in step S60 of FIG. 1 and the upper limit value of the binary edge threshold calculated in step S70 of FIG. 1 .

As shown in FIG. 8 , the secondary histogram H2 is also generated such that the horizontal axis indicates the edgeness of each pixel and the vertical axis indicates the number of pixels, similar to the primary histogram H1 . .

As shown in FIG. 8 , most pixels corresponding to the first group A1 do not appear in the second histogram H2 , and pixels of the second group A2 and the third group A3 remain. . _{When some pixels having an edgeness lower than the lower limit value T L} of the binary edge threshold value are included in the second group A2 , the pixels may not appear in the secondary histogram H2 .

In the step of calculating the upper limit value of the binary edge threshold value ( S70 ), the upper limit value T _U of the binary edge threshold value is calculated based on the secondary histogram H2 generated in step S60 .

_{The upper limit value T U} of the binary edge threshold value may be calculated as an average of the quadratic histogram H2 . That is, the average of the edgeness of each pixel expressed in the secondary histogram H2 may be _{the upper limit value T U of the binary edge threshold value.}

As shown in FIG. 8 , the upper limit value T _U of the binary edge threshold value may be located in the third group A3 . _{The upper limit value T U} of the binary edge threshold value may have a different position depending on the number of pixels belonging to the second group A2 and the third group A3 , but in general, the second group A2 and the third group Values between (A3) or low edgeness within the third group (A3) are formed.

Alternatively, the step of calculating the upper limit value of the binary edge threshold value (S70) does not go through the step (S60) of generating the secondary histogram H2, and is to be performed after the step (S50) of calculating the lower limit value of the binary edge threshold value. can In this case, the secondary histogram H2 is not separately generated, and among the pixels selected in step S30, the edge maps of pixels except for pixels having an edgeness less than or less than _{the lower limit value T L of the binary edge threshold value.} (edgeness) can be summed and divided by the number of pixels _{to calculate the upper limit value (T U} ) of the binary edge threshold.

In the step of selecting the binary edge threshold (S80), as shown in FIG. 8, the lower limit (T _L ) of the binary edge threshold calculated in step S50 and the upper limit (T) of the binary edge threshold calculated in step S70 _A binary edge threshold T may be chosen between U ).

In the step of selecting the binary edge threshold value (S80), the calculated lower limit value (T _L ) of the binary edge threshold value and the upper limit value (T _U ) of the binary edge threshold value are displayed, and the user selects a value between them as the binary edge threshold value ( T) can lead to selection.

Alternatively, the binary edge threshold value T may be automatically calculated based on _{the lower limit value T L} of the binary edge threshold value calculated by a predetermined algorithm and the upper limit value T _{U of the binary edge threshold value.}

As an example, the binary edge threshold value T may be calculated by the following equation.

[Equation 1]

Binary Edge Threshold (T)=(1-α)* Lower Binary Edge Threshold (T _L )+ α* Upper Binary Edge Threshold (T _U )

[Equation 2]

Binary Edge Threshold (T)= α* Lower Binary Edge Threshold (T _L )+ (1-α)* _{Upper Binary Edge Threshold (T U} )

In Equations 1 and 2, α may be input in advance by the user.

Depending on the value of α, the binary edge threshold (T) is the lower bound of the binary edge threshold (T _L ) and the binary edge threshold including the lower bound (T _L ) of the binary edge threshold and _{the upper limit (T U ) of the binary edge threshold.} It is calculated as a value between the upper limit of the value (T _{U ).}

As shown in FIG. 8 , the value of α is set to 0.4 to 0.6 or so that the _{binary edge threshold value T is approximately the middle value between the lower limit value T L} of the binary edge threshold value and the upper limit value T _{U of the binary edge threshold value.} When set within the range of 0.3 to 0.7, a binary edge threshold value T capable of distinguishing the second group A2 and the third group A3 is selected.

In the step S90 of generating the binary edge image, each pixel of the gray scale image is binary-processed as black or white based on the binary edge threshold value T selected in S80.

That is, the edgeness of each pixel of the gray scale image is compared with the binary edge threshold value T, and when the edgeness is greater than the binary edge threshold value T, the binary edge value of the corresponding pixel A grayscale image can be binarized by assigning 1 as the (binary edge value) and assigning 0 as the binary edge value of the pixel when the edgeness is less than the binary edge threshold (T). have. For a pixel whose edgeness is equal to the binary edge threshold value T, it may be set to 0 or 1 as a binary edge value of the pixel according to an algorithm for generating a binary edge image.

Alternatively, conversely, 0 is assigned as a binary edge value to a pixel whose edgeness is greater than the binary edge threshold value T, and the edgeness is smaller than the binary edge threshold value T. A gray scale image may be binarized by assigning 1 as a binary edge value to a pixel.

A binary edge value assigned as 0 or 1 for each pixel is expressed in black and white in the binary edge image.

FIG. 9 illustrates a binary edge image finally generated in step S90 of FIG. 1 .

As shown in FIG. 9 , in the finally generated binary edge image I3 , pixels having an edgeness greater than the binary edge threshold value T are expressed in white, and the edgeness is displayed as a binary edge threshold value. Pixels smaller than (T) may be expressed in black.

Or, conversely, pixels having an edgeness greater than the binary edge threshold value T may be expressed in black, and pixels having an edgeness smaller than the binary edge threshold value T may be expressed in white.

When comparing FIGS. 3, 4, 8 and 9, the part expressed in white in FIG. 9 is the part edge expressed in the lightest color in FIG. 3, and most of it is included in the third group (A3) in FIG. These are pixels, and most of them correspond to pixels included in a valid edge in a part region in FIG. 4 .

Each step-by-step process of the method for obtaining a binary edge image according to an embodiment of the present invention in the above-described embodiment can be implemented by encoding a computer-readable source code, and as a program for sequentially executing each step-by-step process Of course, it can be implemented in the form of being stored in a computer-readable 　 recording medium. Computer recording media includes all kinds of recording media in which programs and data are stored so that they can be read by a computer system, for example, ROM, RAM, CD, DVD-ROM, magnetic tape, floppy disk, optical data storage device, There is a flash memory, etc., and also includes one implemented in the form of transmission through the Internet. That is, these recording media are distributed in computer systems connected through an information and communication network, and computer-readable codes can be stored and executed in a distributed manner.

Those of ordinary skill in the art to which the present invention pertains will understand that the present invention may be embodied in other specific forms without changing the technical spirit or essential features thereof. Therefore, it should be understood that the embodiments described above are illustrative in all respects and not restrictive. The scope of the present invention is indicated by the following claims rather than the above detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalent concepts should be interpreted as being included in the scope of the present invention. do.

H1: 1st order histogram
H2: quadratic histogram
I2: grayscale image processed by Sobel operation
I3: Final generated binary edge image
T: Binary Edge Threshold
T _L : lower bound of binary edge threshold
T _U : upper bound of binary edge threshold

Claims (6)

generating a grayscale image;
selecting N pixels from a plurality of regions in the gray scale image in an order of increasing edgeness for each region;
calculating a lower limit value based on an average of edge maps of the plurality of selected pixels;
calculating an upper limit value based on an average of edge maps of the remaining pixels except for a pixel having an edge map less than or equal to the lower limit among the selected plurality of pixels;
selecting a binary edge threshold between the upper limit value and the lower limit value; and
and binary-processing the gray scale image based on the selected binary edge threshold. According to claim 1,
In the step of generating the gray scale image,
wherein the gray scale image is generated by a Sobel operation. According to claim 1,
wherein the plurality of regions are set to divide the gray scale image into equal parts. According to claim 1,
wherein the plurality of regions is each pixel column or each pixel row of the gray scale image. According to claim 1,
In the step of binary processing the gray scale image,
Among the pixels in the gray scale image, pixels having an edge degree greater than the predetermined binary edge threshold have an edge value set to 1, and pixels having an edge degree smaller than the predetermined binary edge threshold have an edge value. Binary edge image acquisition method, set to zero. A computer-readable recording medium in which a program for executing the binary edge image acquisition method according to any one of claims 1 to 5 is recorded.

Images