JPH064666A - Method and device for extracting boundary - Google Patents

Abstract

PURPOSE:To connect interrupted boundaries and to extract a desired area by integrat ing differential pictures in individual directions by means of a differential operator by individual directions, extracting a boundary picture element which does not contrib ute to division and executing a necessary extension processing. CONSTITUTION:A differential processing part 4 executes the differential processing of the picture in the vertical direction, the horizontal direction. a slanting direction to upper right and slanting direction to lower right based on the picture element which is smoothing-processed in a smoothing processing part 3, and a boundary picture generation part 5 extracts the picture element showing the peak of a differential value in a differential direction from the generated differential picture, and generates the boundary picture of the direction. CPU 6 integrates the boundary pictures in the individual directions, executes a labeling processing and extracts the picture element which does not contributed to division. A segment having prescribed length and directivity and the end point when the boundary picture element which does not contribute to division is two-dimensionally grasped are extracted. When the boundary picture element or the other segment contribute to division does not exist in the prescribed area ahead the end of the segment, the boundary is extended in the direction of the segment with the end as a starting point.

Description

Detailed Description of the Invention

[0001]

INDUSTRIAL APPLICABILITY The present invention is applied to an unmanned guided vehicle and a visual sensor portion of a shape inspection device for various products, and extracts a boundary line in an image from image information of a target object to be present. The present invention relates to an image processing technique for recognizing the shape of a target.

[0002]

2. Description of the Related Art Conventionally, when a boundary line is extracted from image information, the image information is binarized with a certain threshold value and the contour portion thereof is extracted.

Further, as a general application method to various scenes, a boundary line is extracted from image information by using a density difference between pixels, that is, a differentiation, but in that case, a single boundary line is extracted. Boundary lines are extracted using a fixed threshold of.

When these methods are used, the difference between the densities of the measurement object and the background is sufficiently large. For example, as in the case of a document image, the background is white and the character / graphic portion which is the measurement object image. A good measurement environment such as black is required.

As a method for solving these problems, for example, a differential operator for each direction is applied to the input image information to create a differential image for each direction, and a differential value peak is generated for each differential image for each direction. A method of extracting a boundary pixel from the input image information by extracting the boundary pixel from the pixels forming the condition (1) on the condition of the magnitude of the differential value and integrating the boundary pixels extracted for each direction has also been invented.

[0006]

The above-mentioned prior art has the following problems.

Cutting out the observation object in the input image, that is, extracting the area to be observed, or area division processing corresponding to various scenes that actually occur includes, for example, appearance inspection of a product. This is a necessary process for all applications and image recognition. However, when the region boundary is extracted and the region boundary is not closed curve and is interrupted, the desired region is processed. There is a problem that extraction cannot be performed.

In such a case, it is desired to connect the broken boundaries, but there is no effective method for the connection.

[0009] Therefore, an object of the present invention is to provide a method of boundary extraction which can connect a discontinuous boundary and extract a desired region in order to solve the above problem.

[0010]

In order to solve the above problems, the following methods can be considered.

Image information of an object to be measured is input, and a differential operator for each direction is acted on the image information to create a differential image for each direction, and a peak of a differential value in the differential direction for each differential image for each direction. Is checked, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted for each direction are created, and for each area surrounded by the boundary pixels in an integrated image obtained by integrating the boundary pixels extracted for each direction. Labeling is performed, the integrated image is scanned, a certain boundary pixel is focused, and in at least two directions from the focused pixel, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is regarded as a division non-contribution boundary pixel, and A pixel that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is also extracted as the boundary pixel in any other direction. When the pixel of interest is used and the differential value in one of the directions is smaller than the differential value in any of the other directions, one or two pixel blocks composed of adjacent division non-contributing boundary pixels are excluded except for the target pixel. In addition to the above extraction, an end point in the direction perpendicular to the differential direction is obtained, and a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on the axis perpendicular to the certain one direction is larger than a predetermined value for each pixel block. Extracted as a line segment, for each end point of the line segment, check the pixels existing in the predetermined region in the direction of the line segment direction, the boundary pixels contributing to the division, when not connected to other line segments, A method is conceivable in which pixels existing in a predetermined extension range in the direction of the line segment from the end point are added to the boundary pixels, and the boundary pixels extracted for each direction are integrated again.

Further, the image information of the measuring object is inputted, the direction-specific differential operator is applied to the image information to create the direction-specific differential image, and the differential value in the differential direction is applied to each direction-specific differential image. The pixel showing the peak of is examined, and when the differential value in the pixel is larger than a certain set threshold value, the pixel showing the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is created. , A label is applied to each area surrounded by the boundary pixels in the integrated image obtained by integrating the boundary pixels extracted in each direction, the integrated image is scanned, and a certain boundary pixel is focused, In at least two directions, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked,
When there are area pixels within a predetermined range from the pixel of interest in two or more directions and the labels of the area pixels are all equal, the pixel of interest is defined as a division non-contribution boundary pixel, and the boundary pixel extracted by the differentiation processing in a certain direction is Of these, the division non-contributing boundary pixels, and the pixels extracted as the boundary pixels in any other direction are the pixels of interest, and the differential value in the one direction is the differential value in the other direction. If it is smaller, one or more pixel blocks composed of adjacent division non-contribution boundary pixels are extracted excluding the pixel of interest, and end points in the direction perpendicular to the differentiation direction are obtained, and each pixel block has the above-mentioned value. The end points of each line segment are extracted by extracting a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on an axis perpendicular to one direction is larger than a specified value. Then, check the pixels existing in the predetermined area ahead of the line segment direction, when not connected to any boundary pixel, add the pixels existing in the predetermined extension range ahead of the line segment direction from the end point to the boundary pixel, A method of integrating the boundary pixels extracted for each direction again may be used.

Further, the image information of the measuring object is input, the direction-specific differential operator is acted on the image information to create the direction-specific differential image, and the differential value in the differential direction is applied to each direction-specific differential image. The pixel showing the peak of is examined, and when the differential value in the pixel is larger than a certain set threshold value, the pixel showing the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is created. , A label is applied to each area surrounded by the boundary pixels in the integrated image obtained by integrating the boundary pixels extracted in each direction, the integrated image is scanned, and a certain boundary pixel is focused, In at least two directions, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked,
When there are area pixels within a predetermined range from the pixel of interest in two or more directions and the labels of the area pixels are all equal, the pixel of interest is defined as a division non-contribution boundary pixel, and the boundary pixel extracted by the differentiation processing in a certain direction is Of these, the division non-contributing boundary pixels, and the pixels extracted as the boundary pixels in any other direction are the pixels of interest, and the differential value in the one direction is the differential value in the other direction. If it is smaller, one or more pixel blocks composed of adjacent division non-contribution boundary pixels are extracted excluding the pixel of interest, and end points in the direction perpendicular to the differentiation direction are obtained, and each pixel block has the above-mentioned value. The end points of each line segment are extracted by extracting a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on an axis perpendicular to one direction is larger than a specified value. Then, when a pixel existing in a predetermined area in the direction of the line segment is checked and it is not connected to any boundary pixel, and a boundary pixel exists in a predetermined extension range in the direction of the line segment from the end point. In addition, the pixels up to the nearest existing boundary pixel are added to the boundary pixel, and when the boundary pixel does not exist in the predetermined extension range, the processing of adding the boundary pixel is not performed, and the boundary pixels extracted for each direction are integrated again. A method of doing it is also possible.

Further, the image information of the object to be measured is input, the direction-specific differential operator is acted on the image information to create the direction-specific differential image, and the differential value in the differential direction with respect to each direction-specific differential image. The pixel showing the peak of is examined, and when the differential value in the pixel is larger than a certain set threshold value, the pixel showing the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is created. , A label is applied to each area surrounded by the boundary pixels in the integrated image obtained by integrating the boundary pixels extracted in each direction, the integrated image is scanned, and a certain boundary pixel is focused, In at least two directions, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked,
When there are area pixels within a predetermined range from the pixel of interest in two or more directions and the labels of the area pixels are all equal, the pixel of interest is defined as a division non-contribution boundary pixel, and the boundary pixel extracted by the differentiation processing in a certain direction is Of these, the division non-contributing boundary pixels, and the pixels extracted as the boundary pixels in any other direction are the pixels of interest, and the differential value in the one direction is the differential value in the other direction. If the pixel size is smaller than the target pixel, which is an end point in the direction perpendicular to the differentiation direction, a pixel block composed of adjacent division non-contribution boundary pixels is extracted, and the end point in the direction perpendicular to the differentiation direction is obtained. , A pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on an axis perpendicular to the one direction is larger than a predetermined value for each pixel block. Extracted as a segment, and for each end point of each line segment, the pixel existing in a predetermined area in the direction of the line segment is examined, and when it is not connected to the boundary pixel contributing to the division or to another line segment, It is also conceivable that a pixel existing in a predetermined extension range in the direction of the line segment from the end point is added to the boundary pixel and the boundary pixels extracted for each direction are integrated again.

Further, the image information of the measuring object is input, the direction-specific differential operator is applied to the image information to create a direction-specific differential image, and a differential value in the differential direction is created for each direction-specific differential image. The pixel showing the peak of is examined, and when the differential value in the pixel is larger than a certain set threshold value, the pixel showing the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is created. , A label is applied to each area surrounded by the boundary pixels in the integrated image obtained by integrating the boundary pixels extracted in each direction, the integrated image is scanned, and a certain boundary pixel is focused, In at least two directions, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked,
When there are area pixels within a predetermined range from the pixel of interest in two or more directions and the labels of the area pixels are all equal, the pixel of interest is defined as a division non-contribution boundary pixel, and the boundary pixel extracted by the differentiation processing in a certain direction is Of these, the division non-contributing boundary pixels, and the pixels extracted as the boundary pixels in any other direction are the pixels of interest, and the differential value in the one direction is the differential value in the other direction. If the pixel size is smaller than the target pixel, which is an end point in the direction perpendicular to the differentiation direction, a pixel block composed of adjacent division non-contribution boundary pixels is extracted, and the end point in the direction perpendicular to the differentiation direction is obtained. , A pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on an axis perpendicular to the one direction is larger than a predetermined value for each pixel block. Extracted as a segment, for each end point of the line segment, the pixels existing in the predetermined area in the direction of the line segment are examined, and when not connected to any boundary pixel, within the predetermined extension range in the direction of the line segment from the end point. A method may be considered in which existing pixels are added to the boundary pixels and the boundary pixels extracted for each direction are integrated again.

Further, the image information of the object to be measured is inputted, the direction-specific differential operator is applied to the image information to create the direction-specific differential image, and the differential value in the differential direction is applied to each direction-specific differential image. The pixel showing the peak of is examined, and when the differential value in the pixel is larger than a certain set threshold value, the pixel showing the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is created. , A label is applied to each area surrounded by the boundary pixels in the integrated image obtained by integrating the boundary pixels extracted in each direction, the integrated image is scanned, and a certain boundary pixel is focused, In at least two directions, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked,
When there are area pixels within a predetermined range from the pixel of interest in two or more directions and the labels of the area pixels are all equal, the pixel of interest is defined as a division non-contribution boundary pixel, and the boundary pixel extracted by the differentiation processing in a certain direction is Of these, the division non-contributing boundary pixels, and the pixels extracted as the boundary pixels in any other direction are the pixels of interest, and the differential value in the one direction is the differential value in the other direction. If the pixel size is smaller than the target pixel, which is an end point in the direction perpendicular to the differential direction, a pixel block composed of adjacent division non-contribution boundary pixels is extracted, end points are obtained, and A pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on an axis perpendicular to a certain direction is larger than a predetermined value is extracted as a line segment, and the end of each line segment is extracted. Regarding a pixel existing in a predetermined area in the direction of the line segment, when the pixel is not connected to any boundary pixel, and the boundary pixel exists in a predetermined extension range in the direction of the line segment from the end point. In addition, the pixels up to the nearest existing boundary pixel are added to the boundary pixel, and when the boundary pixel does not exist in the predetermined extension range, the processing of adding the boundary pixel is not performed, and the boundary pixels extracted for each direction are integrated again. A method of doing it is also possible. Furthermore, the following means can be considered as an apparatus that realizes the above method. A means for inputting image information of the measuring object, a differential processing means for operating a direction-specific differential operator to create a direction-specific differential image, and a peak of the differential value in the differential direction for each direction-specific differential image. Means for examining the image shown, and when the differential value at the pixel is larger than a certain set threshold value, the pixel showing the peak value is extracted as a boundary pixel, and the boundary pixels extracted for each direction are integrated. A boundary extraction device comprising means for labeling each area surrounded by boundary pixels in the obtained integrated image, scanning the integrated image, paying attention to a certain boundary pixel, From at least two directions, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked, and the label of the at least two directions is 2
When a region pixel exists within a predetermined range from the pixel of interest in the direction or more and the labels of the region pixel are all the same, a unit of determining the pixel of interest as a division non-contribution boundary pixel, and a process of differentiating in one direction Among the boundary pixels, a pixel that is a division non-contributing boundary pixel and is extracted as a boundary pixel in any other direction is set as a pixel of interest, and the differential value in the one direction is in any other direction. If it is smaller than the differential value in, the pixel of interest is excluded, and one or more pixel blocks composed of division non-contributing boundary pixels adjacent to the pixel of interest are extracted, and
A means for obtaining an end point in the direction perpendicular to the differential direction, and a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on the axis perpendicular to the certain one direction for each pixel block is larger than a predetermined value as a line segment Means for extracting, and at least for the end points of each line segment, check the pixels existing in the predetermined area in the direction of the line segment, and when it is not connected to the boundary pixels contributing to the division or to other line segments, A means for adding a pixel existing in a predetermined extension range in the forward direction of the line segment from the end point to a boundary pixel, or a pixel existing in a predetermined area in the forward direction of the line segment for each end point of the line segment, and determining any boundary pixel When not connected, means for adding pixels existing in a predetermined extension range in the direction of the line segment from the end point to the boundary pixel, or for each end point of the line segment, a predetermined direction in the line segment direction. When the pixel existing in the area is examined and it is not connected to any boundary pixel, and the boundary pixel exists in the predetermined extension range in the direction of the line segment from the end point, the nearest existing boundary pixel It is conceivable that a means including one of the three means of adding the pixels up to the above to the boundary pixel and not performing the processing of adding the boundary pixel when the boundary pixel does not exist in the predetermined extension range. Further, a means for inputting image information of the measurement object, a differential processing means for operating a direction-specific differential operator to create a direction-specific differential image, and a differential value in the differential direction for each differential image for each direction. A means for checking a pixel exhibiting a peak, and when a differential value in the pixel is larger than a certain set threshold value, the pixel exhibiting the peak value is extracted as a boundary pixel, and the boundary pixel extracted in each direction is extracted. A boundary extraction device comprising a unit for integrating, a unit for labeling each area surrounded by boundary pixels in the obtained integrated image, and scanning the integrated image, paying attention to a certain boundary pixel, In at least two directions from the pixel of interest, the label of the closest area pixel among the area pixels that are not the boundary pixels existing in the predetermined area is checked, and the label of the at least two directions is 2 There is a region of pixels within a predetermined range from the target pixel in the above direction, when the label of the region pixels are all equal, means for determining a pixel of interest divided not contribute boundary pixels,
Among the boundary pixels extracted by the differential processing in a certain direction, a pixel that is a division non-contribution boundary pixel and is extracted as a boundary pixel in any other direction is a pixel of interest, and When the differential value is smaller than the differential value in any other direction, a pixel block composed of adjacent division non-contribution boundary pixels is excluded except for the pixel of interest that is an end point in the direction perpendicular to the differential direction. Or, a means for obtaining end points in a direction perpendicular to the differential direction while extracting two or more, and a number of pixels when a division non-contribution boundary pixel is shaded on an axis perpendicular to the certain direction for each pixel block from a predetermined value A means for extracting a large pixel block as a line segment, and at least for each end point of each line segment, a pixel existing in a predetermined area ahead of the line segment direction is examined and divided. When it is not connected to the contributing boundary pixel or to another line segment, means for adding a pixel existing in a predetermined extension range in the direction of the line segment from the end point to the boundary pixel, or the end point of each line segment For a pixel existing in a predetermined region in the line segment direction, and when the pixel is not connected to any boundary pixel, a pixel existing in a predetermined extension range in the line segment direction from the end point is added to the boundary pixel. Alternatively, for each end point of each line segment, when a pixel existing in a predetermined region in the direction of the line segment is examined, and when it is not connected to any boundary pixel, the predetermined extension range in the direction of the line segment from the end point is detected. If there is a boundary pixel in, the pixels up to the nearest existing boundary pixel are added to the boundary pixel,
A means provided with one of the three means of not performing the processing of adding the boundary pixel when the boundary pixel does not exist in the predetermined extension range is also conceivable. Further, a means for inputting image information of the measuring object, a differential processing means for operating a direction-specific differential operator to create a direction-specific differential image, and a differential value in the differential direction for each differential image for each direction. A means for checking a pixel exhibiting a peak, and when a differential value in the pixel is larger than a certain set threshold value, the pixel exhibiting the peak value is extracted as a boundary pixel, and the boundary pixel extracted in each direction is extracted. A boundary extraction device comprising a unit for integrating, a unit for labeling each area surrounded by boundary pixels in the obtained integrated image, and scanning the integrated image, paying attention to a certain boundary pixel, The label of the region pixel closest to the pixel of interest in at least two directions, which is a pixel that is not a boundary pixel existing in a predetermined region, is checked, and 2 out of the at least two directions are checked. When there are area pixels within a predetermined range from the pixel of interest above and above, and the labels of the area pixels are all the same, means for judging the pixel of interest as a division non-contribution boundary pixel, and extraction by a differentiation process in a certain direction Of the boundary pixels, a pixel that is a division non-contribution boundary pixel and is extracted as a boundary pixel in any other direction is a pixel of interest,
When the differential value in the one direction is smaller than the differential value in any of the other directions, the pixel of interest is excluded, and one or more pixel blocks composed of adjacent division non-contributing boundary pixels are extracted, A means for obtaining an end point in a direction perpendicular to the differentiation direction, or a division non-contribution boundary pixel among the boundary pixels extracted by the differentiation processing in a certain direction and extracted as a boundary pixel in any other direction. If the differentiated value in the one direction is smaller than the differentiated value in one of the other directions, the pixel is regarded as the pixel of interest, and the pixel of interest is excluded except for the pixel of interest that is the end point in the direction perpendicular to the differentiated direction. Extract one or more pixel blocks composed of non-division-contributing boundary pixels and find the end points in the direction perpendicular to the differential direction. First selecting means for selecting one of the two means, and the number of pixels when a division non-contribution boundary pixel is shaded on an axis perpendicular to the one direction for each pixel block is a predetermined value. For extracting a larger pixel block as a line segment, and the end points of each line segment,
Pixels existing in a predetermined area in the direction of the line segment are examined, and when neither the boundary pixel contributing to the division nor other line segments are connected, the predetermined extension range in the direction of the line segment from the end point. Means to add the existing pixels to the boundary pixels, or
For each end point of each line segment, the pixels existing in a predetermined area in the direction of the line segment are checked, and when not connected to any boundary pixel, the pixels existing in a predetermined extension range in the direction of the line segment from the end point are defined as boundaries. A means for adding to a pixel, or for each end point of each line segment, the pixel existing in a predetermined area in the direction of the line segment is checked, and when it is not connected to any boundary pixel, the end point of the line segment from the end point is detected. Means for not adding a pixel up to the nearest existing boundary pixel to the boundary pixel when the boundary pixel exists in the predetermined extension range, and adding the boundary pixel when the boundary pixel does not exist in the predetermined extension range Means equipped with a second selection means for selecting one of the three means are also conceivable.

[0017]

First, the extraction of the boundary pixels and the creation of the boundary image by the differential processing will be described. First, image information of an object to be measured is input by image information input means such as a television camera, and for example, analog / digital conversion is performed. Next, in the obtained digital image, for example, vertical, horizontal,
The differential operator for each of the four directions of diagonally upward rising and diagonally rightward is applied to create a differential image for each of the four directions. The directions for obtaining the differential images for each direction are not limited to four directions. Here, there are the following methods for detecting the boundary pixels forming the differential image for each direction. A method may be considered in which, for each direction-dependent differential image, a pixel having a peak differential value in that direction is extracted, and a pixel having a differential value greater than a certain threshold value is used as a boundary pixel. To be A boundary image in which the boundary images for each direction thus obtained are integrated is created.

Next, when the boundary pixels are two-dimensionally captured,
A division non-contribution boundary pixel that is a boundary pixel that does not contribute to the division of the region is obtained without forming a closed curve due to the boundary pixel being interrupted.

First, so-called labeling processing is performed on the integrated boundary image. By this processing, labeling is performed for each closed region.

Next, the label of the area in contact with the boundary is checked to determine whether or not that portion contributes to the division of the area (if there is an area to which a different label is added, the area A division non-contribution boundary pixel that is a boundary pixel that does not contribute to the division of the area is extracted.

Whether the target pixel is a division non-contribution boundary pixel is, for example, in at least two directions of the target pixel, a region pixel exists in a predetermined region, and the labels of the region pixels are all the same. Judgment depends on whether or not.

Next, when the division non-contributing boundary pixels are two-dimensionally captured, a line segment (line structure) having a predetermined length and directionality and its end points are extracted.

The extraction is performed as follows.

First, a pixel block which is a set of division non-contribution boundary pixels which are adjacently formed with the same directionality,
And the end points in the direction are extracted. At this time, by setting the following two conditions for the division non-contributing boundary pixels extracted as boundary pixels in a plurality of directions,
A pixel block composed of pixels in the same direction can be accurately detected.

The first condition is that, in a pixel extracted as a boundary pixel in a plurality of directions, the differential value of the boundary image in the direction corresponding to the largest differential direction (for example, in the case of horizontal differential is the vertical direction). The second condition is that when a pixel extracted as a boundary pixel in a plurality of directions has a differential value in one direction smaller than a differential value in any other direction. , The pixel,
It is not to be an end point in a direction perpendicular to the certain one direction. The first condition is stricter than the second condition in terms of directional identity.

Next, if the number of pixels when the pixel block is shaded on the axis perpendicular to the differentiating direction is larger than a predetermined value set in advance for each pixel block, the pixel block is extracted as a line segment.

That is, a group of division-non-contributing boundary pixels having a predetermined length in a certain direction is used as a line segment.

By the processing up to this point, the line segment in the boundary portion that does not contribute to the division and its end points are extracted, and then a predetermined range from each end point in the direction of the line segment (the range is calculated, for example). It is also possible to set it according to the length of the line segment that has been added) to extend the boundary (that is, based on the structure of the extracted boundary) to connect the broken boundaries.

However, the end points of each line segment are only end points for individual line segments, and whether or not they are end points when the boundary is two-dimensionally grasped, that is, the base point of extension when extending the line segment. It is necessary to judge whether or not

The determination as to whether the end point becomes the extension base point can be realized by, for example, the following three methods.

First, the first method will be described.

According to the present method, if there is no boundary pixel contributing to the division or another line segment in a predetermined area beyond the end points of the line segment, even if the division pixel other than the line segment does not contribute. This is a method in which the end point is used as an extension base point even if a pixel exists.

Originally, since the line segment has a structural property that does not contribute to the division, connection or extension processing should be performed, but if the end points are connected to the boundary that already contributes to the division, extension is performed. It is not necessary to perform processing, and if the line is connected to another line segment, the extension process is performed via the line segment, and thus the extension process is not necessary in this case as well. On the other hand, if the end point is connected to a division non-contribution pixel other than a line segment, it is not extended through the division non-contribution pixel. In such a case, extension processing is performed with the end point as an extension base point. It is given. Next, the second method will be described. Unlike the first method, this method is a method in which a division non-contribution boundary pixel other than a line segment does not serve as an extension base point in a predetermined region ahead of the end point of the line segment. That is, this is a method in which the endpoint is set as an extension base point only when the endpoint is an endpoint when the boundary is two-dimensionally captured. Such a method is used because there may be a case where the pixel is connected to a boundary pixel that contributes to the division or another line segment via a division non-contribution boundary pixel other than the line segment. Furthermore, the third method is the same as the second method in the method of determining the extension base point, but is a method of adding a condition and removing unnecessary extension at the time of extension. Finally, the boundary images extracted for each direction are integrated, and the boundary extraction processing ends. As described above, by the series of processes described above, it is possible to provide a boundary extraction method that can connect a disconnected boundary and extract a desired region.

[0034]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows a block diagram of an embodiment of the present invention. In this embodiment, a TV camera 1, an A / D converter 2, a smoothing processing unit 3, a differentiation processing unit 4, a boundary image creating unit 5, a CPU 6 and an RA.
It is configured to include an M7 and a ROM8.

The TV camera 1 is a means for collecting the luminance information which is the image information of the measuring object. In this embodiment,
Although the NTSC signal of the TV camera is used as the image information, the reflection intensity information such as ultrasonic waves and radio waves may be used as the image information. The A / D converter 2 is a circuit that converts an analog signal into a digital signal. The smoothing processing unit 3 is a unit that removes various noises included in the image in order to facilitate feature extraction from the image, and is configured by an electronic device such as various TTL logics.

The differential processing section 4 is means for calculating the difference in luminance information in order to facilitate the extraction of boundaries from the image, and is composed of electronic devices such as various TTL logics.

The boundary image creating section 5 extracts from the differential image for each direction obtained by the differential processing section 4 a pixel whose differential value shows a peak in the differential direction, and the differential value at the pixel is preset. In addition, it is a means for creating a boundary image with a pixel having a value larger than a certain threshold as a boundary pixel, and is composed of, for example, electronic devices such as various TTL logics, clock circuits, and delay circuits.

In the present embodiment, since the boundary image in each direction is created from the differential images in the four directions of the vertical direction, the horizontal direction, the diagonally upward rightward direction, and the diagonally downward rightward direction, each direction is supported. Then, it has four boundary image creation units 5.

In this embodiment, the differential direction is obtained by setting the differential directions to four directions, but the differential direction is not limited to this.

The CPU 6 is a means for performing a series of processing such as integration of boundary pixels, labeling processing, judgment of division non-contributing boundary pixels, extraction of pixel blocks and end points, extraction of line segments, and extension processing, which will be described below. , A semiconductor microprocessor, for example. RAM7 is CPU6
Is a means for temporarily storing various kinds of image processing data, calculation data, etc. processed by the above, and is realized by, for example, a semiconductor element.

The ROM 8 is a means for storing a series of programs processed by the CPU 6, and is realized by, for example, a semiconductor device.

Although not shown, for example, the CPU 6
May be connected to a VRAM, a display device or the like via an interface circuit or the like, and the image processing result may be output to the display device. At this time, the display device may be, for example, a CRT, a liquid crystal display, an EL display, or the like.

FIG. 5 is a process diagram showing the flow of the entire processing.

Hereinafter, with reference to FIG. 1 to FIG. 4, first, the process of creating an integrated image from the boundary image for each direction will be described (steps 10 to 40). First, T
The V camera 1 reads the brightness information of the measuring object. As the image information, for example, an NTSC signal output from a TV camera is used.

Next, the A / D converter 2 converts the analog NTSC signal into a digital signal. Next, the smoothing processing unit 3 performs image smoothing processing based on the obtained digital signal. Here, the smoothing process is a process of removing various noises included in the image in order to facilitate feature extraction from the image. An example of the smoothing process will be described with reference to FIG. 2 (step 10).
Now, considering a 3 × 3 matrix, let the value of the pixel point of interest be P, and let the values of the pixel points around P be A as shown in FIG.
0, A1, A2, A3, A4, A5, A6, A7. Here, P given by the following expression 1 is a smoothing processing value at the pixel point of interest.

P = (A0 + A1 + A2 + A3 + A4 + A5 + A6 + A7) / 8 (Equation 1) The smoothing processing unit 3 performs this smoothing processing with all the pixel points as attention points.

Next, the differential processing section 4 performs differential processing of the image based on the smoothed pixels. Here, the differential processing is arithmetic processing applied to an image in order to facilitate boundary extraction. An example of the differential processing performed in step 20 will be described with reference to FIG.

Considering a 5 × 5 matrix, let P be the value of a pixel point of interest, and let P0, P1 be the values of some pixel points around P, as shown in FIG. P2, P
3, P4, P5, P6 and P7. Also, the “weight matrix” for performing the differential processing is also a 5 × 5 matrix, and the values at the pixel points necessary for the differential processing are shown in FIG.
As shown in (a), A0, A1, A2, A3, A4,
A5, A6 and A7. Here, P given by the following equation 2 is a differential value at the pixel point of interest.

P = P0 × A0 + P1 × A1 + P2 × A2 + P3 × A3 + P4 × A4 + P5 × A5 + P6 × A6 + P7 × A7 (Equation 2) Specific values of A0, A1, A2, A3, A4, A5, A6, and A7 are vertical. The direction, the lateral direction, the diagonally upward rightward direction, and the diagonally downward rightward direction are determined as shown in FIG. 3B. For example, in the case of horizontal differential processing, A
0 = 0, A1 = -1, A2 = -2, A3 = -1, A4 =
Since 0, A5 = 1, A6 = 2, and A7 = 1, the differential value P at the target point can be expressed by the following Expression 3 from Expression 2.

P = P0 × 0 + P1 × (-1) + P2 × (-2) + P3 × (-1) + P4 × 0 + P5 × 1 + P6 × 2 + P7 × 1 Therefore, P = −P1-2 × P2-P3 + P5 + 2 × P6 + P7 ( Formula 3) This differentiation process is performed on all pixel points. Different processing in the vertical direction, the diagonally upward rightward direction, and the diagonally downward rightward direction can be performed by similar processing. next,
From the created differential image, pixels showing the peak of the differential value in the differential direction are extracted to create a boundary image in that direction (step 30). This processing is executed by the boundary image creating unit 5, which will be described with reference to FIG. The description will be made by taking the lateral direction differential image as an example, but exactly the same processing is performed on the differential images in the other three directions. As described above, the differential value for each pixel is obtained by Expression 3 in the image differentially processed in the horizontal direction by the differential processing unit 4. By the way, the fact that the differential value shows a peak means that the change of the luminance information becomes maximum, and the point showing this peak value is very likely to be a point on the boundary. Therefore, the change in the differential value is examined to detect a pixel whose differential value shows a peak (hereinafter, simply referred to as “peak pixel”). A certain pixel point P of interest
However, in order to be a peak pixel, as shown in FIG.
When 0, P, and P1 are set, the condition is that the following expression 4 is satisfied.

P> P0 and P> P1 (Equation 4) By the way, even if the differential value itself is small, if Equation 4 is satisfied,
Although it is a peak pixel, a small differential value means that a change in luminance information is small, and the peak pixel is less likely to be a pixel point on the boundary. Therefore, the differential value at the obtained peak pixel is compared with a certain set threshold value, and when the differential value is larger than the threshold value, the peak pixel is regarded as a boundary pixel, Create a border image. Further, since a pixel that is not a peak pixel is not a boundary pixel, all the pixels that make up the input image are classified into a boundary pixel and a non-boundary pixel, and eventually binarized into a boundary pixel and a non-boundary pixel. An image is obtained (step 40). By the series of processes described above, it is possible to create a boundary image in four directions, which means that the processes from step 10 to step 40 in FIG. 5 have been performed.
In step 50, a process of integrating the boundary images in four directions is performed.

It should be noted that the image obtained by the integration is subjected to other image processing (for example, processing such as removal of a texture area which is unnecessary for boundary line extraction) to prepare for processing starting from the next step. Is also good.

Next, in step 60, so-called area labeling processing is performed.

Here, in general, the "labeling process" means
When a plurality of figures (areas) are present in the obtained image, different label values are added to each connected component of each area. An example of the result of such labeling processing is
As shown in FIG. The example shown in FIG. 6 is a case where the boundary line is a closed curve at two places, and different labels “1”, “2”, and “3” are added to the respective regions. The labeling process is usually performed in three stages of a temporary labeling process, a table operation process, and a main labeling process.

The labeling processing means includes, for example, a CPU for labeling processing, a RAM for temporary labeling processing, a RAM for table operation processing, a RAM for main labeling processing, and the like. In the present embodiment, the CPU 6 is used.
Includes a labeling processing CPU, and the RAM 7 is configured to include a temporary labeling processing RAM, a table operation processing RAM, a main labeling processing RAM, and the like, and a processing program for the labeling processing is , ROM8.

The labeling process will be briefly described below. First, the input image is raster-scanned, and a process of sequentially assigning unused labels to non-boundary pixels to which no label is assigned is performed up to the final scan pixel (referred to as temporary labeling process).

At this time, if it is found that pixels with different labels are connected, label connection information is created.

Whether or not pixels with different labels are connected is determined by, for example, whether or not there is a non-boundary pixel with another label in the vicinity of 8 pixels of the pixel of interest.

Next, a label value correction table is created based on the label connection information (referred to as table operation processing).

Finally, using the label value correction table,
An actual labeling image is created from the temporary labeling image (referred to as an actual labeling process).

The details of the labeling processing method are described in more detail, for example, in "Introduction to Computer Image Processing" (Soken Shuppan Co., Ltd .: supervised by Hideyuki Tamura).

Next, at step 70, the division non-contribution boundary pixel extraction processing, that is, division contribution determination is performed. As a specific method for examining the label of the area in contact with the boundary and determining whether or not the portion contributes to the area division (that is, whether or not there is an area to which a different label is added), There are the following: 2 of a certain pixel of interest
If there are area pixels in a predetermined range in the direction or more and the labels of the area pixels are all equal (that is, they do not touch areas to which different labels are added), the pixel of interest is
It is determined that the boundary pixel does not contribute to the division and is a division non-contribution boundary pixel. As the direction to be considered here, for example, the up, down, left, and right directions can be considered.

Here, the area pixel means a pixel which is not a boundary pixel, and the predetermined range may be set as "the number of pixels forming the width of the boundary + α". , When the number of pixels constituting the width of the boundary is 3 and α = 2,
It should be 5 pixels. Further, when the boundary extends in one direction, all the pixels existing within the predetermined range in that direction are boundary pixels, and in this direction, they do not touch the area.

FIG. 7 shows an example in which the pixel is judged to be a division non-contribution boundary pixel by the above method.

In this case, the labels of the area pixels in the four directions of the division non-contributing boundary pixels are the same ("2").

When a region pixel exists within a predetermined range in one or less of the upper, lower, left, and right directions of the pixel of interest, an algorithm for determining the pixel of interest as a division non-contribution boundary pixel is added to divide the division non-contribution boundary. It may be a process of extracting pixels.

Next, in step 80, a process of extracting connected pixel blocks and end points is performed on the four-direction boundary images.

This process is a process for extracting a pixel block and end points which are composed of connected non-contributing boundary pixels.

There are two possible processes to be performed in this step, which will be described below.

First, the first method will be described.

A vertical boundary pixel (differential direction is horizontal) will be described as an example. Here, the processing target is extracted as a boundary pixel in other directions among the division non-contribution boundary pixels obtained in step 70, and the differential value in that direction is the target direction (in this case, the horizontal direction). Pixels larger than the differential value in the (direction) are excluded.

First, the image is scanned from bottom to top from left to right, and if the target pixel exists, the target pixel exists in the already scanned four pixels located at the bottom right, bottom, bottom left, and left. If not, the pixel is set as the lower end point in the vertical direction (condition 1).

When the target pixel exists in the four pixels located at the lower right, lower, lower left, and left, it is determined whether the left pixel is the lower end point in the vertical direction (of course, the target pixel). If the left pixel is the end point, the lower right pixel is further checked, and if the lower right pixel is not the target pixel, the target pixel is also set as the end point (condition 2).

If the lower right pixel is the target pixel, the image is scanned leftward until it hits a pixel that is not an end point, and all pixels are not end points (condition 3).

The above processing will be specifically described with reference to FIGS. 8A and 8B as follows.

First, in FIG. 8A, when P is the target pixel, P is an end point when P1, P2, P3, and P4 are not the target pixels. When P1 is the end point and P4 is not the target pixel, P becomes the end point. If P1 is an end point and P4 is the target pixel, all end points (including P1) that are in contact with the left side of P are not end points.

Next, in FIG. 8B, A is condition 1
Is the end point. B becomes an end point under the condition 2. C
The conditions 1 and D are end points according to the condition 2, but when the E is scanned, both C and D are not end points due to the condition 3. Note that the hatched portion in the drawing indicates the region formed by the target pixel.

Next, the image is sequentially scanned from the left to the right, and the target pixel is subjected to the labeling process. Since the labeling process is as described above, detailed description thereof is omitted here to avoid redundant description. Note that, during the labeling process, when there is no labeled pixel among the already scanned four pixels on the upper right, upper side, upper left side, and left side, it is assumed that the target pixel is the upper end point in the vertical direction.

When the labeled pixel is present in the four pixels existing in the upper right, upper, upper left, and left, it is checked whether or not the left pixel is the upper end point in the vertical direction, and the left is the end point. Then, the upper right is further examined, and if the upper right is not the labeled pixel, the pixel of interest is also set as the end point.

If the upper right pixel is a labeled pixel, the image is scanned leftward until a pixel that is not an end point and all the pixels are not end points.

By the above processing, the upper and lower end points of the pixel blocks to be connected are obtained at the same time.

Of course, the above processing may be performed by scanning from above, or each scanning may be performed from right to left. Further, the above-mentioned processing has been described for the boundary image by the lateral differentiation,
The same processing may be performed for the vertical direction, the diagonally upward rightward direction, and the diagonally downward rightward direction.

Of course, the boundary pixels extracted by the differential processing in one direction are also extracted as the boundary pixels in any other direction, and the differential value in any other direction is as described above. A method may be considered in which, even if the pixel is larger than the differential value in one direction, the pixel whose differential value in the certain one direction is a predetermined value or more is not excluded, and is determined as the connection determination target.

Next, the second method will be described.

In this method, a constraint condition is set for the end points, and the connected pixel blocks and the end points thereof are obtained at the same time.

According to the present method, when a pixel which is a boundary pixel extracted by a differentiating process in one direction and is also a division non-contribution boundary pixel is also extracted as a boundary pixel in any other direction, The pixel is the pixel of interest,
When the differential value in the one direction is smaller than the differential value in any of the other directions, the labeling is performed while making a determination on the condition that the target pixel which is an end point in the direction perpendicular to the differential direction is excluded. This is a method of performing processing to extract pixel blocks and end points, and a method of obtaining connected pixel blocks and end points with a constraint condition for the end points.

Specifically, the following processing is performed. Note that the labeling process itself will not be described in detail because it is not duplicated. A vertical boundary pixel (differential direction is horizontal) will be described as an example. First, for example, the given boundary image is scanned from left to right in order from the bottom, and when there are division non-contribution boundary pixels, the lower right, lower, lower left, and left four pixels that have already been scanned. When no black pixel (which is a pixel indicating “1” in the digital binarized image) does not exist therein, among the division non-contribution boundary pixels, it is extracted as a boundary pixel also in the other differential direction, and in that direction. A pixel other than a pixel whose differential value is larger than the differential value in the target differential direction (horizontal direction) is set as a black pixel, and at the same time, the pixel is set as a lower end point in the vertical direction.

When there are black pixels in the lower right, lower, lower left, and left four pixels, the pixels that do not contribute to division are black pixels, and the left pixel is the lower end point in the vertical direction (the end point is a black pixel). If the left pixel is the end point, the lower right pixel is further checked, and if the lower right pixel is not a black pixel, the target pixel is also set as the end point. If the lower right pixel is a black pixel, the image is scanned leftward until all the pixels are not end points, and all pixels are not end points. Next, for example, the given boundary image is scanned from left to right in order from the top, and among black pixels, it is extracted as a boundary pixel in other directions, and the differential value in that direction is noted. Pixels that are not larger than the differential value in the differential direction (horizontal direction) and pixels that are extracted as boundary pixels in other directions and whose differential value in that direction is larger than the differential value in the target differential direction. , Upper right, already scanned,
The labeling process is performed on the target pixels in which the labeled pixels are present among the upper, upper left, and left four pixels. The labeling process itself may be the same process as described above. At this time, if there is no labeled pixel among the four pixels that have already been scanned on the upper right, upper side, upper left side, and left side, the pixel of interest is regarded as the upper end point in the vertical direction.

When the labeled pixel is present in the four pixels located on the upper right, upper, upper left, and left, it is checked whether the left pixel is the upper end point in the vertical direction, and the left pixel is If it is an end point, the upper right pixel is further checked, and if the upper right pixel is not the labeled pixel, the target pixel is also set as the end point.

If the upper right pixel is a labeled pixel, the image is scanned leftward until the pixels that are not the end points, and all the pixels are not the end points.

By the above processing, the upper and lower end points in the vertical direction are extracted as boundary pixels in other directions, and the differential value in that direction is not a pixel larger than the differential value in the target direction. Blocks can be extracted.

At the same time, its end points are also obtained. Of course, the preceding process may be a scan from above, the subsequent process may be a scan from below, and the scan itself may be a scan from right to left. Although the above-described processing has been described with respect to the horizontal differentiation, pixel blocks are extracted by the same processing in the vertical direction, the diagonally rightwardly upward direction, and the diagonally rightwardly downward direction.

In the second method, the boundary pixel extracted by the differential processing in one certain direction and also the division non-contribution boundary pixel is extracted as the boundary pixel in any other direction. And the above 1
A pixel whose differential value in one direction is smaller than the differential value in any other direction does not become an end point in the direction perpendicular to the differential direction, but the differential value in the certain one direction is predetermined. A method of extracting pixel blocks and end points by allowing pixels having a value or more to be end points may be considered. According to this method, more stable extraction becomes possible.

Further, in the first and second methods, the boundary pixels extracted from the magnitude of the differential value in the peak pixel of the differential value are generally thinned (that is,
Since the line width is 1 pixel), the method of extracting the end points is simplified,
For example, a method of extracting the end point under the condition that the lower end point in the vertical direction is “an end point when the lower left, lower, and lower right three pixels are not the target pixel” can be considered.

Next, a process for extracting a block in which each pixel block has a pixel number larger than a predetermined value when shaded on an axis perpendicular to the differential direction as a "line segment" (step 9)
0). An example of this processing is shown in FIG. This processing example is an example of the result of extracting a line segment from the differential image in the horizontal direction. Pixels hatched with diagonal lines form a pixel block. The number of pixels when shaded in the direction (vertical direction) perpendicular to the differentiation direction is 4 pixels. For example, when the threshold value of the number of pixels for determining whether or not it is a line segment is set to 3 pixels in advance, a line segment in the vertical direction is extracted from the pixel block shown in FIG.

Next, the extension base point extraction and extension processing performed in steps 100 and 110 will be described collectively.

By the way, the end points of each line segment are only end points for individual line segments, and whether or not they are end points when the boundary is two-dimensionally grasped, that is, the base point of extension when extending the line segment. It is necessary to judge whether or not

The determination as to whether the end point becomes the extension base point can be realized by, for example, the following three methods.

First, the first method will be described. In this method, a boundary pixel contributing to the division is present in a predetermined area beyond the end point of the line segment, or if there is no other line segment, a non-contribution pixel other than the line segment exists. Even if it is done, it is a method of using the end point as an extension base point.

The "inside predetermined area" will be described with reference to FIG.

FIGS. 10A, 10B, 10C and 10D.
The end points in the vertical direction (horizontal differential), the horizontal direction (vertical differential), the diagonally right upward direction (diagonal rightward downward direction), and the diagonally right downward direction (diagonal rightward upward direction) are specified by An example in the area is shown. The hatched pixels are an example within a predetermined area ahead of the end point of the line segment in each direction (three pixels per end point in the present embodiment, but not limited to this). In addition, within a predetermined area beyond the end points of the line segment,
In the second and third methods, the same one is used as an example.

By the way, since the line segment originally has a structure that does not contribute to the division, connection processing, that is, extension processing should be performed, but if the end points are connected to the boundaries that have already contributed to the division. If it is connected to another line segment, the extension process is performed through the line segment, so that the extension process is not necessary in this case either.

On the other hand, when the end point is connected to a division non-contribution pixel other than a line segment, it is not extended through the division non-contribution pixel. In such a case, the end point is set as an extension base point. The extension process is performed. The extension range extended from the extension base point is obtained in, for example, step 90,
It may be set according to the length of the line segment (the number of pixels when it is shaded on the axis perpendicular to the differentiation direction). For example, the length of the line segment is 4
It is conceivable that the extension range is set to one-tenth.

For the sake of simplicity of processing, the extension range may be a constant value regardless of the length of the line segment. The concept of the extension range is similarly used in the second and third methods described below.

Next, the second method will be described. This method is different from the first method in that a predetermined non-contribution boundary pixel other than the line segment exists in a predetermined area ahead of the end point of the line segment, but the end point is not the extension base point. is there.

That is, this is a method in which the end point is set as an extension base point only when the boundary is two-dimensionally regarded as an end point. Through division non-contribution boundary pixels other than line segments,
This method is used because it may be connected to a boundary pixel that contributes to the division or another line segment. Furthermore, the third method is the same as the second method in the method of determining the extension base point, but is a method of adding a condition and removing unnecessary extension at the time of extension.

That is, when the extension processing is performed from the extension base point, no processing is performed when no boundary pixel exists within the predetermined extension range, and the pixels up to the closest boundary pixel existing within the predetermined extension range are determined. This is a process of adding to the boundary pixel. Finally, in step 120, step 1
The boundary images obtained by the processes up to 10 are calculated for each direction, and the entire boundary image is created. The integration process is performed by, for example, an OR operation.

The processing from steps 60 to 120 is also performed by the CPU 6, ROM 8 and RAM 7.

By the series of processing described above, it is possible to provide a valid boundary extraction processing method by connecting the broken boundaries to extract a predetermined area.

An apparatus using this boundary extraction method is, for example,
By installing in an obstacle avoidance sensor or the like of an automated guided vehicle, it is possible to provide an automated guided vehicle that can avoid obstacles with high accuracy. However, the application example is not limited to the automatic guided vehicle and can be applied to various systems.

[0111]

As described above, according to the present invention, the direction-dependent differential operator is operated to integrally create the direction-specific differential images, the division non-contribution boundary pixels are extracted, and the necessary extension processing is performed to cause interruption. A boundary extraction method capable of connecting boundaries and extracting a desired boundary can be provided.

[Brief description of drawings]

FIG. 1 is an explanatory diagram of a configuration of an embodiment of the present invention.

FIG. 2 is an explanatory diagram of smoothing processing.

FIG. 3 is an explanatory diagram of a differential operator.

FIG. 4 is an explanatory diagram of peak image extraction.

FIG. 5 is an explanatory diagram according to a flowchart of a processing example of the present invention.

FIG. 6 is an explanatory diagram of an example of labeling processing.

FIG. 7 is an explanatory diagram of a process of extracting a division non-contribution boundary pixel.

FIG. 8 is an explanatory diagram of a processing example of the present invention.

FIG. 9 is an explanatory diagram of line segment extraction processing.

FIG. 10 is an explanatory diagram of extension base point extraction processing.

[Explanation of symbols]

1 ... TV camera, 2 ... A / D converter, 3 ... Smoothing processing path,
4 ... Differentiation processing unit, 5 ... Boundary image creation unit, 6 ... CPU, 7
... RAM, 8 ... ROM

Claims (9)

[Claims] 1. Inputting image information of an object to be measured, applying a direction-specific differential operator to the image information to create a direction-specific differential image, and for each direction-specific differential image, a differential value in the differential direction. The pixel showing the peak of is examined, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted by direction are created, and the boundary pixels extracted by direction are integrated. Labeling is performed, the integrated image is scanned, attention is paid to a certain boundary pixel, and in at least two directions from the pixel of interest, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is defined as a division non-contribution boundary pixel, and An image that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is extracted as a boundary pixel in any other direction. Is a pixel of interest, and the differential value in the one direction is smaller than the differential value in any of the other directions, the pixel block excluding the target pixel is defined as one or more pixel blocks composed of adjacent division non-contribution boundary pixels. A pixel block in which two or more are extracted and the end points in the direction perpendicular to the differential direction are obtained, and the number of pixels when the division non-contribution boundary pixel is shaded on the axis perpendicular to the certain one direction for each pixel block is larger than a predetermined value. Is extracted as a line segment, and for each end point of the line segment, the pixels existing in a predetermined area in the direction of the line segment are examined, and even for the boundary pixels contributing to the division,
When it is not connected to other line segments, pixels existing in a predetermined extension range in the direction of the line segment from the end point are added to the boundary pixels, and the boundary pixels extracted for each direction are integrated again to input the image information. Boundary extraction method to extract the boundary from the image. 2. Inputting image information of an object to be measured, applying a direction-specific differential operator to the image information to create a direction-specific differential image, and for each direction-specific differential image, a differential value in the differential direction. The pixel showing the peak of is examined, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted by direction are created, and the boundary pixels extracted by direction are integrated. Labeling is performed, the integrated image is scanned, attention is paid to a certain boundary pixel, and in at least two directions from the pixel of interest, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is defined as a division non-contribution boundary pixel, and An image that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is extracted as a boundary pixel in any other direction. Is a pixel of interest, and the differential value in the one direction is smaller than the differential value in any of the other directions, the pixel block excluding the target pixel is defined as one or more pixel blocks composed of adjacent division non-contribution boundary pixels. A pixel block in which two or more are extracted and the end points in the direction perpendicular to the differential direction are obtained, and the number of pixels when the division non-contribution boundary pixel is shaded on the axis perpendicular to the certain one direction for each pixel block is larger than a predetermined value. Is extracted as a line segment, and for each end point of the line segment, the pixels existing in a predetermined region in the direction of the line segment are checked. When not connected to any boundary pixel, a predetermined extension in the direction of the line segment from the end point is extracted. A boundary extraction method in which pixels existing in the range are added to the boundary pixels, and the boundary pixels extracted for each direction are integrated again to extract the boundaries from the input image information. 3. Inputting image information of an object to be measured, applying a direction-specific differential operator to the image information to create a direction-specific differential image, and for each direction-specific differential image, a differential value in the differential direction. The pixel showing the peak of is examined, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted by direction are created, and the boundary pixels extracted by direction are integrated. Labeling is performed, the integrated image is scanned, attention is paid to a certain boundary pixel, and in at least two directions from the pixel of interest, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is defined as a division non-contribution boundary pixel, and An image that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is extracted as a boundary pixel in any other direction. Is a pixel of interest, and the differential value in the one direction is smaller than the differential value in any of the other directions, the pixel block excluding the target pixel is defined as one or more pixel blocks composed of adjacent division non-contribution boundary pixels. A pixel block in which two or more are extracted and the end points in the direction perpendicular to the differential direction are obtained, and the number of pixels when the division non-contribution boundary pixel is shaded on the axis perpendicular to the certain one direction for each pixel block is larger than a predetermined value. Is extracted as a line segment, and for each end point of the line segment, the pixels existing in a predetermined area in the direction of the line segment are examined, and when the boundary pixel is not connected to the end point in the direction of the line segment from the end point. If there is a boundary pixel in the specified extension range, the pixels up to the closest existing boundary pixel are added to the boundary pixel, and if there is no boundary pixel in the specified extension range, Without the process of adding the element, the boundary pixels extracted by direction again, the boundary extraction method for extracting boundaries from integration to the input image information. 4. A differential value in a differential direction for each direction differential image is created by inputting image information of a measuring object, applying a direction-specific differential operator to the image information, and creating a direction-specific differential image. The pixel showing the peak of is examined, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted by direction are created, and the boundary pixels extracted by direction are integrated. Labeling is performed, the integrated image is scanned, attention is paid to a certain boundary pixel, and in at least two directions from the pixel of interest, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is defined as a division non-contribution boundary pixel, and An image that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is extracted as a boundary pixel in any other direction. If the differential value in the one direction is smaller than the differential value in any one of the other directions, the target pixel that is the end point in the direction perpendicular to the differential direction, A pixel obtained by extracting a pixel block composed of contributing boundary pixels, finding the end points in the direction perpendicular to the differentiation direction, and dividing each non-contribution boundary pixel along the axis perpendicular to the one direction. A pixel block whose number is larger than a predetermined value is extracted as a line segment, and for the end points of each line segment, the pixels existing in a predetermined region in the direction of the line segment are examined, and even for boundary pixels contributing to the division,
When it is not connected to other line segments, pixels existing in a predetermined extension range in the direction of the line segment from the end point are added to the boundary pixels, and the boundary pixels extracted for each direction are integrated again to input the image information. Boundary extraction method to extract the boundary from the image. 5. Inputting image information of a measuring object, applying a direction-specific differential operator to the image information to create a direction-specific differential image, and for each direction-specific differential image, a differential value in the differential direction. The pixel showing the peak of is examined, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted by direction are created, and the boundary pixels extracted by direction are integrated. Labeling is performed, the integrated image is scanned, attention is paid to a certain boundary pixel, and in at least two directions from the pixel of interest, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is defined as a division non-contribution boundary pixel, and An image that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is extracted as a boundary pixel in any other direction. If the differential value in the one direction is smaller than the differential value in any one of the other directions, the target pixel that is the end point in the direction perpendicular to the differential direction, A pixel obtained by extracting a pixel block composed of contributing boundary pixels, finding the end points in the direction perpendicular to the differentiation direction, and dividing each non-contribution boundary pixel along the axis perpendicular to the one direction. A pixel block whose number is larger than a predetermined value is extracted as a line segment, and for the end points of each line segment, the pixels existing in a predetermined area ahead of the line segment direction are examined. Pixels existing in the predetermined extension range in the direction of the line segment are added to the boundary pixels, and the boundary pixels extracted for each direction are integrated again to extract the boundaries from the input image information. Boundary extraction method. 6. Inputting image information of an object to be measured, applying a direction-specific differential operator to the image information to create a direction-specific differential image, and for each direction-specific differential image, a differential value in the differential direction. The pixel showing the peak of is examined, and when the differential value at the pixel is larger than a certain set threshold value,
Pixels indicating the peak value are extracted as boundary pixels, boundary pixels extracted by direction are created, and the boundary pixels extracted by direction are integrated. Labeling is performed, the integrated image is scanned, attention is paid to a certain boundary pixel, and in at least two directions from the pixel of interest, the nearest area pixel of the area pixels that are not the boundary pixels existing in the predetermined area The label is examined, and when there are area pixels within a predetermined range from the pixel of interest in two or more directions among the at least two directions and the labels of the area pixels are all the same, the pixel of interest is defined as a division non-contribution boundary pixel, and An image that is a division non-contribution boundary pixel among the boundary pixels extracted by the differential processing in the direction and is extracted as a boundary pixel in any other direction. If the differential value in the one direction is smaller than the differential value in any one of the other directions, the target pixel that is the end point in the direction perpendicular to the differential direction, A pixel obtained by extracting a pixel block composed of contributing boundary pixels, finding the end points in the direction perpendicular to the differentiation direction, and dividing each non-contribution boundary pixel along the axis perpendicular to the one direction. When a pixel block whose number is larger than a predetermined value is extracted as a line segment, and the end point of each line segment is examined for pixels existing in a predetermined region ahead of the line segment direction, and when it is not connected to any boundary pixel, When a boundary pixel exists in a predetermined extension range in the direction of the line segment from the end point, the pixels up to the closest existing boundary pixel are added to the boundary pixel to define a boundary in the predetermined extension range. Without processing to add boundary pixel if the pixel does not exist, a boundary extracting method for extracting a boundary boundary pixels extracted by direction again from the integration to the input image information. 7. A means for inputting image information of an object to be measured,
A differential processing means for operating a direction-specific differential operator to create a direction-specific differential image, a means for examining a pixel showing a peak of a differential value in the differential direction in each direction-specific differential image, and a differential value in the pixel Is a threshold value greater than a certain set threshold value, the pixel indicating the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is integrated. A unit for labeling each area surrounded by boundary pixels in the acquired integrated image, scanning the integrated image, focusing on a certain boundary pixel, and existing in a predetermined area in at least two directions from the target pixel The label of the closest area pixel among the area pixels that are not the boundary pixels is checked, and the area within the predetermined range from the pixel of interest in at least two directions out of the at least two directions. When a pixel exists and the labels of the area pixels are all the same, a method of determining the pixel of interest as a division non-contribution boundary pixel, and a division non-contribution boundary pixel among the border pixels extracted by the differential processing in one direction If the pixel extracted as the boundary pixel in any other direction is the pixel of interest, and the differential value in the one direction is smaller than the differential value in any of the other directions, the pixel of interest is selected. Except for extracting one or two or more pixel blocks composed of adjacent division non-contributing boundary pixels, and obtaining an end point in the direction perpendicular to the differentiation direction, and an axis perpendicular to the one direction for each pixel block. Means for extracting a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded is larger than a predetermined value as a line segment, and at least the end points of each line segment. Then, a pixel existing in a predetermined region in the direction of the line segment is examined, and when it is not connected to a boundary pixel contributing to the division or to another line segment, a predetermined amount in the direction of the line segment from the end point is determined. A means for adding a pixel existing in the extension range to a boundary pixel, or an end point of each line segment is checked for a pixel existing in a predetermined area in the direction of the line segment, and when not connected to any boundary pixel, from the end point A means for adding a pixel existing in a predetermined extension range in the direction of the line segment to the boundary pixel, or for each end point of the line segment, checking the pixel existing in the predetermined region in the direction of the line segment, and connecting to any boundary pixel. When there is a boundary pixel in the predetermined extension range in the direction of the line segment from the end point, the pixels up to the nearest existing boundary pixel are added to the boundary pixel, and the boundary pixel is included in the predetermined extension range. exist A boundary extraction device comprising one of three means that does not perform processing of adding a boundary pixel when not doing so. 8. A means for inputting image information of a measuring object,
A differential processing means for operating a direction-specific differential operator to create a direction-specific differential image, a means for examining a pixel showing a peak of a differential value in the differential direction in each direction-specific differential image, and a differential value in the pixel Is a threshold value greater than a certain set threshold value, the pixel indicating the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is integrated. A unit for labeling each area surrounded by boundary pixels in the acquired integrated image, scanning the integrated image, focusing on a certain boundary pixel, and existing in a predetermined area in at least two directions from the target pixel The label of the closest area pixel among the area pixels that are not the boundary pixels is checked, and the area within the predetermined range from the pixel of interest in at least two directions out of the at least two directions. When a pixel exists and the labels of the area pixels are all the same, a method of determining the pixel of interest as a division non-contribution boundary pixel, and a division non-contribution boundary pixel among the border pixels extracted by the differential processing in one direction If the pixel extracted as the boundary pixel in any other direction is the pixel of interest and the differential value in the one direction is smaller than the differential value in the other direction, the pixel is differentiated. Except for the pixel of interest that is an end point in the direction perpendicular to the direction, one or more pixel blocks composed of adjacent division non-contribution boundary pixels are extracted, and means for determining the end point in the direction perpendicular to the differential direction, For each pixel block, a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on an axis perpendicular to the one direction is larger than a predetermined value is extracted as a line segment. Means for outputting, and at least for the end points of each line segment, check the pixels existing in a predetermined area in the direction of the line segment, and when it is not connected to the boundary pixel contributing to the division or to another line segment, A means for adding a pixel existing in a predetermined extension range in the forward direction of the line segment from the end point to a boundary pixel, or a pixel existing in a predetermined area in the forward direction of the line segment for each end point of the line segment, and determining any boundary pixel When not connected, means for adding a pixel existing in a predetermined extension range in the direction of the line segment from the end point to a boundary pixel, or for each end point of the line segment, existing in a predetermined region in the direction of the line segment ahead When the pixel is examined and it is not connected to any boundary pixel, and the boundary pixel exists in the predetermined extension range in the direction of the line segment from the end point, the pixels up to the nearest existing boundary pixel are selected. In addition to the boundary pixel, the boundary extracting device is provided with one of three means that does not perform processing of adding the boundary pixel when the boundary pixel does not exist in the predetermined extension range. 9. A means for inputting image information of a measuring object,
A differential processing means for operating a direction-specific differential operator to create a direction-specific differential image, a means for examining a pixel showing a peak of a differential value in the differential direction in each direction-specific differential image, and a differential value in the pixel Is a threshold value greater than a certain set threshold value, the pixel indicating the peak value is extracted as a boundary pixel, and the boundary pixel extracted for each direction is integrated. A unit for labeling each area surrounded by boundary pixels in the acquired integrated image, scanning the integrated image, focusing on a certain boundary pixel, and existing in a predetermined area in at least two directions from the target pixel The label of the closest area pixel among the area pixels that are not the boundary pixels is checked, and the area within the predetermined range from the pixel of interest in at least two directions out of the at least two directions. When a pixel exists and the labels of the area pixels are all the same, a method of determining the pixel of interest as a division non-contribution boundary pixel, and a division non-contribution boundary pixel among the border pixels extracted by the differential processing in one direction If the pixel extracted as the boundary pixel in any other direction is the pixel of interest, and the differential value in the one direction is smaller than the differential value in any of the other directions, the pixel of interest is selected. Except for extracting one or more pixel blocks composed of adjacent division non-contribution boundary pixels and obtaining end points in the direction perpendicular to the differentiation direction, or boundary pixels extracted by a differentiation process in a certain direction. Of the division non-contribution boundary pixels, the pixel extracted as the boundary pixel in any other direction is the pixel of interest, and When the differential value in one direction is smaller than the differential value in any other direction, the pixel is composed of adjacent division non-contribution boundary pixels except for the pixel of interest that is the end point in the direction perpendicular to the differential direction. First selection means for extracting one or more pixel blocks and selecting one of the two means for obtaining an end point in the direction perpendicular to the differential direction, and the one direction for each pixel block Means for extracting a pixel block in which the number of pixels when the division non-contribution boundary pixel is shaded on the axis perpendicular to the line segment is larger than a predetermined value as a line segment, and each end point of the line segment exists in a predetermined region ahead of the line segment When the pixel is examined, and it is not connected to the boundary pixel contributing to the division or to another line segment, the pixel existing in the predetermined extension range in the direction of the line segment from the end point is defined as the boundary image. Or the end point of each line segment is checked for pixels existing in a predetermined area in the direction of the line segment, and when not connected to any boundary pixel, a predetermined extension range in the direction of the line segment from the end point. Is added to the boundary pixel, or the end point of each line segment is checked for a pixel existing in a predetermined area in the direction of the line segment, and the end point is not connected to any boundary pixel. If there is a boundary pixel in the predetermined extension range from the
The pixel up to the nearest existing boundary pixel is added to the boundary pixel, and when the boundary pixel does not exist in the predetermined extension range, one of three means of not performing the processing of adding the boundary pixel is selected. A boundary extraction device comprising selection means.