JP2000293666A - Method and device for detecting flaw and recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To suppress the occurrence of erroneous detection even when the flaws of a plurality of objects to be detected with different sizes are continuously detected by comparing the luminance of a pixel in a comparison area provided in a central part with the luminance of a pixel in a comparison area provided in an edge part and setting the object area of flaw detection in the image of the object to be detected on the basis of obtained comparison results. SOLUTION: A camera 14 of a 1st image pickup device 11 picks up the upper part of an object S to be detected directly and a lateral image indirectly and stores the obtained image signal in a 1st frame memory 4. An image extracting part 7 binarizes the obtained image with a preliminarily set threshold and specifies the contour of the object S included respectively in the upper image and both lateral images of the object S on the basis of an obtained binarized image. Then, the image of the object S is respectively extracted from the upper image and the both lateral images before binarization in the memory 4 on the basis of the specified image. A detecting part 8 detects the existence/ absence of a flaw and stores the results in a memory 10.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for detecting a flaw generated on the surface of an object to be detected based on an image obtained by imaging the object to be detected such as fruits and vegetables, and an apparatus used for carrying out the method.
The present invention also relates to a recording medium on which a computer-readable program for causing a computer to detect a flaw is recorded.

[0002]

2. Description of the Related Art In a sorting place where a plurality of fruits and vegetables are sorted according to a predetermined grade, a surface flaw of the fruits and vegetables, which is one element of the grade judgment, is detected as follows. A plurality of spherical fruits and vegetables, such as oranges or oysters, are placed in a row at predetermined intervals on a conveyor. At an appropriate position in the longitudinal direction of the conveyor, an imaging device for imaging fruits and vegetables, and a light emitting device including a straight tube lamp, a reflector, and the like are provided facing the transport area of the conveyor, and the imaging device performs the imaging of the imaging device. Each time a fruit or vegetable is transported to the area, it is imaged and the resulting image is provided to a flaw detection device via an analog / digital converter.

[0003] The flaw detection device stores an image given from the imaging device, binarizes the image to generate a binarized image, and determines a fruit and vegetable image. The flaw detection device extracts an image of a fruit or vegetable from the stored image, and calculates an average luminance that is an average value of the luminance of a plurality of pixels constituting the obtained image of the fruit or vegetable.

[0004] The flaw detection device generates a threshold value which is lower than the average luminance by a predetermined ratio based on the calculated average luminance. The flaw detection device compares the luminance of each pixel with a threshold, and when there is a pixel with a luminance lower than the threshold,
It is determined that the image is a flaw image.

[0005]

However, when a plurality of objects of different sizes are imaged by a conventional flaw detection device, it is difficult to irradiate the surface of each object with uniform intensity. Therefore, on the surface of each object to be detected, there is a portion where light having an intensity lower than the intensity of light applied to other portions is irradiated, and the brightness of an image related to the portion is smaller than a threshold value. In some cases, the image was incorrectly determined.

The present invention has been made in view of such circumstances, and a purpose thereof is to set comparison regions each having a predetermined number of pixels at a center portion and an edge portion of an image of an object to be detected.
The luminance of the pixel of the comparison area set at the center and the luminance of the pixel of the comparison area set at the edge are compared, and based on the obtained comparison result, the target area of the flaw detection is set on the image of the object to be detected. At the same time, the target area is divided to generate a plurality of divided areas, and the presence or absence of an image relating to the flaw is determined for each divided area, thereby continuously detecting a plurality of flaws of a plurality of objects having different sizes. It is an object of the present invention to provide a flaw detection method capable of suppressing occurrence of erroneous detection even in such a case, a device used for implementing the method, and a recording medium on which a computer-readable program for detecting a flaw is recorded.

[0007]

A flaw detection method according to a first aspect of the present invention extracts an image of a detected object from an image including the detected object,
In the method for detecting a flaw of an object to be detected based on the luminance of a plurality of pixels constituting an image of the object to be detected, a comparison area having a predetermined number of pixels is provided at a center and an edge of the image of the object to be detected. Are set, and the luminance of the pixel of the comparison area set at the center and the luminance of the pixel of the comparison area set at the edge are compared.
Based on the obtained comparison result, a target area for flaw detection is set on the image of the object to be detected, and the target area is divided to generate a plurality of divided areas.
It is characterized in that the presence or absence of an image relating to a flaw is determined.

In the flaw detection method according to a second aspect of the present invention, in the first aspect, a first comparison area and a second comparison area are set at appropriate positions on the edge, and the luminance of the pixels in both the comparison areas and the center are set. Comparing the luminance of the pixels of the set comparison area with the luminance of the pixels, and setting a quadrangular target area based on both the obtained comparison results and the positions where the first comparison area and the second comparison area are set. And

A flaw detection method according to a third aspect of the present invention is the flaw detection method according to the second aspect, wherein the ratio of the luminance of the pixel of the first comparison area to the luminance of the pixel of the comparison area set at the center and the pixel of the second comparison area are determined. The ratio between the luminance and the luminance of the pixel in the comparison area set at the center is determined, and the ratios corresponding to the determined ratios are determined by using a table in which a plurality of ratio ranges and the number of divisions are associated and stored in advance. The target area is divided by a division number determined by a ratio related to the first comparison area in a first direction determined by a position where the first comparison area is set, and the second comparison area is defined by a division number. The target area is divided by a division number determined by a ratio related to a second comparison area in a second direction determined by a set position.

According to a fourth aspect of the present invention, in the flaw detection method according to the second or third aspect, another image that partially overlaps the image is obtained, and the obtained other image corresponds to the size of the target area. Another target area having a rectangular shape is set, and the set other target area is divided according to the divided area.

According to a fifth aspect of the present invention, there is provided a flaw detection device which extracts an image of a detected object from an image including the detected object, based on luminance of a plurality of pixels constituting the obtained image of the detected object. A flaw detection device comprising: a determination unit configured to determine the presence / absence of an image related to a flaw of the detected object; and a comparison area for setting a predetermined number of comparison areas at a center portion and an edge of the detected object image. Setting means, comparing means for comparing the luminance of the pixels of the comparison area set at the center with the luminance of the pixels of the comparison area set at the edge, and the image of the object to be detected based on the obtained comparison result. A target area setting means for setting a target area for flaw detection, and a divided area generating means for generating a plurality of divided areas by dividing the target area, the determination means, for each obtained divided area, It should be noted that it is intended to determine the presence or absence of an image To.

In a flaw detection apparatus according to a sixth aspect, in the fifth aspect, the comparison area setting means includes means for setting a first comparison area and a second comparison area at appropriate positions on the edge. The target area setting means includes means for comparing the luminance of the pixels of both comparison areas with the luminance of the pixels of the comparison area set at the center, respectively, the two comparison results obtained, and the first
Means for setting a rectangular target area based on both the positions where the comparison area and the second comparison area are set.

[0013] In a flaw detection device according to a seventh aspect of the present invention, in the sixth aspect, the comparing means determines a ratio between the luminance of the pixel in the first comparison area and the luminance of the pixel in the comparison area set at the center. And a means for calculating a ratio between the luminance of the pixel in the second comparison area and the luminance of the pixel in the comparison area set in the center portion, and a table in which a plurality of ratio ranges and the number of divisions are stored in advance is stored. Storage means, wherein the divided area generating means determines the number of divisions corresponding to each of the determined ratios based on the table, and a first direction determined by a position at which the first comparison area is set. In contrast, means for dividing the target area by the number of divisions determined by the ratio according to the first comparison area, and for the second direction defined by the position where the second comparison area is set, Division determined by ratio In characterized by comprising a means for dividing the target area.

According to an eighth aspect of the present invention, in the flaw detection apparatus according to the sixth or seventh aspect, a means for obtaining another image partially overlapping the portion where the image is obtained, It is characterized by comprising means for setting another target area of a square having a size corresponding to the size of the area, and means for dividing the set other target area according to the divided area.

According to a ninth aspect of the present invention, there is provided a recording medium comprising: program code means for causing a computer to extract an image of an object from an image including the object; and luminance of a plurality of pixels constituting the obtained image of the object. On the basis of, in a recording medium in which a computer-readable program including a determination program code means for causing a computer to determine the presence or absence of an image related to the flaw of the detected object, the central portion of the image of the detected object and A comparison area setting program code means for causing a computer to set a comparison area having a predetermined number of pixels at an edge; And a comparison program code means for making a comparison with the computer and a target area for flaw detection in the image of the object to be detected based on the obtained comparison result. Target area setting program code means, and a divided area setting program code means for dividing the target area into a computer to generate a plurality of divided areas, the determination program code means, for each obtained divided area, A computer-readable program for causing a computer to determine the presence or absence of an image related to a flaw is recorded.

A recording medium according to a tenth invention is the recording medium according to the ninth invention, wherein the comparison area setting program code means causes a computer to set a first comparison area and a second comparison area at appropriate positions on the edge. Has,
The target area setting program code means is a program code means for causing a computer to compare the luminance of the pixels of both comparison areas and the luminance of the pixels of the comparison area set at the center, and the obtained both comparison results and the A computer readable program having program code means for causing a computer to set a rectangular target area based on both positions where the first comparison area and the second comparison area are set is recorded.

According to an eleventh aspect of the present invention, in the recording medium according to the tenth aspect, the comparison program code means stores in the computer the ratio of the luminance of the pixel in the first comparison area to the luminance of the pixel in the comparison area set at the center. Program code means for causing the computer to determine the ratio between the luminance of the pixel in the second comparison area and the luminance of the pixel in the comparison area set at the center, wherein the division area setting program code means A program code means for causing a computer to determine the number of divisions corresponding to each of the determined ratios using a table preset in association with a plurality of ratio ranges and the number of divisions; In the first direction determined by the position where the area is set, the first
Program code means for causing a computer to divide the target area by the number of divisions determined by the ratio of the comparison area, and a second direction corresponding to a second direction determined by a position where the second comparison area is set. A computer readable program having program code means for causing a computer to divide the target area by the number of divisions determined by the ratio is recorded.

A recording medium according to a twelfth aspect of the present invention is the recording medium according to the tenth or eleventh aspect, wherein a program code means for causing a computer to obtain another image partially overlapping the image, Computer including program code means for causing a computer to set another target area of a quadrangle having a size corresponding to the size of the area, and program code means for causing the computer to divide the set other target area according to the divided area A readable program is recorded.

According to the first, fifth and ninth aspects of the present invention, a comparison area having a predetermined number of pixels is provided at the center and the edge of the image of the object extracted from the image obtained by imaging the object. Set. The luminance of the pixel of the comparison region set at the center of the image of the detected object is compared with the luminance of the pixel of the comparison region set at the edge of the image of the detected object, and the difference between the former and the latter is large. Is
A relatively small target area is set as the image of the detected object, and when the difference between the former and the latter is small, a relatively large target area is set as the image of the detected object. As a result, a portion having a smaller luminance than the luminance of the pixel at the center of the image of the object is removed from the target area as much as possible, and occurrence of erroneous detection is suppressed.

When the difference between the former and the latter is large, the target region is divided into a relatively large number of divided regions, and when the difference between the former and the latter is small, the target region is set. The area is divided into relatively few divided areas. As a result, the number of divisions of the target area can be reduced as much as possible, and the time required for the calculation described later can be shortened.

An average value of the luminance of a plurality of pixels included in the divided region is calculated for each divided region, and a luminance lower than the obtained average by a predetermined ratio is set as a threshold. The luminance of a plurality of pixels included in each divided region is compared with a corresponding threshold, and if there is a luminance lower than the threshold, it is determined that the image is a flaw image.

When an image of a substantially spherical object is picked up, the brightness of the pixel at the edge is lower than the brightness of the pixel at the center of the image of the object. The larger the difference between the brightness of the pixel and the edge pixel, the more finely the target area is divided, so that the brightness of a plurality of pixels included in each divided area is limited to an appropriate brightness range, and the center of the image of the object is detected. Even when the difference between the luminance of the pixel of the part and the luminance of the pixel of the edge part is large, erroneous detection due to the difference between the two luminances is avoided. Therefore, even when a plurality of substantially spherical flaws of different sizes are continuously detected, occurrence of erroneous detection can be suppressed.

In the second, sixth and tenth aspects of the present invention, the first comparison area is set at an appropriate position on the edge of the image of the detected object,
A second comparison area is set at a position different from the position of the edge. For example, a first comparison area is set in the x-axis direction passing through the center of the image of the detection target, and a second comparison area is set in the y-axis direction passing through the center of the detection target image.

The luminance of the pixel in the first comparison area is compared with the luminance of the pixel in the comparison area set at the center. If the difference between the former and the latter is large, a rectangular target area having a small dimension in the x-axis direction is used. Is set, and when the difference between the former and the latter is small, a rectangular target area having a large dimension in the x-axis direction is set. Similarly, the luminance of the pixel in the second comparison area is compared with the luminance of the pixel in the comparison area set at the center, and if the difference between the former and the latter is large, the rectangular target area having a small dimension in the y-axis direction Is set, and when the difference between the former and the latter is small, a rectangular target area having a large dimension in the y-axis direction is set. Thus, the above-described target area can be easily set.

According to the third, seventh and eleventh aspects of the invention, the ratio of the luminance of the pixel in the first comparison area to the luminance of the pixel in the comparison area set at the center, and the ratio of the pixel in the second comparison area The ratio between the luminance and the luminance of the pixel in the comparison area set at the center is determined. In addition, a table in which a plurality of ratio ranges and the number of divisions are associated with each other is stored in a memory in advance, and the number of divisions corresponding to the obtained two ratios is determined with reference to the table.

The x-axis direction (first direction) of the rectangular target area set as described above is divided by the number of divisions determined according to the first comparison area, and the y-axis direction (second direction) in the target area is divided. Direction) is divided by the number of divisions determined according to the second comparison area. Thereby, the first direction and the second direction of the rectangular target area can be respectively divided according to the ratio of the luminance of the central part and the luminance of the edge part of the image of the detection target, and are included in each divided area. The brightness of many pixels is limited to a reasonable range.

According to the fourth, eighth and twelfth aspects of the invention, another image partially overlapping the image of the object to be detected is obtained. For example,
The imaging device captures an object from directly above to obtain an image including a planar image of the detection object, and a mirror disposed on a side of the detection object transfers an image of the side of the detection object to the imaging device. By forming an image, an image including a side image of the object is obtained. From the obtained image, a plane image (an image of the ratio object) and a side image (another image of the ratio object) of the object are respectively obtained. Extract. A part of the planar image and a part of the side image of such an object overlap each other.

As described above, the comparison area and the quadrangular target area are set in the plane image of the object to be detected, and the set target area is divided to generate a divided area. Further, another quadrangular target area whose one side is the dimension in the first direction or the dimension in the second direction of the target area set in the planar image is set as the side image of the detected object. Then, among the division numbers obtained by dividing the target area in the first direction or the second direction, another target area is divided by the division number in the direction applied to the setting of another target area. This makes it possible to easily set the target area in another image of the detected object, and to generate the divided area in a short time.

[0029]

Embodiments of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram showing the configuration of a flaw detection device according to the present invention, in which 21 is a conveyor. A plurality of trays T, T,
Are provided at predetermined intervals. Each of the trays T, T,... Stores a spherical object S, S,... Such as oranges or oysters, respectively, and the conveyor 21 includes the trays T, T,. ... in the direction of the arrow.

The first imaging device 11 and the second imaging device 12 for imaging the upper and side portions of the detected object S are spaced apart in the longitudinal direction of the conveyor 21 in the transport area of the detected objects S, S,. It is provided. The first imaging device 11 and the second imaging device 12 include a light-shielding housing 18 having an opening through which the conveyor 21 is inserted,
Equipped with 18.

In the housing 18, upper light emitters 15, 15 each having a straight tubular halogen lamp and a reflector are provided at a distance in the longitudinal direction of the conveyor 21 so as to mainly emit light to the upper part of the detection object S. The camera 14 is arranged so as to be orthogonal to the longitudinal direction of the conveyor 21, and a camera 14 for picking up an image of the detection object S is provided in the center between the upper light projectors 15, 15.

A little obliquely above the side edges of the conveyor 21, four side projectors 16, 16, 16, 16 each having a spherical halogen lamp and a reflector are positioned at each vertex of a square parallel to the conveyor 21. Provided, each side projector 16,
Light is projected from 16, 16, 16 to the side of the object S. Also,
Two pairs of side projectors 16, which make a pair in the longitudinal direction of the conveyor 21,
At the center between the side projectors 16 and the side projectors 16, there are provided reflectors 17, 17 for forming a side image of the object S on the camera 14, respectively.

A rotation device 13 is disposed between the first imaging device 11 and the second imaging device 12, and the rotation device 13 is
Each time the trays T, T,... Are transported to a predetermined position, the tray T and the detection target S stored in the tray T are rotated by 90 °. This makes it possible to detect a flaw on a portion other than the portion of the detection target S facing the tray T.

When the object S is conveyed to an appropriate position in the first and second imaging devices 11 and 12 by the conveyor 21, the cameras 14 and 14 move the trays T and T and the upper portions of the objects S and S. Directly, and indirectly capture the side images of the trays T, T and the objects S, S reflected on the reflecting mirrors 17, 17, and convert the obtained image signals into analog / digital (A / D) signals. Converter 2,
Give 2 to each. A / D converters 2 and 2 are both cameras
The image signal given from the camera 14 of the first image pickup device 11 is converted into a digital signal from the image signal given from 14 or 14 to the first frame memory 4 of the flaw detector 3.
The image signal provided from the camera 14 of the imaging device 12 is provided to the second frame memory 5 and stored therein.

Each time an image is stored in the first frame memory 4 or the second frame memory 5, the image is stored in the image extraction unit 7.
The image extraction unit 7 binarizes the given image with a preset threshold value, and
Based on the binarized image, the outline of the object included in the upper image of the object directly obtained by imaging and the outline of the object included in both side images of the object obtained by indirect imaging are specified. . The image extracting unit 7 extracts images of the detected object from the upper images and the two side images before binarization in the corresponding frame memories 4 and 5 based on the specified outline of the detected object, respectively. This is given to the detection unit 8. Then, the detection unit 8 detects the presence or absence of a flaw as described below, gives the result to the memory 10, and stores it there.

FIG. 2 is a flowchart showing a procedure for detecting a flaw by the detection unit 8 shown in FIG. The detection unit 8 includes, at the center of the image of the detected object extracted from the upper image, an edge in the x-axis direction passing through the center, and an edge in the y-axis direction passing through the center,
Areas E _O , _Ex , E having the same area (number of pixels) determined in advance
_y (comparison area) is set (steps S1, S2, S
3) That is, the center of the area E _O coincides with the center of the image of the detected object, and the areas _Ex and E _y are inscribed in the outline of the image of the detected object S, and the respective areas E _O , _Ex and. calculating a plurality of average luminance which is the average of the luminance of the pixels constituting the E _y (step S4). Further, the detection unit 8 obtains the maximum length in the x-axis direction and the maximum length in the y-axis direction of the image of the object extracted from the upper image (step S5).

The detecting section 8 has an area E_xAverage brightness and area
A E_OOf the average luminance (E_x/ E _O), Area E_yof
Average brightness and area E_OOf the average luminance (E_y/ E_O)
Are obtained (steps S6 and S7). Detection unit 8
As shown in Table 1 below, the detection unit memory 9 stores the average luminance.
And the dimensional ratio that determines the target area for flaw detection
Is stored, and the detection unit 8
Dimensions corresponding to the ratio of both average luminances calculated in step S4
The ratio is determined (step S8), and the maximum length in the x-axis direction is determined.
And the maximum length in the y-axis direction multiplied by the corresponding dimensional ratio
To calculate the x-axis dimension and the y-axis dimension
(Step S9). Then, the detection unit 8 performs the operation shown in FIG.
As shown in the figure, the image P of the detected object extracted from the upper image_U
The image P _UThe same center as the center of
And the target area 31 of the dimension in the y-axis direction (step S1)
0).

[0038]

[Table 1]

[0039] For example, the average luminance of the area E _x the average luminance of the area E _O set at the center of the image of the detected object S 100 was set on the edge of the x-axis direction passing through the center of the image of the detected object S If but 50, the average luminance of the area E _y set in the edge of the y-axis direction passing through the center of the image of the detected object S was 90, x
The axial dimension is (2/3) × (maximum length in x-axis direction), and the y-axis dimension is (4/5) × (maximum length in y-axis direction)
It is.

As shown in Table 2 below, a table in which the number of divisions into which the target area 31 is divided and the ratio of the average luminance are stored in the detection unit memory 9.
Determines the number of divisions corresponding to the ratio of the both average luminances calculated in step S4 (step S11), and
Are divided into a number of divisions corresponding to the x-axis direction and the y-axis direction, and a plurality of divided regions 32, 32,... Are set (step S12).

[0041]

[Table 2]

When the setting of the target area and the divided area in the image of the object extracted from the upper image in this way is completed, the object area and the divided area are added to the image of the object extracted from the side image as follows. Set the area.

[0043] Figure 5 is an explanatory diagram for explaining a method of setting a target area and divided area on the image of the detected object extracted from the side image, in the figure, the P _U is the detection object is extracted from the top image image, P _S represents the image of the object to be detected extracted from the side image. As shown in Figure 5, the image P _U of the detected object extracted from the upper image has a substantially circular, the said image P _U, the first object region 31 rectangular is set as the horizontal axis in the x-axis It is. In the first target area 31, a plurality of first divided areas 32, 32,..., Which divide the x-axis direction and the y-axis direction into 3 × 3, for example, are set.

[0044] Approximately the image P _S of the detected object extracted from the side image, and forms a shape connecting arcs across the greater than 180 ° center angle with a straight line, the straight bottom of said image P _S The longitudinal direction is the image P _{U of the} object extracted from the upper image.
In the same direction as the x-axis direction.

[0045] To set the second target area 41 rectangular to the image P _S of the detected object extracted from such side image,
The detecting unit 8 sets the dimension r ₁ of the first target area 31 in the x-axis direction to the second
The size r _{1 in} the x-axis direction of the target area 41 is set (step S2
0). The detection unit 8, determine the maximum radius r of the image P _S of the detected object extracted from the side image (step S21), and they are substituted into the following equation (1) z-axis direction of a portion size r Ask for ₂ . ^{_{r 2 = r 1 2 + r}} 2 2 ... (1)

The detection unit 8 extracts from the side image parallel to the x-axis.
The image P of the object to be detected_SThe coordinates of the line segment passing through the center of the arc
The image P_SFrom the coordinates of the linear base of
(Step S22), and the value is set to the partial dimension r._TwoJoin
To obtain the dimension in the z-axis direction (step S
twenty three). The detection unit 8 detects the image P of the detected object extracted from the side image.
_SAnd the dimension r in the x-axis direction obtained as described above.₁And z-axis
The second target area 41 having a rectangular shape having a dimension in the direction
Image of the detected object extracted from the center and side images of the target area 41
P_S(Step S2)
Four). Then, the detecting unit 8 sets the second target area 41 in the x-axis direction
And the image P of the object extracted from the upper image _UX-axis direction of
Divided into the same number of divisions as
.. Are set (step S25).

In this way, when the setting of the target area and the divided areas in the image of the object to be detected extracted from the upper image and the side image is completed, the detecting unit 8 separates the divided areas for each divided area. Is calculated (step S30), and a luminance lower than the obtained average by a predetermined ratio is set as a threshold (step S31).
The luminance of a plurality of pixels included in the divided area is compared with the threshold, and if there is a pixel having a luminance lower than the threshold, it is determined that the image is a flaw image (step S32).

The program for detecting the above-mentioned flaw is provided by a computer RO having a configuration as shown in FIG.
M, as shown in FIG. 6, a recording medium RM such as a compact disk or a flexible disk.
, The recording medium RM may be loaded into a disk drive of a computer and loaded, and flaw detection may be performed.

[0049]

As described in detail above, the first, fifth, and ninth embodiments
In the present invention, the number of divisions of the flaw detection target area is reduced as much as possible based on the result of comparison between the luminance of the pixel at the center of the image of the detected object and the luminance of the pixel at the edge. Thus, the time required for flaw detection can be shortened. Further, the luminance of a plurality of pixels included in each divided region is limited to an appropriate luminance range, and the difference between the luminance of the central pixel and the luminance of the peripheral pixel of the image of the detected object is large. Also, erroneous detection due to the difference between the two luminances is avoided.

According to the second, sixth and tenth aspects of the present invention, it is possible to easily set the target area according to the shape of the image of the detected object.

In the third, seventh and eleventh aspects of the present invention, since the luminance of many pixels included in each divided area is limited to an appropriate value range, a substantially spherical object to be detected may be used. This can also prevent occurrence of erroneous detection.

According to the fourth, eighth, and twelfth aspects of the present invention, the target area can be easily set in another image of the detected object, and the divided area can be generated in a short time.
In addition, the present invention has excellent effects, such as hardly overlooking flaws, because another image that partially overlaps with the image of the detected object is obtained and the presence or absence of a flaw image is determined for this other image. To play.

Although reference numerals have been entered in the claims to facilitate comparison with the drawings, it is needless to say that the present invention is not limited to the configuration corresponding to the entered reference numerals.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a configuration of a flaw detection device according to the present invention.

FIG. 2 is a flowchart showing a procedure for detecting a flaw by the detection unit shown in FIG. 1;

FIG. 3 is a flowchart showing a procedure for detecting a flaw by the detection unit shown in FIG. 1;

FIG. 4 is a flowchart showing a procedure for detecting a flaw by the detection unit shown in FIG. 1;

FIG. 5 is an explanatory diagram illustrating a method of setting a target area and a divided area in an image of a detection object extracted from a side image.

FIG. 6 is a schematic diagram showing another embodiment.

[Explanation of symbols]

 Reference Signs List 3 flaw detector 7 image extraction unit 8 detection unit 9 detection unit memory 10 memory 11 first imaging device 12 second imaging device 13 rotation device 14 camera 15 upper projector 16 side projector 17 reflector 18 housing 21 conveyor S cover Detected object T tray

Continued on the front page F term (reference) 2G051 AA05 AB02 AB07 BA01 BA20 CA04 CA07 DA01 DA06 DA08 EA11 EA12 EA14 EA20 EC03 ED07 5B057 AA02 AA15 BA30 CC02 DA03 DA08 DB02 DB09 DC22 5L096 AA06 BA03 FA14 GA19 GA51 9A001H57

Claims (12)

[Claims] 1. A method for extracting an image of an object to be detected from an image including the object to be detected, and detecting a flaw of the object to be detected based on luminance of a plurality of pixels constituting the obtained image of the object. In each, a comparison region of a predetermined number of pixels is set at the center and the edge of the image of the detection object, respectively, and the luminance of the pixel of the comparison region set at the center and the luminance of the pixel of the comparison region set at the edge are set. And, based on the obtained comparison result, set a target area for flaw detection in the image of the object to be detected, and divide the target area to generate a plurality of divided areas. A flaw detection method characterized by determining the presence or absence of a flaw image for each area. 2. A first comparison area and a second comparison area are set at appropriate positions on the edge, and the luminance of the pixels in both comparison areas is compared with the luminance of the pixels in the comparison area set in the center. 2. The flaw detection method according to claim 1, wherein a quadrangular target area is set based on the obtained two comparison results and both positions where the first comparison area and the second comparison area are set. 3. The ratio of the luminance of the pixel of the first comparison area to the luminance of the pixel of the comparison area set at the center, and the luminance of the pixel of the second comparison area and the luminance of the pixel of the comparison area set at the center. Are determined, and a plurality of ranges of ratios and the number of divisions are associated with each other, and the number of divisions corresponding to each ratio is determined using a table stored in advance, and the first comparison area is set. The target area is divided by a division number determined by a ratio related to the first comparison area with respect to a first direction determined by the determined position, and the second direction is determined by a position set by the second comparison area. 3. The flaw detection method according to claim 2, wherein the target area is divided by the number of divisions determined by a ratio related to two comparison areas. 4. Obtaining another image partially overlapping the image, setting another square target area having a size corresponding to the size of the target area in the obtained other image, Target area,
4. The flaw detection method according to claim 2, wherein the division is performed according to the divided area. 5. An apparatus for extracting an image of a detected object from an image including the detected object, and detecting a flaw of the detected object based on luminance of a plurality of pixels constituting the obtained image of the detected object. A flaw detection device comprising: a determination unit configured to determine the presence or absence of the image; a comparison region setting unit configured to set a comparison region having a predetermined number of pixels at a center portion and an edge portion of the image of the detected object; A comparison unit that compares the luminance of the pixel of the comparison area with the luminance of the pixel of the comparison area set at the edge, and sets a target area for flaw detection on the image of the object based on the obtained comparison result. Target area setting means, and divided area generating means for dividing the target area to generate a plurality of divided areas, wherein the determining means determines presence or absence of a flaw image for each of the obtained divided areas. A flaw detection device characterized by what it does. 6. The comparison area setting means includes means for setting a first comparison area and a second comparison area at appropriate positions on the edge, and the target area setting means includes a luminance of a pixel in each of the comparison areas. Means for comparing the luminance of the pixels of the comparison area set at the center with the obtained two comparison results and the positions of the first and second comparison areas, respectively. 6. The flaw detection device according to claim 5, further comprising means for setting a target area. 7. The comparison means for calculating a ratio between the luminance of the pixel in the first comparison area and the luminance of the pixel in the comparison area set at the center, and the luminance of the pixel in the second comparison area and the luminance at the center of the second comparison area. Means for calculating the ratio of the luminance of the pixels in the comparison area set in the storage area, and storage means (9) in which a table in which a plurality of ranges of ratios are associated with the number of divisions is stored in advance. The area generating means is configured to determine the number of divisions corresponding to each of the determined ratios based on the table, and to determine the number of divisions corresponding to each of the determined ratios in a first direction determined by a position where the first comparison area is set. Means for dividing the target area by the number of divisions determined by the ratio, and the target area by the number of divisions determined by the ratio of the second comparison area in a second direction determined by the position where the second comparison area is set. Means for dividing The flaw detection device according to claim 6. 8. A means for obtaining another image that partially overlaps with the part where the image is obtained, and setting another target area of a square having a size corresponding to the size of the target area in the obtained other image. The flaw detection device according to claim 6, further comprising: a unit that divides another set target region according to the divided region. 9. A program code means for causing a computer to extract an image of a detected object from an image including the detected object, and an object to be detected based on luminance of a plurality of pixels constituting the obtained image of the detected object. A determination program code means for causing a computer to determine the presence or absence of an image related to the flaw of a computer-readable recording medium, wherein a predetermined number of pixels are provided at the center and the edge of the image of the object to be detected. Comparison area setting program code means for setting a comparison area by a computer, and comparison program code means for causing a computer to compare the luminance of the pixels of the comparison area set at the center with the luminance of the pixels of the comparison area set at the edge. And a target area setting program for causing a computer to set a target area for flaw detection in the image of the detection target based on the obtained comparison result. Ram code means, and a divided area setting program code means for dividing the target area by a computer to generate a plurality of divided areas, wherein the determination program code means comprises: A recording medium characterized by recording a computer-readable program for causing a computer to determine the presence or absence of a program. 10. The comparison area setting program code means includes program code means for causing a computer to set a first comparison area and a second comparison area at appropriate positions on the edge, and the target area setting program code means Program code means for causing a computer to compare the luminance of the pixels of both comparison areas and the luminance of the pixels of the comparison area set at the center, respectively, the obtained two comparison results, the first comparison area and the second comparison area 10. The recording medium according to claim 9, wherein a computer-readable program having program code means for causing a computer to set a quadrangular target area based on both positions where the area is set is recorded. 11. The comparison program code means includes: a program code means for causing a computer to determine a ratio between a luminance of a pixel in a first comparison area and a luminance of a pixel in a comparison area set at a center portion; Program code means for causing a computer to determine the ratio between the luminance of the pixel and the luminance of the pixel in the comparison area set at the center.The divided area setting program code means sets a plurality of ratio ranges and the number of divisions. Using a table preset in association with
Program code means for causing a computer to determine the number of divisions corresponding to each of the determined ratios, and a ratio determined by a ratio of the first comparison region with respect to a first direction determined by a position where the first comparison region is set. Program code means for causing the computer to divide the target area by the number of divisions, and a division number determined by a ratio related to the second comparison area in a second direction determined by a position where the second comparison area is set. 11. A computer-readable program having program code means for dividing the target area by a computer is recorded.
The recording medium according to the above. 12. A program code means for causing a computer to obtain another image that partially overlaps with the image, and the other obtained image includes another target area of a square having a size corresponding to the size of the target area. 12. A computer-readable program including program code means for causing a computer to set and program code means for causing a computer to divide another set target area according to the divided area is recorded. Recording medium.