JPH0810194B2 - Threshing processing grain inspection device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Industrial field of use] The present invention relates to an imaging means for imaging grains after threshing processing,
Threshing-processed grains including image processing means for extracting a region corresponding to the grain based on the imaged information of the image pickup means, and determination means for determining the presence or absence of a twig based on the information of the extracted region Inspection equipment.

[Conventional technology]

If the grains after threshing treatment include those with twigs, there is a risk of causing inconveniences such as damaging the grains with twigs when dried with a dryer, for example, grains with twigs. Performance tests of threshing equipment are being conducted to check the mixing rate of grains. By the way, if the threshing process is insufficient, it will have twigs. It is desired that the mixing rate with twigs is 3% or less. Therefore, in order to check whether the threshing process is being performed properly, samples of about 4000 to 5000 grains are taken out from the grain after the threshing process, and the presence or absence of twigs in these samples is checked for each grain. Need to determine.

However, if small grains are checked manually one by one, the inspection will take time and the inspection accuracy may vary depending on the skill level of the operator. Is desired.

By the way, if the grain has a twig, the area of the grain imaged by the twig becomes large, so in the past, the area of the region corresponding to the grain extracted using image processing A means for discriminating the presence or absence of a twig is proposed based on the size of the above (see, for example, Japanese Patent Application No. 1-125662 previously proposed by the applicant).

[Problems to be Solved by the Invention]

However, there are individual differences in grain size, and it is desirable to be able to determine the length of the twig in order to make an accurate determination.

However, if the twig infarct is long, the twig may be imaged in a curved state. In order to accurately determine the length of the curved twig, there is a disadvantage that the image processing becomes complicated and the processing time becomes long. Of course, if the image processing is simplified in order to shorten the processing time, the risk of misjudgment increases.

The present invention has been made in view of the above circumstances,
The purpose thereof is to reduce the processing time required for the discrimination without deteriorating the accuracy of discrimination with a twig.

[Means for solving the problem]

An inspection device for threshing-treated grains according to the present invention is an image pickup unit for picking up an image of the grain after threshing, an image processing unit for extracting a region corresponding to the grain based on image pickup information of the image pickup unit, and an extraction unit. And a discriminating means for discriminating the presence or absence of a twig on the basis of the information of the region thus obtained, and its characteristic configuration is as follows.

That is, a twig branch extracting means for extracting a twig branch region corresponding to only a twig branch from the region, and a length discriminating means for discriminating the length of the twig branch based on information of the twig branch extracting means are provided. If the area of the region is larger than a set value, and if the length of the twigs discriminated by the length discriminating means is larger than a set value, the discriminating means discriminates that there is a twig. It is configured as follows.

[Work]

In the case where the twig is long such as curved, even if the grain is small, the size of the region corresponding to the extracted grain is shorter than when the twig is not attached and when the twig is not attached. Greatly large.

Therefore, if the area of the extracted region is larger than the set value, even if it is determined that the twig infarction is included without discriminating the length of the twig, the one without twig is attached to the twig. There is no fear of making a mistake.

However, since there is individual difference in grain size, if the area of the extracted region is smaller than the set value, the length of the twig branch region corresponding to only the twig branch is determined. However, as the image processing for the length determination, an error may occur in the length determination for a curved long twig, so that the length of a relatively short twig can be accurately determined. A simple process with a short time can be used. As a result, the processing time as a whole can be shortened as compared with the case of using the image processing for accurately determining the length of the twig of all the grains.

〔The invention's effect〕

As a result, the processing time required for the discrimination can be shortened without lowering the discrimination accuracy with the twig.

〔Example〕

Hereinafter, an example in which the present invention is applied to a paddy inspection device after threshing treatment will be described with reference to the drawings.

If the grain after the threshing treatment contains unpolished rice and grains with twigs, the rate of barrel cracks increases and the number of grain damages increases when dried in a dryer. Due to its disadvantages, the present inspection device will be used when performing a test for checking the mixing rate of brown rice and twigs on the paddy after threshing treatment as a performance test of the threshing device.

By the way, if the threshing process is excessive, it will be removed and turned into brown rice, and if insufficient, it will become paddy with twigs. And
Brown rice is less than 0.5%, with twigs is 3%
The following is desired. Therefore, in order to check whether the threshing process is being performed properly, samples of about 4000 to 5000 grains are taken out from the grain after the threshing process, and those samples are attached with paddy, brown rice, and twigs. It will be determined for each grain which one of them.

As shown in FIG. 10 and FIG. 11, the inspection device for threshing-treated grains includes a vibrating grain feeding device (1) as grain feeding means, and a transport end of the grain feeding device (1). Belt-type conveyor (2) as grain transporting means for transporting grain fed from the lower part, and two-dimensional image as image capturing means for repeatedly capturing the grain transported by the conveyor (2) from above. Sensor (3) and said supply device (1)
And a control device (4) using a microcomputer for controlling the operation of the conveyor (2) and discriminating a defective product based on the imaging information of the image sensor (3).

In addition, as will be described in detail later, a region corresponding to the grain (A) by utilizing the control device (4) to binarize the imaging information of the image sensor (3) based on the brightness value thereof. Image processing means (10) for extracting (X) (see FIG. 3)
0), based on the information of the twig infarct extracting means (102), which extracts the twig infarct region (Xb) corresponding to only the twig infarct based on the information of the extracted region (X) Each of the length discriminating means (103) for discriminating the length of the twig and the discriminating means (101) for discriminating the presence or absence of the twig based on the information of the respective means is configured.

By the way, the conveying surface of the conveyor (2) is painted black in order to prevent the influence of reflected light, and the image pickup range of the image sensor (3) is illuminated by the illumination device (5) so as to have a set brightness. It is supposed to be done. Further, the image sensor (3) repeats at a cycle faster than the convey speed of the conveyor (2) in order to capture the state in which the grain (A) is located in the imaging visual field (B). It is configured to be imaged.

However, the control device (4) captures the imaging information for each imaging cycle of the image sensor (3) and determines whether or not the grain (A) is within the imaging field of view (B). If there is a grain (A), the processing of the captured image information is automatically started.

The image sensor (3) has a red signal (R), a green signal (G), and a blue signal (B) corresponding to the three primary color components, in addition to the luminance signal (Y) corresponding to the brightness of the imaged information. ), So that each can be output, is configured in color. However, the image processing will be described later.

Explaining the presence / absence determination of the grain (A), since the conveyance surface of the conveyor (2) that is the background of the imaging field of view (B) is black, the grain ( When there is no A), the brightness of the imaging information is low, and when the grain (A) is within the imaging field of view (B), the brightness is high.

Therefore, when it is determined that there is one whose brightness is higher than the set threshold value based on the imaging information, the setting required for the grain (A) to move from the edge of the imaging field of view (B) to the vicinity of the center. As time passes, the image pickup information is fetched and image processing is started.

Therefore, the grain (A) conveyed by the conveyor (2)
Is aligned in a line along the transport direction, and
It is necessary to make a space for each grain along the transport direction.

Therefore, the vibrating feeder (1) uses a structure similar to that of a so-called vibrating parts feeder to arrange the fed grain groups in a line until reaching the transportation end portion. The conveyor (2) is configured to drop and feed from the transport end portion to the transport start end portion of the conveyor (2), and the transport speed of the conveyor (2) is set to the above in order to separate the dropped and fed grain into each grain. It is set to be higher than the supply speed of the supply device (1). That is, the interval between the grains conveyed by the conveyor (2) is determined by the conveyor (2).
The intervals are set so as to correspond to the difference between the transporting speed of 1 and the supplying speed of the supplying device (1).

As shown in FIGS. 10 and 11, the vibrating type feeding device (1) has a main body part (6) formed in a substantially cylindrical shape as a whole.
And a drive part (7) for driving the main body part (6) in a state of slight vibration along the circumferential direction.

As shown in FIGS. 12 and 13, the main body (6)
Is formed into a cylindrical container that is inclined downward from the bottom surface toward the inner peripheral wall portion and has an open upper end.

On the inner peripheral wall of the main body (6), there is formed a spiral shelf (8) rising from the lower bottom surface toward the upper opening, and from the upper end of the spiral shelf (8). A V-shaped guide groove (9) is formed in a portion extending to a portion above the conveyance start end of the conveyor (2) and extending to the conveyance end. The V-shaped guide groove (9) is formed so that the width gradually changes from the large side to the small side and the depth changes from the large side to the small side toward the conveying end portion.

That is, the main body part (6) is subjected to a slight vibration by the driving part (7), so that the grain (A) put therein is provided with the spiral shelf along the inner peripheral wall part. By transporting from the lower part to the upper part along the part (8) and carrying it along the V-shaped guide groove (9) following the spiral shelf (8), By utilizing the shape change of the V-shaped guide groove (9) whose width becomes narrower and the depth becomes shallower toward the side and the vibration of the main body (6), the fed grain group is conveyed to the end. The lines are arranged in a line between the parts.

The grain (A) overflowing from the V-shaped guide groove (9) has an opening (10) for grain recovery formed below the conveying end of the V-shaped guide groove (9). ) To the inner bottom surface of the main body (6) for re-transportation.

As shown in FIGS. 14 and 15, the feeding device (1)
At the end of the V-shaped guide groove (9), which serves as the transfer end of the, the grain (A) falling from the V-shaped guide groove (9) is received and transferred by the conveyor (2). A relay member (12) as a guide body having a V-shaped guide groove (11) for guiding to the start end of the surface forms the V-shaped guide groove (11) serving as the guide surface in the conveyor ( It is provided in a posture inclining toward the conveyance surface of 2).

Then, a reflection type of detecting the presence or absence of the grain (A) falling from the relay member (12) at a position between the transport end of the relay member (12) and the transport surface of the conveyor (2). An optical sensor (13) is provided, and this optical sensor (1
When 3) detects the falling grain (A), the vibration of the supply device (1) is controlled so as to be weakened for a set time, so that a plurality of grains (A) are At the same time, it is prevented from being supplied to the conveyor (2). Then, the transfer speed of the conveyor (2) is
A state in which the grains (A) arranged in a line are spaced from each other along the conveying direction of the conveyor (2) by setting the feeding speed of the feeding device (1) to be higher. It is designed to be transported in.

Further, if dust or dirt is mixed in the grain (A) supplied from the supply device (1), there is a risk of misjudgment,
Dust and dust mixed in the grain (A) can be automatically removed.

To add a further explanation, the end portion of the V-shaped guide groove (9) of the supply device (1) is formed such that its depth gradually increases in a state of being inclined downward in the transport direction, and An interval (K) smaller than the width of the grain (A) is provided between the end of the V-shaped guide groove (9) and the start of the relay member (12). That is, the dust and the dust carried in the V-shaped guide groove (9) are dropped and eliminated through the interval (K), and only the grain (A) is received by the relay member (12). It is handed over.

In addition, even if dust or dust is erroneously delivered to the relay member (12), the dust is not transferred to the relay member (12) in order to prevent the dust or dust from falling onto the conveyor (2). A slit-shaped opening (14) for dropping and removing dust is formed at the bottom of the V-shaped guide groove (11) that is formed, and the slit-shaped opening (14) falls on the lower surface of the opening (14). A plate-like member (15) for guiding dust and dirt toward the lateral side of the conveyor (2) is provided.

Explaining the drive part (7), although not described in detail, the main body part (6) is vibrated with a frequency proportional to the drive frequency and an amplitude proportional to the drive power. ing. That is, the grain supply speed of the supply device (1) is configured to increase as the drive frequency or drive power increases. Therefore, when the optical sensor (13) detects the fall of the grain (A),
The driving frequency or the driving power is changed to the smaller side and the supply speed is changed.

In FIG. 10, (16) is an electric motor for driving the conveyor (2), and (17) is the grain (A) supplied to the conveying start end of the conveyor (2). Is a guide member that guides the vehicle so that it does not fall outward, and is provided at both left and right sides of the conveyance start end of the conveyor (2).

 The image processing will be described below.

As described above, first, based on the value of the luminance signal (Y), it is determined that the grain (A) is in the state of being located within the imaging visual field (B) of the image sensor (3). The captured image information is 32 x 32 per screen.
Converted to binary image information with pixel density, and the grain (A)
The area (X) corresponding to is extracted (see FIG. 3).

However, since the twigs are much narrower than the grain (A), when the image is picked up in the focused state, the image information is quantized relatively coarsely in 32 × 32 pixels as shown in FIG. In addition, due to the quantization error in the image processing, the extracted region (X) is divided into a grain region (Xa) corresponding to the grain (A) and a twig branch region (Xb) corresponding to the twig. However, there is a risk of misjudgment.

Therefore, by intentionally imaging the image sensor (3) in a non-focused state, as shown in FIG.
An image signal component corresponding to a grain or a twig is expanded to prevent a quantization error from occurring.

That is, based on the value of the luminance signal (Y), the captured imaging information is converted into binarized image information at a density of 32 × 32 pixels per screen, and the area (X) corresponding to the grain (A) is converted. )
The processing for extracting the image data corresponds to the image processing means (100).

Next, image processing for determining whether the grain (A) has brown rice, paddy, or twigs will be described.

As described above, if the threshing treatment is insufficient, the twigs will be attached. As a result, the extracted region (X)
Is larger than the case without a twig infarction by the area of the twig infarct region (Xb) corresponding to the twig. On the contrary, when the threshing treatment is excessive, the grain (A) is dehusked to form brown rice, so that the grain size is smaller than that of the unhulled rice.

Therefore, basically, when the area of the region (X) is larger than the size of the unhulled rice that has been threshed appropriately, it is determined that the rice has twigs. I am allowed to do it. Since the size of the area (X) is proportional to the number of pixels belonging to the area (X),
The larger the number of pixels, the larger the area.

However, since the size of the grain (A) varies due to individual differences even if it is the same type, it is possible to determine whether it is paddy or brown rice only from the size of the region (X). It may be inaccurate.

Therefore, the information of the color signal output from the image sensor (3) is used to distinguish between paddy and brown rice.

In addition, if attention is paid to the difference in color between the case where the grain (A) is paddy and the case where the grain is brown rice, the brown rice looks whiter than the paddy because it has been removed.

Therefore, comparing the distribution states of the three primary color components, as shown in FIG. 8, the value of the red signal (R) is not significantly different between the case of paddy and the case of brown rice, but the blue signal ( The value of B) is much larger in brown rice than in paddy. Although not shown, since the value of the green signal (G) is not so different between the paddy and the brown rice as in the case of the red signal (R), the green signal (r) is used instead of the red signal (R). G) may be used.

Therefore, when a region in which the value obtained by subtracting the blue signal (B) from the red signal (R) of the pixel belonging to the region (X) is larger than the set value for noise removal is extracted, the area is Regardless of the difference, the amount of unhulled rice is significantly larger than that of unhulled rice, and it is possible to discriminate between unhulled rice and unhulled rice with approximately 100% probability (see FIG. 9).

That is, from the red signal (R) to the blue signal (B)
The area of the area where the value obtained by subtracting is larger than the set value for noise removal is determined by determining whether or not the area is larger than the set threshold for distinguishing brown rice and paddy. The normalization of the brightness and the determination of whether it is paddy or unpolished rice can be performed at the same time.

However, in order to speed up image processing, when the image pickup information of the image sensor (3) is taken in, the red signal (R) to the blue signal are transferred in parallel with the quantization of the luminance signal (Y). The value obtained by subtracting (B) is quantized with the same pixel density as the luminance signal (Y) and stored in advance in an image memory (not shown).

Then, if it is determined to be paddy, the branch branch region (Xb) (see FIG. 3) corresponding to only the branch branch is re-extracted from the region (X), its length is calculated, and the calculated length is calculated. Those with a length longer than the set value are discriminated as having a twig.

Next, processing for extracting the twig infarct region (Xb) and discriminating its length will be described.

Since the twigs are much narrower than the main body of the grain (A), the number of pixels belonging to the twig region (Xb) is smaller than the number of pixels belonging to the grain region (Xa).

Therefore, as shown in FIG. 3, when the frequency distribution of pixels having the same coordinate values in the x-axis and the y-axis of the imaging visual field (B) is obtained, the grain area (in the x-axis and the y-axis) is calculated. Xa) has the maximum number of pixels at the center, and the number of pixels gradually decreases toward the outer edge of the region (X), and is in the direction along the longitudinal direction of the twig. The distribution width becomes large.

Therefore, on the side where the distribution width of the frequency distribution of pixels having the same coordinate value on each of the x-axis and the y-axis is large, the frequency is larger than the set frequency corresponding to the size of the paddy, and the width thereof is larger. The center value of the coordinates that maximizes is obtained as the value corresponding to the center of the grain area (Xa).

Next, from the center of the grain area (Xa), toward the outer end of the distribution, which is the far side, where the distance is first smaller than the set frequency set corresponding to the width of the twig. Is calculated as the boundary (a) between the grain region (Xa) and the twig branch region (Xb).

By the way, when the twig infarct is extremely long, as shown in FIG. 5, it may be imaged with its end curved. Thus, if the end of the twig is curved, the correct boundary (a) can be obtained only from the frequency distribution.
Not only is it difficult to distinguish from the false boundary (c) having the same coordinate value as, but in the frequency distribution, the end (b) of the twig branch region (Xb) overlapping the grain region (Xa) may be lost. is there.

However, if the adjacency state of the pixel with respect to the obtained boundary (a or c) is checked on both the obtained x-axis and y-axis, it is determined whether it is the correct border (a) or the false border (c). Although it is possible to make a distinction, there is a possibility that the processing becomes complicated, the processing time becomes long, and the transportation speed of the grain (A) cannot be followed.

By the way, when the end of the twig is long such that it curves, the region (X) extracted based on the luminance signal (Y) is obtained.
The area of is larger than that of large paddy without twig.

Therefore, if the area of the region (X) extracted based on the luminance signal (Y) is larger than the set value in the case of a long branch branch, it is determined that the branch branch is included. Under the condition, the process of discriminating the length of the twig is simplified, and the entire process can be simplified and speeded up.

Explaining the length determination of the twig branch, after the start point of the twig branch region (Xb) is determined, the pixel to which the twig branch region (Xb) belongs is separated from the start point (length direction of the twig branch). By sequentially differentiating toward the line, the line is thinned to the width of one pixel, and the number of pixels, that is, the length of the thinned pixel is obtained.

However, as shown in FIG. 4, when the extracted pixels are adjacent to each other in a direction parallel to the x-axis or the y-axis, 1 corresponding to one pixel is added as a unit distance, and the pixel is diagonally moved. If they are adjacent, 1.4 is added as the unit distance, which corresponds to the diagonal distance of the square. That is, when the pixels belonging to the twig branch region (Xb) are diagonally adjacent to each other in the image, the length per pixel is set to be larger than that in the vertical and horizontal directions when the pixels are adjacent to each other. By making it possible to discriminate the length, it is possible to accurately determine the length of the twig even if the twig is bent or the twig is imaged in an oblique direction on the image. -ing

Further, when the branch branch is bent or extended in a direction intersecting with the coordinate direction to be searched, it is determined that the adjacency of the pixels belonging to the branch branch area (Xb) is interrupted by the thinning process by the differentiation. In that case, the value (n) corresponding to the difference in the number of broken pixels is added to the unit distance (1 or 1.4). In addition, when the adjacency of the pixels belonging to the twig infarct region (Xb) is interrupted by the thinning process by the differentiation or the quantization error, when the length is obtained, the interrupted pixels are approximated by a straight line. May be.

Based on the length of the obtained twig, for example, less than 8 mm (15 pixels
If it is (dot equivalent), it is determined that there is no twig infarction.

That is, the process of obtaining the boundary (a) between the grain region (Xa) and the twig branch region (Xb) based on the frequency distribution of pixels having the same coordinate value on each of the x-axis and the y-axis is Corresponding to the extraction means (102), the process of thinning the extracted twig infarct region (Xb) and obtaining its length as the number of pixels corresponds to the length determination means (103).

Next, the operation of the control device (4) will be described based on the flowchart shown in FIG.

Based on the value of the luminance signal (Y) output from the image sensor (3), it is determined that the grain (A) is within the imaging field of view (B). In order to separate from the background, the captured image information of the image sensor (3) is quantized with a set pixel density,
The region (X) corresponding to only the grain (A) is extracted (see FIG. 3).

After extracting the area (X), first, in order to prevent misjudgment due to foreign matter such as dust, the number of pixels belonging to the area (X) is a value significantly smaller than the size of the grain (A). It is determined whether the number of pixels is less than or equal to the set number of pixels (set to 20 dots), and if the area is smaller than the value corresponding to the set number of pixels, the imaged object is dust. It is determined that the foreign matter is such a foreign matter and is excluded from the processing target.

After removing foreign matter such as dust from the processing target, as described above,
The number of pixels belonging to the region (X) is set to a value larger than the size when there is no twig (140)
It is determined whether or not the number of pixels is greater than or equal to (set in dots), and if the number of pixels is greater than or equal to the set number of pixels, it is determined that a twig is included.

When the area of the region (X) is less than the set number of pixels, the value of the difference between the red signal (R) and the blue signal (B) of the three primary color information output from the image sensor (3). Based on whether or not the area of the region extracted based on is less than or equal to the set value (set to 15 dots), it is determined to be brown rice when it is less than or equal to the set value, and is larger than the set value. The case is determined to be paddy.

Then, if it is determined that it is paddy, then, if the length of the twigs obtained as described above is less than the set value (15 dots), it is determined that there is no twig, and the set value is If it is larger, it will be judged as paddy.

That is, when the area of the region (X) is equal to or larger than the set value, and when the calculated length of the twig is larger than the set value, the process for discriminating with the twig is the discriminating means (10).
It corresponds to 1).

Although not described in detail, the process described above is automatically performed until the number of processed samples reaches a set value, based on information such as the presence or absence of a grain (A) determined based on the imaging information. Will be repeated.

[Another embodiment]

In the above-described embodiment, in order to convey the grains (A) to be inspected in a state in which the grains are spaced apart from each other, the vibrating type feeding device (1) and the belt type conveyor (2) having different feeding speeds. ) Is used as an example, but the specific configuration such as a means for capturing an image of each grain (A) can be changed in various ways.

It should be noted that reference numerals are added to the claims for convenience of comparison with the drawings, but the present invention is not limited to the structures of the accompanying drawings by the entry.

[Brief description of drawings]

The drawings show an embodiment of a threshing treatment grain inspection apparatus according to the present invention. Fig. 1 is a block diagram of a control configuration, Fig. 2 is a flow chart of control operation, and Figs. 3 to 9 are explanations of image processing. FIG. 10, FIG. 10 is a schematic plan view of the entire apparatus, FIG. 11 is a schematic side view of the entire apparatus, FIG. 12 is a II-II arrow view of FIG.
10 is an enlarged perspective view of the conveying terminal portion of the supply device, and FIG. 15 is an enlarged side view of the same. (3) ... Imaging means, (X) ... Area corresponding to grain,
(Xb) …… Twig branch region, (100) …… Image processing means, (1
01) …… discrimination means, (102) …… twig extraction means, (10
3) …… A means to determine the length.

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Katsuya Usui 64 Ishizukita-machi, Sakai City, Osaka Prefecture Kubota Iron Works Co., Ltd.Sakai Factory (72) Inventor Toshio Tominaga, 64, Ishizukitamachi, Sakai City, Osaka Prefecture Kubota Iron Works Co., Ltd. Sakai Factory (56) Reference JP-A-2-302878 (JP, A)

Claims (1)

[Claims] 1. An image pickup means (3) for picking up an image of a grain after threshing processing, and an image processing means (3) for extracting a region (X) corresponding to the grain based on image pickup information of the image pickup means (3). 100) and a discriminating means (101) for discriminating the presence or absence of a twig based on the information of the extracted region (X), which is an inspection device for threshing-treated grains. A twig infarct extraction means (102) for extracting a twig infarct region (Xb) corresponding to only the infarction, and a length discrimination means for discriminating the length of the twig infarct based on the information of the twig infarct extraction means (102) ( Ten
3) is provided, and when the area of the region (X) is larger than a set value, the discriminating means (101) determines that the length of the twig is discriminated by the length discriminating means (103). A threshing-processed grain inspection device configured to determine that a twig is present when the value is larger than a set value.