JPH09318547A - Appearance inspection method and apparatus for farm product - Google Patents

Abstract

PROBLEM TO BE SOLVED: To enable accurately drawing of the external contour of a farm product and detecting of light information for obtaining information on color and damage in the contour by photographing the farm product on a carrying conveyor by near infrared rays(NIR) to compare NIR image information of pixels of an NIR image with a predetermined threshold. SOLUTION: Image information in wavelength ranges of NIR, R (red light) and G (green light) per pixel of an image taken by a photodetecting means 5 containing a camera is inputted into a camera measurement controller 6. The controller 6 binary codes a signal of NIR image information by a contour pixel extraction processing part and compares image information of each pixel detected as NIR image information with a specified threshold stored to extract the area of the pixel indicating a value higher than the threshold. The extracted NIR binary image undergoes a computation to extract the image information of the pixel within the contour of the farm product for use in the judgment of color and damage in the image signals NIR, R and G, thus accomplishing a judging processing for the judgment of color, detection of obstacles and the like.

Description

Detailed Description of the Invention

[0001]

TECHNICAL FIELD The present invention relates to apples, persimmons, pears,
The present invention relates to a method and an apparatus for inspecting the appearance of agricultural products used in a sorting field for sorting, packaging, and shipping agricultural products such as tomatoes, eggplants, cucumbers, and mandarins.

[0002]

2. Description of the Related Art The selection of agricultural products in a sorting field is roughly classified into
A method based on class factors such as size and area, color,
Either or both of the method based on grade factors such as scratches are used, and in the judgment based on the class factor of the former, the factors (eg size, weight) used for this are compared. Since it is easy to detect by mechanical means, it has been popular since ancient times, and the optical method and device for detecting this element even in the non-contact method that does not damage recent agricultural products is relatively easy to implement and automate. Many automated devices have been proposed and put to practical use.

However, in the latter case of the grade determination for grade factors such as color and scratches, there is still a problem to be solved although an automatic device has been proposed and partly put to practical use.

Known as a conventionally proposed apparatus for detecting an element such as color information for grade determination is as follows.
For example, by using a CCD camera or the like to capture a color image of an agricultural product that is transported on a transport conveyor, red (R),
A method of detecting color information of green (G) and blue (B) ternary colors (Japanese Patent Publication No. 61-7360 and Japanese Patent Publication No. 61-11).
No. 679, Japanese Patent Publication No. 1-33225, and Japanese Patent Laid-Open No. Hei 4
No. 147042, Japanese Unexamined Patent Publication No. 4-238252, etc.) and a method of detecting a “scratch” on the surface of an agricultural product by a similar method (Japanese Unexamined Patent Publication No. 4-307357).

In these methods, the bucket, which is a conveyer, the saucer (so-called free tray), the narrow slat conveyer, etc., are made black so that the brightness difference from the conveyed produce is easily detected, and the color information about the produce is properly detected. A configuration that allows for this is generally adopted.

[0006] However, agricultural products come in a wide variety of colors and have low reflected lightness (typical examples are black purple "eggplant", dark purple "starking", dark green "blue cut oranges").
When determining the grade, it is possible to accurately capture the outline of the agricultural product because the boundary between the two is not clear even if the color information is detected while the agricultural product is being transported by the transportation means having a black color as described above. In many cases, the problem is that the accuracy of determination is low. For example, there is one in which an image of an agricultural product captured by using a CCD camera or the like is used as a pixel signal in row a and column b as a data signal for determining a color or a scratch, but a blackish agricultural product or dust is attached. The outlines of agricultural products such as potatoes were not clear, and as a result, the reflected light of the conveyor portion, which was originally unrelated to the agricultural products, was captured as a part of the agricultural products, and the reliability of the determination accuracy was sometimes lowered.

In order to deal with this, for example, there is a method of extracting a conveyor or the like with a background color opposite to that of the object to be measured, but the conveyor has image extraction performance in a short time due to dirt or the like. Is deteriorated, and if the conveyor must be replaced for each target agricultural product, the versatility of the device will be reduced, leading to a cost disadvantage, and the device parts at the sorting site. There is a problem that it is extremely difficult to deal with the actual situation in which a large number of sorting fields are scattered, because an engineer with considerable knowledge is required for the exchange of.

By the way, as a method of analyzing the color of the surface of an agricultural product, which is a grade judgment element, the reflected light detected by the above-mentioned imaging is divided into several wavelength ranges and extracted as R, G, B.
In addition to the known method for analyzing saturation, hue, and lightness, green light (G) and red light (R) cause a change in color as in Japanese Patent Publication No. 3-21235. Infrared light (IR) is less sensitive to color changes than it is sensitive, so there is a known method to judge the color from the result of calculating IR / R and IR / G by utilizing the difference in the characteristics. There is. It is also known that infrared light has a high reflectance for agricultural products but a low reflectance for black resin conveyors and the like.

[0009]

The present inventor has accomplished the present invention by earnestly conducting research under the above-mentioned conventional techniques.

That is, in the above-mentioned conventional grade determination for selecting agricultural products such as colors and scratches, generally, in order to calculate and analyze the colors and scratches by a computer, the reflection of visible light region sensitive to the colors from the agricultural products is generally performed. Since it is necessary to measure the light, the reflected light in the visible light region from the agricultural product is received by an imaging means such as a CCD camera, and the grade factor such as the color of the agricultural product is determined based on this light information. In some agricultural products (for example, black-based agricultural products), the visible light color information does not clearly show the boundaries of the agricultural products, and whether the color information near the outline of the agricultural products indicates whether it is derived from the conveyor or the color and scratches of the agricultural products. It is not clear, which causes a problem that colors and scratches cannot be accurately determined including the vicinity of the outline of agricultural products.

Therefore, the present invention is a method and apparatus capable of accurately cutting out the outer contour of an agricultural product, accurately detecting the color and scratch of the agricultural product in the vicinity of the contour, and enabling highly accurate grade determination. It is made for the purpose of providing.

Another object of the present invention is to provide a method and a device capable of accurately cutting out the outer contour of an agricultural product and detecting optical information for detecting color and scratch information in the contour with a simple device. Where it is.

[0013]

The features of the present invention for achieving the above object are as set forth in the claims of the appended claims.

In the invention of the appearance inspection method for agricultural products according to claim 1 of the present application, the agricultural products on the conveying means are imaged with near-infrared light to obtain a near-infrared light image consisting of pixels in row a and column b. The near-infrared light image information of each pixel is compared with a predetermined threshold value to extract the pixels in the contour area of the agricultural product, and the visible light image obtained by imaging the agricultural product on the conveying means with visible light. It is characterized in that the color and / or the scratch of the agricultural product is determined based on the visible light image information of each pixel in the agricultural product outline.

The agricultural products to which the present invention is applied include colors,
The fruits, vegetables and the like whose grade is determined by measuring the scratches can be mentioned without particular limitation.

In the present invention, near-infrared light for imaging agricultural products is insensitive to color changes and has a high reflectance for agricultural products and a high reflectance for resin materials that constitute a conveying means such as a conveyor. Light having a low property in a wavelength range, generally light having a wavelength of 750 nm to 900 nm is preferably used, but the light is not limited as long as the property is satisfied. Further, as the visible light, light having a wavelength suitable for detecting and determining the color and scratch of the agricultural product is used.

According to the above-mentioned invention, the near-infrared light image information of each pixel received and detected by the near-infrared light is insensitive to the change in color and has a high reflectance for agricultural products and for a conveyor or the like. Has a low reflectance, a predetermined threshold value that is sufficiently lower than the reflected light from the agricultural products and sufficiently higher than the reflected light from the conveyor, and the near-infrared light image of each pixel. By comparing the information, it is possible to clearly distinguish whether each pixel is located inside or outside the outline of the agricultural product.
That is, by this, it is possible to accurately extract the pixels (group) existing in the area within the outline of the agricultural product (extract the outline accurately).

Then, only the pixels (groups) in the contour area that are accurately cut out (extracted) in this way (that is, the pixels outside the contour that have not been extracted are excluded) are received with color-sensitive visible light. Since information on the detected visible light image is used to determine either or both of color and scratch, it is possible to perform highly accurate grade determination. The elements and the calculation method used for determining the color and the scratch can be performed by using a conventional method.

As described above, the contour of the agricultural product is cut out by using the near-infrared light, which is insensitive to the change of the color, but obtains the image information which is sensitive to the difference between the type of the agricultural product and the other object, and the change of the color is detected. Since the color and the flaw are determined based on the information of the pixel (group) in the contour of visible light that can obtain sensitive image information, accurate and efficient determination can be performed.

In the invention of claim 2 of the present application, the visible light information for judging the color and / or the scratch of the agricultural product is at least light information in the wavelength range of red light (R) and green light (G). It is characterized in that it includes, and if necessary, optical information in the wavelength range of the blue light (B).

The invention according to claim 3 of the present application is the same as the invention according to claim 1 described above, in which the pixels in the area inside the outline of the agricultural product extracted by comparing the near-infrared light image information with a predetermined threshold value,
The visible light image information of each pixel in the agricultural product contour of the visible light image obtained by imaging the agricultural product on the transport means with visible light, and the near infrared light image information of each pixel in the agricultural product contour, the agricultural product It is characterized by determining the color and scratches of, or scratches.

According to this invention, in addition to the invention of claim 1, the near-infrared light image information is used for detecting scratches on agricultural products.
It is possible to detect scratches with higher accuracy. in this case,
Scratch detection by near-infrared light image information and visible light image information, it is also possible to determine the scratch of the agricultural product by the logical sum of the pixels that are regarded as scratches detected by both, and the logical product of the pixels It is also possible to judge the scratch. The latter method using the logical product can be preferably adopted to avoid the problem that non-true scratches such as mud adhesion are judged to be scratches. The method of using both pixel information for scratch detection may be designed by empirically and experimentally confirming the relationship with what is recognized as an actual scratch according to the characteristics of various agricultural products.

The invention of claim 4 of the present application is characterized in that imaging of agricultural products by near infrared light and imaging of agricultural products by visible light are performed simultaneously by a single imaging means.

According to the present invention, the appearance inspection can be performed quickly, and the operation of matching the image captured by the near infrared light with the image captured by the visible light is unnecessary. It is possible to accurately ensure that the visible light image information for determining the scratch is based on the pixels within the outline of the produce.

According to the invention of claim 5 of the present application, the agricultural products on the conveying means are imaged with near-infrared light to obtain a near-infrared light image of the pixels in row a and column b. By comparing the image information with a predetermined threshold value to extract the pixels in the area within the outline of the agricultural product, it is possible to detect the scratch of the agricultural product by the near-infrared light image information of each pixel within the extracted agricultural product outline. Characterize.

According to the present invention, for agricultural products for which color determination is not necessary, extraction of pixels in the contour accurately cut out based on only the near-infrared light image information and proximity of each pixel in the contour. It is possible to make a highly accurate determination based on the infrared light image information.

According to a sixth aspect of the present invention, there is provided an apparatus for inspecting an appearance of agricultural products, which comprises a conveying means for conveying the agricultural products, a light source for irradiating the conveyed agricultural products with illumination light including a near infrared light region and visible light, and Light receiving means for extracting near-infrared light image information and visible light image information for each pixel in row a, column b of an image of an agricultural product illuminated by a light source, and near-infrared light image information for each pixel are determined in advance. An in-contour pixel extracting means for extracting pixels in an in-contour region of the agricultural product by comparing with the threshold value, and a visible light image information extracting means for extracting visible light image information of each pixel extracted by the in-contour pixel extracting means; And a determining unit that determines the color and / or the scratch of the agricultural product based on the visible light image information extracted by the visible light image information extracting unit.

The conveying means is a tray (so-called free tray) which is conveyed on the belt conveyor, roller conveyor, bucket conveyor, belt conveyor, etc. without being constrained,
Examples include so-called slat conveyors in which a large number of narrow slats are aligned in the transport direction.

As a light source for illuminating agricultural products, a light source for emitting near infrared light and a light source for emitting visible light can be used.
It is also possible to use a band-pass filter for transmitting a near-infrared light region and visible light in the light receiving optical system by using, for example, a halogen lamp whose emission wavelength region is not limited.

A CCD camera, a video camera or the like can be used as the image pickup means, and the image information of the agricultural products picked up by this image pickup means and the image information in the area outside the contour are divided into pixels in a matrix of a rows and b columns. Is a collection of information of
The size of the matrix of row a and column b is determined by design depending on the resolution of the light receiving optical system and the size of the agricultural product. In addition, it is preferable that the size of each pixel is smaller because it can cut out pixels that match the actual outline of the agricultural product, but on the other hand, the number of pixels increases and it takes time for image processing. What is designed by design according to requirements such as cost, processing speed, and grade determination accuracy is used.

The light receiving means for extracting the near-infrared light image information and the visible light image information is a lens system that collects the reflected light from the agricultural product, and a CCD on which an optical image from this lens system is projected.
It can be constituted by an image pickup means such as a CCD camera or a video camera provided with an image pickup element such as the above, and an image processing means for image-processing the image pickup information obtained by the image pickup means by using a computer. For this purpose, the structure can be designed in various ways although it is related to the light source, for example, a light source that emits near-infrared light for obtaining infrared image information, an imaging unit that receives the light source, and a visible light image. It is possible to separately provide a light source that emits visible light for obtaining information and an image pickup unit that receives the light, and use a common light source whose emission wavelength range is not limited, and the near infrared light image information and the visible light image. Information can be obtained by receiving light by separate image pickup means, or near infrared light image information and visible light can be obtained by using the common light source and the spectroscopic means (a predetermined bandpass filter if necessary). Image information can also be obtained through a common lens system.

The in-contour pixel extracting means sets image information for each pixel divided into a matrix of a rows and b columns of a near infrared light image obtained by picking up an image of a farm product with a threshold value set in advance in a memory or the like. It is a means for comparison and is generally constructed by using a computer, and the threshold value for the comparison can be experimentally and empirically determined.

According to the present invention, there is provided a device for accurately cutting out pixels in the outer contour of a farm product based on near-infrared light image information, and detecting the color and scratches of the farm product with high accuracy based on the visible light image information of the cut out pixels. Can be provided.

According to a seventh aspect of the present invention, there is provided a conveying means for conveying an agricultural product, a light source for irradiating the conveyed agricultural product with illumination light including a near infrared light region and visible light, and an agricultural product illuminated by the light source. Light receiving means for extracting near-infrared light image information and visible light image information for each pixel in row a, column b of the captured image;
Near-infrared light image information of each pixel is compared with a predetermined threshold value to extract a pixel in the contour area of the agricultural product, and a visible light of each pixel extracted by the contour pixel extraction means. Visible light image information extraction means for extracting image information,
A first scratch judging means for judging a scratch on an agricultural product based on the visible light image information extracted by the visible light image information extracting means, and a near infrared light image information for each pixel extracted by the in-contour pixel extracting means. An infrared light signal extracting means and a second scratch judging means for judging a scratch on an agricultural product based on the near infrared light image information extracted by the near infrared light signal extracting means are provided, and the first and second scratches are provided. A feature of the present invention is that the damage of the agricultural product is judged by integrating the results of the judgment means.

According to the present invention, it is possible to provide a device for detecting a scratch on an agricultural product with higher accuracy by using either or both of near infrared light image information and visible image information.

The invention according to claim 8 of the present application provides a transporting means for transporting agricultural products, a light source for irradiating the transported agricultural products with illumination light including a near infrared light region and visible light, and the agricultural products illuminated by the light source. Light receiving means for extracting near-infrared light image information and visible light image information for each pixel in row a, column b of the captured image;
Near-infrared light image information of each pixel is compared with a predetermined threshold value to extract a pixel in the contour area of the agricultural product, and a visible light of each pixel extracted by the contour pixel extraction means. Visible light image information extraction means for extracting image information,
The first determination means for determining the color and / or the scratch of the agricultural product based on the visible light image information extracted by the visible light image information extraction means, and the near infrared light image information of each pixel extracted by the in-contour pixel extraction means It is characterized by comprising near-infrared light signal extracting means for extracting and second judging means for judging a scratch on the agricultural product based on the near-infrared light image information extracted by the near-infrared light signal extracting means.

According to the present invention, the detection of the color of agricultural products,
It is possible to provide an apparatus for highly accurately detecting a flaw using either or both of near-infrared light image information and visible image information.

[0038] The invention of claim 9 of the present application is directed to a transport means for transporting agricultural products, a light source for irradiating the transported agricultural products with illumination light including near-infrared light, and an image of an image of the agricultural products illuminated by the light source. By comparing the near-infrared light image information of each pixel of each pixel in row a and column b with the near-infrared light image information of each pixel with a predetermined threshold value, the pixels in the area within the contour of the agricultural product are detected. In-contour pixel extracting means for extracting, near-infrared light signal extracting means for extracting near-infrared light image information of each pixel extracted by this in-contour pixel extracting means, and for extracting by the near-infrared light signal extracting means It is characterized by comprising a determining means for determining a scratch of the agricultural product based on the near-infrared light image information.

According to the present invention, it is possible to provide a device for accurately cutting out the outer contours of agricultural products and detecting scratches with high accuracy by using only near-infrared light image information.

The invention according to claim 10 of the present application is characterized in that, in the apparatus according to each of the above-mentioned claims, the near-infrared light image and the visible light image are simultaneously received and imaged by one imaging means.

According to the present invention, the imaged image by the near infrared light for cutting out the outer contour of the agricultural product by the near infrared light and the imaged image by the visible light for detecting the color and the scratch are separated by the different cameras (etc. Compared to a device that uses the image pickup means (1), it is superior in that it is not necessary to align the cameras and the like that match the angle of view of both images, and high detection accuracy of colors and scratches can be secured. Further, according to the present invention, since the near-infrared light and the visible light are simultaneously imaged by one imaging means, even if the agricultural product may move on the transportation means such as the transportation conveyor, the near-infrared light image information and It is possible to detect colors and scratches with high accuracy without having to consider the image shift of visible light image information at all.

[0042]

BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1 FIG. 1 shows a schematic configuration of a fruit selection equipment to which a device for selecting grades of fruits (agricultural products) to which the appearance inspection device for agricultural products of this example is applied is shown.
Are apples, which are carried one by one on free trays (containers) 3 which are carried by the conveyor 2 which rotates endlessly without any gap.

Reference numeral 4 denotes a light source lamp such as a halogen lamp arranged near the image pickup stage for detecting colors and scratches, which is light including the wavelength range of near infrared light and the visible light range.
Illuminate.

Reference numeral 5 denotes a light receiving means including a camera (imaging means) having a CCD as a light receiving element, and details of the light receiving means including this camera are shown in FIG. That is, this FIG.
The light receiving means 5 shown in FIG.
The reflected light from the lens 51 passes through the spectroscope 52 to near infrared light (“NEAR INFRARED RAYS”: hereinafter abbreviated as “NIR”), red light (hereinafter abbreviated as “R”), green light (hereinafter (Hereinafter abbreviated as "G"), each projected image is divided into NIR CCD 531, R CCD 532, and G C.
Project onto CD533. The image projected and picked up by each CCD is passed through the amplification circuits 541, 542, 543, and then the A / D conversion circuits 551, 552, 553,
It is output as NIR digital output, R digital output, and G digital output. In order to extract the spectrum of each wavelength range of NIR, R, G, in this example, a bandpass filter (not shown) is provided inside the spectroscope 52 for transmitting only the light of the corresponding predetermined wavelength range. ing. Further, the CCDs 531, CCDs 532, and CCDs 533 of the light receiving means 5 are provided so as to have the same number of pixels, and the above-mentioned amplifier circuits and A / D conversion circuits are synchronized by the drive circuit 56 that receives the drive signal input. It is provided so as to be driven, whereby the pixels at the corresponding addresses of the matrix of row a and column b of the image picked up by each CCD are controlled so as to have a matching relationship.

In this way, the light receiving means 5 including the camera
The image information of the NIR, R, and G wavelength regions for each pixel of the image captured in (1) is input to the camera measurement control device 6 including a computer.

Then, in the camera measurement control device 6, the grade of the color and scratch of the apple which is the processing target agricultural product of this example is judged by a predetermined arithmetic processing based on the input signal described later, and the judgment result is the agricultural product. It is output as a sorting signal of. Normally, the class element is measured, and the class determination based on the class element is performed to perform the classification determination based on the class / class, but the description of the class is omitted in this example.

[0047] _{7 1, 7 2, ···· 7} i is the solenoid of the sorting sorting device which receives a sorting signal output from the camera measurement control unit 6, the agricultural products to grading to Grade 1 Grade i When the free tray 3 installed at a predetermined position for sorting and sorting and the agricultural products of the corresponding grade classification come to the sorting position, the free tray 3 is discharged to a boxing branch discharge conveyor (not shown). Works like.

Reference numeral 8 is a synchronization signal generator for generating a signal in synchronization with the conveyance of the conveyer 2. The conveyer conveyer synchronization signal generated by this is inputted to the camera measurement controller 6 and is received from the light receiving means 5. It is used to output the sorting position of a predetermined agricultural product 1 whose grade is determined by the input imaging information of the agricultural product to the sorting and sorting devices 7 ₁ , 7 ₂ , ... 7 _i .

The camera measurement control device 6 is configured as the image processing / control circuit shown in the block diagram of FIG. That is, the NI extracted by the light receiving means 5 in FIG.
Signals of R, R, and G image information and synchronization signals (these are collectively referred to as "camera signals") are input to the video input I / F 601 shown in FIG.
R, R, and G are stored in the frame memory 602 as image information of each signal. In addition, this frame memory 6
Each image information stored in 02 is output as a video monitor signal from the image bus 603 to the video monitor (not shown) via the video D / A 604, and is also sent from the image bus 603 to the image processor 605.
Reference numeral 606 denotes a LUT look-up table memory for pre-processing, which performs pre-processing such as enhancement of each extracted image and removal of noise, and then stores this image in a memory. A plane memory 607 functions as a buffer for temporarily storing color information of each pixel obtained by processing by the image processor 605. Image processor 605, LUT memory 606, plane memory 60 described above
Based on the signal of No. 7, the processing unit (CPU) 609 performs image processing via the CPU bus 608. Also, I / O6
A sorting and sorting signal is sent via 10 to the sorting and sorting device 7
It outputs to the solenoids of ₁ , 7 ₂ , ..., 7 _i . 611 is a ROM and 612 is a RAM.

The image processing performed by the processing device 609 will be described with reference to FIG. That is, in this example, the signal of NIR image information (multivalued image information in which each pixel has various values) is first binarized by the in-contour pixel extraction processing unit. This binarization process compares a predetermined threshold value stored in the ROM 611 (or the RAM 612) with the image information of each pixel detected as NIR image information, and this NIR is higher than that of a resin material such as a conveyor. By using the property of exhibiting a high reflectance with respect to, the region of pixels that show a value higher than the threshold value is extracted as the outline of the agricultural product, and the extracted pixel groups are numbered in the order of their connected regions. (Labeling process). After that, for each label image, after it is determined whether or not the label image is the inspection object, only the image of the inspection object is extracted.

The NIR binary image obtained by the in-contour pixel extraction processing section is sent to the class element shape processing section in this example, and is processed to measure class elements such as size. In addition, the NIR binary image includes each image information signal NI.
An AND operation between pixels is performed in the detection area specifying processing unit in order to extract the image information of the pixel in the agricultural product contour used for judging the color and the scratch among R, R, and G, and the NIR detection area and the R detection area, respectively. , Image information of pixels in the G detection area is extracted. The image of the extracted pixels in each wavelength range is a multivalued image in which the value of each pixel depends on the detection value.

Then, N in which these detection areas are specified
Predetermined image information of the image information of each wavelength range of IR, R, G is output to the color determination processing unit or the injury detection processing unit, and based on this, color determination (coloring degree, color uniformity, etc.),
Performs injury detection (maximum area, injury area, etc.). As a processing method performed in the color determination processing unit and the damage detection processing unit, a conventionally known method can be used. For example, in the damage detection processing, a differential processing method or a color difference processing method for comparing with a known damage color value. And so on.

In FIG. 4, all the image information in the NIR, R, and G wavelength regions are input to the color determination processing section and the injury detection processing section, but this is a signal transmission / reception system. It is shown and does not mean that all signals are always available. Although unnecessary signal transmission / reception paths can be omitted depending on the device, these are actually modularized contents of the processing performed using the computer technology described in FIG. Therefore, it is not necessary to have each of these processing units in terms of hardware.

FIG. 5 shows the inspection operation performed by using the above-described agricultural product appearance inspection apparatus according to the above claims 1, 2, 4, 6, and 10.
The flowchart for explaining as an example corresponding to is shown. First, in step 101, a near infrared light image (NIR 750 to 900 nm) and a visible light image (R, G) are captured. In this case, in this case, the near infrared light and the visible light are simultaneously captured by one camera (imaging unit) in this example.

Next, in step 102, based on the near-infrared light image, the threshold value predetermined for cutting out the contour of the agricultural product is compared with the near-infrared light image information of each pixel to be binarized, and the contour is contoured. Pixels in the inner area are extracted. Next, in step 103, the AND operation between the extracted pixels in the contour area and the visible light image is performed to specify the visible light image information of the contour area.

In step 104, the color of the in-contour region (hence the color of the agricultural product) is calculated and calculated based on the visible light image information of the in-contour region specified in step 103. For example, a color histogram is created according to a conventionally known color calculation formula, and the color uniformity and the color at the center of the image are detected.

Finally, in step 105, the detected color is evaluated and judged, the grade of the color of the corresponding agricultural product is evaluated, and the process is terminated. The above processing is sequentially repeated for the agricultural products (apples) transported by the transport conveyor 2.

Embodiment 2 FIG. 6 shows a flow chart for explaining an inspection operation performed by using an appearance inspection apparatus for agricultural products as an example corresponding to claims 5 and 9.

In this example, in step 201, a near infrared light image (NIR 750 to 900 nm) is picked up. In step 202, based on the near-infrared light image, binarization is performed by comparing a predetermined threshold value for cropping the outline of the agricultural product with the near-infrared light image information of each pixel, and the pixels in the in-contour region are extracted. Extract.

Next, in step 203, the near-infrared light image information of the in-contour region is specified by performing an AND operation between the extracted pixels in the in-contour region and the near-infrared light image.

In step 204, based on the near-infrared light image information of the contour area specified in step 203, binarization is performed by, for example, a known binarization method to create a scratch image.

Finally, in step 205, the detected scratch is evaluated and judged, and the grade of the scratch of the corresponding agricultural product is evaluated, and the process is terminated. The above processing is sequentially repeated for the agricultural products (apples) transported by the transport conveyor 2.

Embodiment 3 FIG. 7 shows a flow chart for explaining an inspection operation performed by using an appearance inspection apparatus for agricultural products as an example corresponding to the above claims 3 and 8. First, step 3
01 near infrared light image (NIR 750-900n
m) and a visible light image (R, G). In this case, in this case, the near infrared light and the visible light are simultaneously captured by one camera (imaging unit) in this example.

Next, in step 302, based on the near-infrared light image, the threshold value predetermined for cutting out the outline of the agricultural product is compared with the near-infrared light image information of each pixel to be binarized to obtain the outline. Pixels in the inner area are extracted.

Next, in step 303, AND operation is performed between the extracted pixels in the contour area and the near-infrared light image and the visible light image to perform near-infrared image information and visible light image information in the contour area. Specify.

In step 304, based on the near-infrared light image information of the in-contour region identified in step 303, binarization is performed by, for example, a well-known binarization method to create a scratch image, and an injury is generated. Specify and calculate the scratch area.

In step 305, the color of the in-contour region (hence the color of the agricultural product) is calculated and calculated based on the visible light image information of the in-contour region specified in step 303. For example, a color histogram is created according to a conventionally known color calculation formula, and the color uniformity and the color at the center of the image are detected.

In step 306, the above-mentioned detected scratches are evaluated and judged, and the grade of the scratches of the corresponding agricultural product is evaluated.

In step 307, the detected color is evaluated and judged, and the grade of the color of the corresponding agricultural product is evaluated, and the process is terminated. The above processing is sequentially repeated for the agricultural products (apples) transported by the transport conveyor 2.

Embodiment 4 FIG. 8 shows a flow chart for explaining an inspection operation performed by using an appearance inspection apparatus for agricultural products as an example corresponding to claim 7 above. First, step 401.
Near-infrared light image (NIR 750-900nm)
And a visible light image (R, G) is captured. In this case, in this case, the near infrared light and the visible light are simultaneously captured by one camera (imaging unit) in this example.

Next, at step 402, based on the near-infrared light image, the threshold value predetermined for cutting out the contour of the agricultural product is compared with the near-infrared light image information of each pixel to be binarized to obtain the contour. Pixels in the inner area are extracted.

Next, in step 403, the AND pixels between the extracted pixels in the contour area and the near-infrared light image and the visible light image are respectively subjected to AND operation to obtain the near-infrared light image and the visible light image in the contour area. Identify the information.

In step 404, the flaw image (A) is created by binarizing the near-infrared light image information of the contour area specified in step 403 by, for example, a known binarization method. To do.

In step 405, a flaw image (B) in the contour area is created based on the visible light image information of the contour area specified in step 403, for example, by a known differentiation method or color difference method.

In step 406, based on the detected scratch images (A) and (B), the scratches are comprehensively evaluated and the grade of the scratch of the corresponding agricultural product is evaluated, and the process is terminated. The above processing is sequentially repeated for the agricultural products (apples) transported by the transport conveyor 2.

Embodiment 5 FIG. 9 shows a flow chart for explaining an inspection operation performed by using an appearance inspection apparatus for agricultural products as an example corresponding to the above claims 3 and 8. First, step 5 is shown.
01 near infrared light image (NIR 750-900n
m) and a visible light image (R, G). In this case, in this case, the near infrared light and the visible light are simultaneously captured by one camera (imaging unit) in this example.

Next, in step 502, based on the near-infrared light image, the threshold value predetermined for cutting out the contour of the agricultural product is compared with the near-infrared light image information of each pixel to be binarized to obtain the contour. Pixels in the inner area are extracted.

Then, in step 503, AND operation between the extracted pixels in the contour area and the near-infrared light image and the visible-light image is performed respectively to perform the near-infrared light image and the visible-light image in the contour area. Identify the information.

In step 504, a flaw image (A) is created by performing binarization by a known binarization method, for example, based on the near-infrared light image information of the contour area specified in step 503. Then, the injury is identified and the wound area is calculated.

In step 505, a flaw image (B) in the contour area is created based on the visible light image information of the contour area specified in step 503, for example, by a known differentiation method or color difference method.

In step 506, the color of the in-contour region (hence the color of the agricultural product) is calculated and calculated based on the visible light image information of the in-contour region specified in step 503. For example, a color histogram is created according to a conventionally known color calculation formula, and the color uniformity and the color at the center of the image are detected.

At step 507, the comprehensive evaluation of the scratches is made based on the detected scratch images (A) and (B), the grade of the scratch of the corresponding agricultural product is evaluated, and the process is terminated. The above processing is sequentially repeated for the agricultural products (apples) transported by the transport conveyor 2.

In step 508, the detected color is evaluated and judged, and the grade of the color of the corresponding agricultural product is evaluated.
The process ends. The above processing is sequentially repeated for the agricultural products (apples) transported by the transport conveyor 2.

[0084]

According to the invention of claim 1 of the present application, it is possible to accurately cut out the outer contour of an agricultural product by using an image picked up by near infrared light, and at the same time an image picked up by visible light (necessarily Depending on the near infrared light image), it is possible to accurately detect the color and scratches in the outline of the produce, and to judge the grade with high accuracy. The effect that the optical information for detecting the color and scratch information can be detected with a simple device, and further the following effects can be obtained.

According to the invention of claim 3, in addition to the invention of claim 1, the near-infrared light image information is used for detecting scratches on agricultural products, so that scratches can be detected with higher accuracy. In addition, it is also possible to detect a scratch by determining the scratch of the agricultural product by the logical sum of the pixels considered to be the scratch detected by the near infrared image information and the visible light image information, and the scratch of the agricultural product by the logical product of the pixels. It is also possible to make a judgment, and when the latter logical product is used, it is possible to avoid the problem of judging that something that is not a true scratch, such as mud adhesion, is a scratch.

According to the invention of claim 4, wherein the image pickup by the near-infrared light and the visible light is simultaneously performed by a single image pickup means, the appearance inspection can be carried out quickly, and the image picked up by the near-infrared light,
It is not necessary to match the images picked up with visible light, and it is possible to more accurately determine the color and scratches in the contour of the agricultural product.

According to the fifth aspect of the present invention, regarding the agricultural products for which color determination is unnecessary, the scratches are generated based on the near-infrared light image information of the pixels in the contour that are accurately cut out based on only the near-infrared light image information. Can be determined with high accuracy.

According to the invention of the appearance inspection apparatus for agricultural products according to claim 6, an accurate contour cutting of the contour of the agricultural product based on the near-infrared light image information and the agricultural product utilizing the visible light image information of the pixels in the cut-out contour. Highly accurate detection of color and scratches.

According to the invention of claim 7, it is possible to provide an apparatus for detecting a scratch of an agricultural product with higher accuracy by using either or both of near-infrared light image information and visible image information.

According to the eighth aspect of the present invention, it is possible to provide an apparatus for highly accurately detecting the color of an agricultural product and the flaw detection using either or both of the near infrared light image information and the visible image information.

According to the ninth aspect of the present invention, it is possible to provide an apparatus for accurately cutting out the outer contour of an agricultural product and detecting a scratch with high accuracy by using only near infrared light image information.

According to the tenth aspect of the invention, the imaged image by the near infrared light for cutting out the outer contour of the agricultural product by the near infrared light and the imaged image by the visible light for detecting the color and the scratch are separated. Compared to a device that uses a camera (such as image capturing means),
It is possible to eliminate the need for alignment of the cameras and the like for matching the angle of view of both images, and moreover, an excellent effect of ensuring high detection accuracy of colors and scratches can be obtained. Further, since the near-infrared light and the visible light are simultaneously imaged by one imaging means, even if the produce moves on the transportation means such as the transportation conveyor, the image of the near-infrared light image information and the visible light image information is displayed. It is possible to detect colors and scratches with high accuracy without having to consider the deviation of the color at all.

[Brief description of drawings]

FIG. 1 is a diagram showing an example of a schematic configuration of a fruit selection facility to which the present invention is applied.

FIG. 2 is a diagram showing a configuration of an image pickup unit which is a part of a light receiving unit used in the present invention.

FIG. 3 is a block diagram showing an example of a schematic configuration of a camera measurement control device of the present invention.

FIG. 4 is a diagram showing an example of the processing contents of the method for inspecting the appearance of agricultural products according to the present invention in a modular form.

FIG. 5 is a flowchart for explaining the first embodiment of the inspection operation performed using the agricultural product visual inspection apparatus.

FIG. 6 is a flowchart illustrating a second embodiment of the inspection operation performed by using the agricultural product appearance inspection device.

FIG. 7 is a flowchart illustrating a third embodiment of the inspection operation performed using the agricultural product appearance inspection device.

FIG. 8 is a flowchart illustrating a fourth embodiment of an inspection operation performed by using the agricultural product appearance inspection device.

FIG. 9 is a flowchart illustrating a fifth embodiment of the inspection operation performed using the agricultural product appearance inspection device.

[Explanation of symbols]

1 ... Apple (agricultural product) 2 ... Conveyor 3 ... Free tray (container) 4 ... Light source lamp 5 ... Light receiving means 51 ... Lens 52 ... Spectrometer 531 ... CCD for NIR 532 ... CCD for R 533 ... CCD for G 541 ... Amplification circuit 542 ... Amplification circuit 543 ... Amplification circuit 551 ... A / D conversion circuit 552 ... A / D conversion circuit 553 ... A / D conversion circuit 56 ... Driving circuit 6 ... Camera measurement control device 601 ... Video I / F 602 ... Frame memory 603 ... Image bus 604 ... Video D / A 605 ... Image processor 606 ... LUT memory 607 ... Plane memory 608 ... CPU bus 609 ... Processing device 610 ... I / O 611 ... ROM 12 ··· RAM 7 ₁ ~7 _i ··· sorting sorting device (solenoid) 8 ... synchronous signal generator

Claims (10)

[Claims] 1. A near-infrared light image consisting of pixels in row a and column b is obtained by imaging the agricultural products on the conveying means with near-infrared light, and the near-infrared light image information of each of these pixels is predetermined. While extracting the pixels in the outline region of the agricultural product by comparing with the threshold value,
Appearance inspection of the agricultural product, characterized in that the color and / or the scratch of the agricultural product is determined based on the visible light image information of each pixel in the agricultural product outline of the visible light image obtained by imaging the agricultural product on the conveying means with visible light. Method. 2. The visible light information for determining the color and / or scratch of an agricultural product according to claim 1, includes at least light information in the wavelength range of red light (R) and green light (G), and Accordingly, a method for inspecting the appearance of agricultural products, characterized by including light information in the wavelength range of blue light (B). 3. A near-infrared light image consisting of pixels in row a and column b is obtained by imaging the agricultural products on the conveying means with near-infrared light, and the near-infrared light image information of each of these pixels is predetermined. While extracting the pixels in the outline region of the agricultural product by comparing with the threshold value,
The visible light image information of each pixel in the agricultural product contour of the visible light image obtained by imaging the agricultural product on the transport means with visible light, and the near infrared light image information of each pixel in the agricultural product contour, the agricultural product A method for inspecting the appearance of agricultural products, which comprises determining the color and scratches of the, or scratches. 4. The agricultural product according to claim 1, wherein the imaging of the agricultural product with near infrared light and the imaging of the agricultural product with visible light are performed simultaneously by a single imaging means. Appearance inspection method. 5. A near-infrared light image consisting of pixels in row a and column b is obtained by imaging the agricultural products on the conveying means with near-infrared light, and the near-infrared light image information of each of these pixels is predetermined. While extracting the pixels in the outline region of the agricultural product by comparing with the threshold value,
A method for inspecting the appearance of a farm product, comprising detecting a scratch on the farm product based on the near-infrared light image information of each pixel within the extracted contour of the farm product. 6. A conveyance means for conveying agricultural products, a light source for irradiating the conveyed agricultural products with illumination light including a near-infrared light region and visible light, and a row b of an image of the agricultural products illuminated by the light source. Light-receiving means for extracting near-infrared light image information and visible-light image information for each pixel in a row, and comparing the near-infrared light image information for each pixel with a predetermined threshold value, and pixels in the area within the contour of the agricultural product And a visible light image information extracting means for extracting visible light image information of each pixel extracted by the in-contour pixel extracting means, and a visible light image extracted by the visible light image information extracting means. An appearance inspection device for agricultural products, comprising: a determining unit that determines color and / or scratches of agricultural products based on information. 7. A transporting means for transporting agricultural products, a light source for irradiating the transported agricultural products with illumination light including a near-infrared light region and visible light, and a row b of an image of the agricultural products illuminated by the light source. Light-receiving means for extracting near-infrared light image information and visible-light image information for each pixel in a row, and comparing the near-infrared light image information for each pixel with a predetermined threshold value, and pixels in the area within the contour of the agricultural product And a visible light image information extracting means for extracting visible light image information of each pixel extracted by the in-contour pixel extracting means, and a visible light image extracted by the visible light image information extracting means. First scratch judging means for judging a scratch of an agricultural product based on information, near infrared light signal extracting means for extracting near infrared light image information of each pixel extracted by the in-contour pixel extracting means, and the near infrared light Agricultural products based on near-infrared light image information extracted by signal extraction means The second aspect of the present invention is a visual inspection apparatus for agricultural products, comprising: a second scratch determining means for determining the damage of the agricultural product, wherein the results of the first and second scratch determining means are combined to determine the damage of the agricultural product. 8. A conveyance means for conveying agricultural products, a light source for irradiating the conveyed agricultural products with illumination light including a near-infrared light region and visible light, and a row b of an image obtained by imaging the agricultural products illuminated by the light source. Light-receiving means for extracting near-infrared light image information and visible-light image information for each pixel in a row, and comparing the near-infrared light image information for each pixel with a predetermined threshold value, and pixels in the area within the contour of the agricultural product And a visible light image information extracting means for extracting visible light image information of each pixel extracted by the in-contour pixel extracting means, and a visible light image extracted by the visible light image information extracting means. First determining means for determining the color and / or scratch of the agricultural product based on the information; near-infrared light signal extracting means for extracting near-infrared light image information of each pixel extracted by the in-contour pixel extracting means; Near-infrared light image information extracted by infrared light signal extraction means The second aspect of the present invention is a visual inspection device for agricultural products, comprising: 9. A transporting means for transporting agricultural products, a light source for irradiating the transported agricultural products with illumination light including near-infrared light, and each pixel in row a, column b of an image of the agricultural product illuminated by the light source. Light receiving means for extracting near-infrared light image information for each, and in-contour pixel extracting means for extracting pixels in the in-contour region of the agricultural product by comparing the near-infrared light image information of each pixel with a predetermined threshold value And a near-infrared light signal extracting means for extracting near-infrared light image information of each pixel extracted by the in-contour pixel extracting means, and near-infrared light image information extracted by the near-infrared light signal extracting means An appearance inspection apparatus for agricultural products, comprising: a determination unit for determining a scratch on an agricultural product. 10. The appearance inspection apparatus for agricultural products according to claim 6, wherein the light receiving unit has a single image pickup unit that simultaneously receives a near infrared light region and visible light. .