JP4002387B2 - Agricultural product grade judging method and grade judging device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
In consideration of the fact that micro-scratches derived from sunspot disease known in agricultural products such as citrus fruits have an impact on the evaluation of agricultural products in the market, the agricultural products having such micro-scratches are selected and sorted in a fruit selection field. It relates to technology that can be reflected in work.
[0002]
[Prior art]
The value of agricultural products is usually determined by the amount of demand and evaluation in the trading market, and the evaluation results are reflected in the labor value (consideration) of producers, which in turn leads to increased work motivation. In order to reflect the results in the trading market as much as possible in consideration for the producers, the items of selection / sorting in the selection / sorting workshop in the selection area close to the production site are selected.
[0003]
Such selection items generally include class elements belonging to geometric categories such as size and weight of agricultural products, and emotional grade elements such as taste such as sugar content and satisfaction given by visual elements. It is widely known that there is a device that can detect the above-mentioned class elements and grade elements as accurately and accurately as possible in the process of selecting, judging and sorting agricultural products, Various methods and devices that have been devised to quickly perform the work process have been proposed in the past. For example, in the conventional sorting and sorting of agricultural products, the presence or extent (size) of the surface of the fruit is considered. Have also been proposed (Japanese Patent Laid-Open No. 3-226658).
[0004]
[Problems to be solved by the invention]
However, unlike industrial products that require dimensional accuracy in consideration of tolerances, agricultural products have the unique problem that their size and maturity are inevitably different.
[0005]
For example, in the case of citrus fruits as an example, as shown in FIG. 1, it is known that there is a disease called sunspot disease in which many small black spots appear as small scratches 2 on the fruit skin. Citrus 1 etc., which is dotted with many black spots caused by disease (think of this as “scratches” as described above), has a low reputation in the trading market, but conventionally, this is a concrete countermeasure for this problem. No method has been proposed, and the problem of considering grade evaluation for such sunspot problems has not been recognized.
[0006]
In other words, in conventional wound screening, grade evaluation is performed using the total area (or ratio) of detected wounds as a guide, so large wounds called general wounds, normal wounds 3 etc., and black spot disease It is not particularly distinguished from small scratches 2 where the area of each scratch is small, and even if there are a large number of small scratches 2, the total area does not become very large, so it should be classified into a lower grade There was no. Therefore, in the mandarin orange 1 in which one large wound such as the ordinary wound 3 exists, and in the mandarin orange 1 that does not have the ordinary wound 3 but has a large number of micro-scratches 2 of the sunspot disease, the total area of sun spots of the latter mandarin orange is If it is small, it will be sorted into a higher grade than the former, which may lead to results that are not consistent with the market ratings.
[0007]
As described above, the present invention is considered as one of the elements of sorting and sorting of agricultural products, and the evaluation in the trading market is considered as one of the elements of sorting and sorting agricultural products. It was made for the purpose of being able to accurately reflect the consideration to the producer.
[0008]
[Means for Solving the Problems]
In order to achieve the above object, the features of the present invention are as follows.
[0009]
(1) A plurality of scratch images existing in an image of agricultural products are individually detected, and the area of each scratch image is compared with a predetermined micro-scratch area threshold, and the area of the micro-scratch and the area larger than that is determined. A method for determining the grade of agricultural products, characterized by discriminating large scratches, and using the total total area and / or total number of micro-scratch groups determined to be micro-scratches as an element of grade determination for sorting and sorting the agricultural products. .
[0010]
In the above, the method for determining whether a single scratch is a micro-scratch, or a normal scratch other than that, or a general scratch is not limited, but, for example, an image obtained by capturing a mandarin orange as a dot matrix The saturation (or hue) of each pixel is binarized to first determine the difference between the skin portion and the scratched portion of a normal mandarin orange, and then the continuous size (width) of the pixels determined to be scratched Is compared with a predetermined threshold value to determine whether it is a minute wound, a normal wound, or a general wound larger than that. Whether or not it is a microscratch derived from a sunspot disease depends on the resolution of the captured image, but according to the experience of the inventor of the present application, for example, about 2 mm 2 / piece can be set as the upper limit guide for the sunspot.
[0011]
When determining the grade of the agricultural product based on the group of micro-scratches determined to be micro-scratches, at least one of the total total area or the total number of micro-scratch groups can be used. Whether to use the total area or the total number is unclear depending on the target agricultural products, etc., but even if the number is the same, for example, when there are small scratches of 1 mm ² or less, it is close to 2 mm ² In the case where minute flaws are scattered, it is often preferable to use the total area because it is inferior in appearance to those having flaws close to 2 mm ² . Both of these can also be used for evaluation. In this case, it is preferable to adopt a grade determination result that is shown to have a lower evaluation.
[0012]
Although the present invention specifies that a micro-flaw is used as one of the elements for grade determination, other grade determination elements, for example, other elements such as color, internal quality, irregular shape, etc. are used for the grade determination. Of course, it is not excluded that they are used together, and it is natural that class elements such as size and weight are not excluded from the combined sorting of agricultural products.
[0013]
The present application is characterized by the following invention in addition to the above invention (1).
[0014]
(2) The above-mentioned (1) characterized in that the area threshold value for discriminating the wounds included in the agricultural product picked-up image of (1) is set so as to include the black spot disease-derived wounds in the micro wound group. Agricultural product grade judgment method.
[0015]
(3) When the total total area and / or the total number of micro-scratches is greater than a predetermined grade rating reference value, the grade of the agricultural product is determined to be downgraded ((3) 2) Agricultural product grade determination method (4) The grade determination method described in any one of (1) to (3) above is used for grade determination of citrus fruits such as oranges, apples or peas. Judgment method.
[0016]
The present application is characterized by the following agricultural product grade determination device, apart from the above-described inventions of the respective grade determination methods.
[0017]
(5) Agricultural product imaging means, and a plurality of scratch images that are partially divided in the captured image are individually detected, and each detected scratch image is compared with a predetermined area threshold and a micro-scratch First comparing means for discriminating the above-mentioned large scratches, and second comparing means for comparing the total total area and / or total number of the discriminated micro-flaw groups with a predetermined rating rating reference value; An agricultural product grade judging device comprising grade rating means for judging the grade rating of the agricultural product in accordance with the output of the second comparing means.
[0018]
(6) The agricultural product first determination means of the invention (5) is characterized in that the area threshold is set so that the wound caused by sunspot disease is included in the group of minute wounds.
[0019]
DETAILED DESCRIPTION OF THE INVENTION
Embodiment 1
The present example shown in FIG. 2 to FIG. 6 does not have the mandarin orange 1i having either the normal wound 3 which is the large wound shown in FIG. 1 or the micro-scratch 2 derived from sunspot disease, the mandarin orange 1j having both wounds. This is to explain how these are graded considering the oranges 1k. In the present example, the coloration degree of the mandarin orange is detected together with the presence or absence of these scratches, and is shown as the case of the grade determination.
[0020]
In FIG. 2 showing this example, the appearance of the mandarin orange 1 conveyed on the sorting conveyor 10 for sorting and sorting is imaged by the CCD camera 11 at a predetermined imaging position. As this imaging method and means (illumination system 12, mirror 13, etc.), known ones can be used, and if necessary, only a plane image, three surfaces including a side image, five surfaces, or even a bottom surface are included. Although multi-plane imaging is performed, the area of the imaging plane is not directly related to the present invention.
[0021]
In this example, the captured image captured by the CCD camera 11 is input to a computer image processing system (not shown) as image information including a set of pixels of a dot matrix.
[0022]
In this image processing system, first, based on the image input, each pixel signal of the image information is binarized in step 201 (hereinafter referred to as “S”; the same applies hereinafter) 201 in order to distinguish the background from the agricultural products. In step S202, the outline of the agricultural product is determined and the processing area is determined in step S203. The size of the agricultural product is determined in step S204 based on the external shape determination information, and the class is determined in step S205. Further, the degree of deformation is determined in S206.
[0023]
On the other hand, based on the image input, second, color conversion of each pixel signal of the image information is performed in S210, scratch detection is performed in S211, and a scratch in the processing region determined in S203 is detected ( S212: Injury detection). Further, a color histogram is taken using the information after color conversion of each pixel signal of the image information (S213), and the color level is detected (S214).
[0024]
In this example, the grade is determined from the degree of deformation of the mandarin orange (information in S206), the injury (information in S212), and the degree of coloring (information in S214) detected as described above (S216: Grade). Judgment).
[0025]
Then, based on the results of the class determination (S205) and the class determination (S216), the final sorting classification is determined and a sorting output (sorting sorting) signal is output.
[0026]
The image processing of the picked-up image for sorting and sorting as described above and the standard for discrimination of equal classes have been set in various ways. In this example, characteristic processing is performed in S211 for detecting a scratch.
[0027]
That is, the normal epidermis color of the mandarin orange and the non-colored portion of the mandarin orange are detected from the image information subjected to color conversion in S210, and the latter is recognized as a flaw. Further, as shown in FIG. 3, the detected / recognized flaw is a minute flaw if the continuous portion of the portion recognized as a flaw in this example is in the range of 0 to 2 due to the size of continuous pixels. The image processing system of the computer is set so as to determine that it is a normal scratch if it is 3 or more. Then, the scratches included in the micro-scratches and the scratches included in the normal scratches are counted (added) as shown in FIG.
[0028]
Then, as described above, when each of the number (or area) of normal scratches, the number (or area) of fine scratches, and the degree of coloring is used as an element for grade determination as described above, for example, as shown in FIG. , Grade category 1, grade category 2, ..... If the grade judgment standard according to each grade category called grade category n is determined according to each grade category, which of these is applicable? By comparing the above, it is possible to determine the corresponding grade of each tangerine. FIG. 6 shows an example of the determination procedure for such grade determination.
[0029]
According to the grade determination method performed as described above, the fact that the injury caused by sunspot disease has been evaluated as a grade downgrade in the trading market can be reflected in the sorting stage in the selection area. It is possible to bring about an excellent effect that problems that have not been made can be reflected in the evaluation of the producer.
[0030]
【The invention's effect】
According to the invention of the present application, the sunspot that appears in the agricultural product epidermis derived from sunspot disease, which has not been shown even the recognition of the problem in the past, effectively solved the problem that the grade evaluation was downgraded in the trading market. Thus, there is an advantage that it can be reflected in a more accurate evaluation for the producer.
[Brief description of the drawings]
FIG. 1 is a diagram for explaining the appearance of a mandarin orange in which a black spot of a black spot disease that is a subject of the present invention appears on the epidermis.
FIG. 2 is a diagram showing an overview of a system when the method of the present invention is applied.
FIG. 3 is a diagram for explaining an example for discriminating minute scratches.
FIG. 4 is a diagram for explaining how to handle depending on the size of a wound;
FIG. 5 is a diagram showing an example of grade classification.
FIG. 6 is a diagram showing a cascading determination procedure when determining a classification.

Claims (6)

A plurality of scratch images present in images of agricultural products are individually detected, and the area of each scratch image is compared with a predetermined micro-scratch area threshold, and the micro-scratch and a scratch having a larger area than that are And a total total area and / or total number of micro-scratch groups determined to be micro-scratches is used as a class determination element for sorting and sorting the agricultural products. The method according to claim 1, wherein an area threshold value for discriminating micro-scratches included in the farm product captured image is set so as to include black spot-derived scratches in the micro-scratch group. . The grade of the said agricultural product is judged to be downgraded when the total total area of the said micro wound group and / or the total number of micro scratches are larger than the predetermined grade rating reference value. Grade judgment method. The grade judging method according to any one of claims 1 to 3 is used for judging grades of citrus fruits such as mandarin oranges, apples or peas. Agricultural product imaging means and a plurality of wound images that are partially divided and present in the captured image are individually detected, and each detected scratch image is compared with a predetermined area threshold and a small scratch and an area larger than that First comparing means for discriminating a large flaw, and second comparing means for comparing the total total area and / or total number of the discriminated micro-flaw group with a predetermined grading reference value, An agricultural product grade judging device comprising grade ranking means for judging the grade of the agricultural product according to the output of the comparison means. 6. The agricultural product grade determination apparatus according to claim 5, wherein the agricultural product first comparing means has an area threshold set so as to include a wound caused by sunspot disease in a small wound group.

Images