JPH1076233A - Apparatus for detecting calyx direction of farm produce - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce labor by automatically detecting the directions of calyxes of egg apples before sorting measurement by an automatic egg apple sorting system to dispense with arranging the directions of calyxes of egg apples to be sorted placed on the buckets of a feed conveyor. SOLUTION: Prior to measuring the shapes of egg apples for automatic sorting, a brightness image F2 and a hue image F3 are formed from the RGB camera image F1 of each of the egg apples and, further, three images, that is, a multivalue image F5 obtained by the differential processing of the brightness image F2, a shining removed binarized image F4 obtained by binarizing the brightness image F2 on the basis of a predetermined brightness threshold value in order to remove shining pixels and a yellow injury removed binarized image F6 obtained by binarizing the hue image F3 on the basis of a predetermined threshold value of a yellow hue in order to remove pixels of a yellow injury are subjected to AND processing to form a multivalent image F7. The respective sum totals PDL, PDR on the left and right sides of the almost central point (X coordinate xm) of the length in the axial direction of the egg apple related to projection arrangement data PD wherein the multivalue image F7 is projected in the Y-axis direction crossing the axis of the egg apple almost at a right angle are compared and a calyx is judged to be present on the side of the larger sum total.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to the measurement and quality (e.g., etc.) of an agricultural product such as an eggplant, which has a substantially axially symmetrical shape and is provided with an end on one end of the shaft, for example, an eggplant. Prior to this measurement and quality determination, the automatic fruit sorting system that performs the determination of the classification of the agricultural products placed on the conveyor bucket of the conveyor so that the axial direction thereof is oriented in a predetermined direction. The present invention relates to an apparatus for detecting the direction of the agricultural product, which detects which side of the predetermined direction is in the predetermined direction.

[0002] In the drawings, the same reference numerals indicate the same or corresponding parts.

[0003]

2. Description of the Related Art Conventionally, in an automatic eggplant sorting system, an operator places eggplants on a bucket of a conveyor for transporting the eggplants, the axes of the eggplants are oriented in a substantially constant direction, and the eggplant gadgets have a fixed left and right side. (That is, the direction of the gaku is aligned).

[0004]

However, as described above, the work of the operator placing eggplants on the conveyor bucket in the same direction as that of the conveyor is naturally time-consuming. Accordingly, the present invention has been made to solve the above-described problem, and to provide an apparatus for detecting the gaku direction of agricultural products, which enables automatic fruit sorting without aligning the left and right directions of the gaku of agricultural products to be sorted. I do.

[0005]

According to a first aspect of the present invention, there is provided an apparatus for detecting the direction of an agricultural product, which has a substantially axially symmetric outer shape, and has an axial portion (01a) at one end of the shaft. An image of the agricultural product (01) is taken by a monochrome or color imaging camera (1), and a gak direction detection device for detecting the gak direction of the agricultural product from the image. (F2), a differential image (differentiated multi-valued image F5) is generated from the brightness image, and the projection array data (PD) of the differential image in a direction substantially orthogonal to the axis of the agricultural product. And the sum of the projection array data (left projection array sum PD _L , right projection array sum PD _R ) for each of both sides of a point (X coordinate xm) substantially at the center of the axial length of the agricultural product. Compare It is determined that the larger side of the sum is calyx.

According to a second aspect of the present invention, in the first aspect, the brightness image is binarized with a predetermined threshold value, and pixels having a brightness value equal to or higher than the threshold value are obtained. From the image (AND-processed multi-valued image F7) obtained by the logical product of the binarized image with illuminated removal and the differential image. Ask for.

According to a third aspect of the present invention, in the first aspect, the camera is a color imaging camera, and an image of the hue is further obtained from an image captured by the camera. (F3) is created, this hue image is binarized with a predetermined threshold of a predetermined hue, and a predetermined color flaw removal binarized image (yellow yellow) for removing pixels of this predetermined hue above this threshold Flaw removal binarized image F6 etc.)
Is generated, and a logical AND operation with the predetermined color flaw removal binarized image is performed in advance on the image of the projection target from which the projection array data is obtained.

According to a fourth aspect of the present invention, there is provided the gaku direction detection device according to any one of the first to third aspects, wherein the produce is eggplant. According to a fifth aspect of the present invention, the predetermined direction for binarizing the hue image is set to yellow.

[0009]

FIG. 2 is a block diagram of the hardware of an automatic fruit sorting system as one embodiment of the present invention. In the figure, 01 is an agricultural product to be sorted (eggplant in this example),
Reference numeral 6 denotes a main conveyor for transporting the agricultural products 01 to the sorting position via the bucket 6a. Reference numerals 7 (71, 72, 73) denote the main conveyor 6 from the main conveyor 6 based on the determination results of the grades A, B, and C based on the shape measurement and the like of the agricultural products 01, respectively. This is a discharge conveyor that conveys and discharges the agricultural products 01 dropped to a position corresponding to the class.

In this example, the bucket 6a is set in advance so that the axis of the agricultural product 01 is substantially perpendicular to the plane of the drawing (in other words, the axis of the agricultural product 01 is substantially orthogonal to the conveying direction of the main conveyor 6). It is assumed that it is carried on the upper part. Next, 1 is a color imaging TV camera for imaging the agricultural product 01 conveyed by the main conveyor 6 to a predetermined position near the entrance of the automatic fruit sorting system, and 2 is the imaging from the TV camera 1. An RGB video signal obtained by raster-scanning an image is input and image-processed,
As described later, an image processing apparatus that detects the direction of the gak of the agricultural product 01 in advance, measures the shape and the like of the agricultural product 01 using the detection result, and determines the class of the agricultural product 01, and 3 determines the image processing apparatus 2 An emission control device that performs emission control of the agricultural product 01 based on the result 5.

That is, when the agricultural product 01 is conveyed in the direction of the arrow via the main conveyor 6 and the front end of the bucket 6a reaches the point P, a judgment start signal 4 is inputted to the image processing device 2 via a position detecting means (not shown). Is done. Thus, the image processing apparatus 2 takes in the video signal from the TV camera 1 to detect the direction of the gaku of the agricultural product 01 and determine the grade, and outputs the determination result 5. The discharge control device 3 drops the agricultural product 01 from the main conveyor 6 to the corresponding discharge conveyor 7 (that is, any one of 71 to 73) based on the determination result 5, and then places the packing position (not shown) via the discharge conveyor 7. Transport to

FIG. 1 is an explanatory diagram of a process of detecting the direction of the gak of the agricultural product 01 by the image processing apparatus 2, and FIG. 1A shows a flow of the process. Referring to FIG. 1 with reference to FIG. 2, the image processing apparatus 2
Is the RG for the agricultural product 01 from the TV camera 1
The multi-value lightness image F2 and the multi-value hue image F3 are obtained by color difference conversion of the camera image F1 composed of video signals for each color B.
Create

Next, the brightness image F2 is illuminated (light:
A slang term that refers to the appearance of a bright glow due to the reflection of light.)
A binarization process of removing a portion, that is, a process of setting a pixel having brightness equal to or higher than a predetermined threshold value to “0” and setting other pixels to “1”, to obtain an illuminated binary image F4. At the same time, the brightness image F2 is similarly subjected to a differentiation process (in this example, a process of obtaining a difference between adjacent pixels in each of the X-axis and Y-axis directions) to create a differential-processed multivalued image F5.

The image processing apparatus 2 performs a binarization process for removing the yellow flaw from the hue image F3, that is, sets only pixels having a density of a yellow hue equal to or higher than a predetermined threshold value to “0”. A process for setting the other pixels to "1" is performed to create a yellow-stain removal binarized image F6. Next, the image processing device 2
Are the illuminated binary image F4 and the differentially processed multi-valued image F
5. An AND-processed multi-valued image F7 composed of the logical product of the three images of the binarized image F6 including the yellow scratches is created.

In other words, the AND-processed multi-valued image F7 is a multi-valued image mainly composed of the outer contour portion and the black portion of the agricultural product 01, except for the portion that looks shining and the portion that looks yellow. Next, from the AND-processed multivalued image F7, the projection array data PD in the Y-axis direction (that is, data in which the sum of the pixel values of each pixel on the same X coordinate is arranged in the order of the X coordinate value) PD is obtained. In this example, the axis of the agricultural product (eggplant) 01 is substantially in the X-axis direction as shown in FIG. 1B, and the entire image of the agricultural product (eggplant) 01 is displayed on the screen of the image processing apparatus 2. In order to fit the window area for measurement processing set in advance appropriately (that is, the length of the blank portion of the window area is about a fraction of the total length of the agricultural product (eggplant) 01 in the axial direction). It is assumed that it is placed on the bucket 6a.

Next, based on the projection array data PD,
From the X coordinate xs of the start point of the window to the middle point of the window
Of the projection array data PD up to the X coordinate xm
Left projection array sum PD_LAnd the X coordinate xm of this midpoint
To the X coordinate xe of the end point of the window
Right projection array sum PD as sum of array data PD _R
Is calculated, and this array sum PD_LAnd PD_RAnd the sum
The side with the larger is the direction of gaku.

In this example, it is detected that the right-side projected array sum PD _R is larger than the left-side projected array sum PD _L , and that the gaku 01a of the agricultural product (eggplant) 01 is on the right side. The image processing apparatus 2 measures the shape of the agricultural product (eggplant) 01 (for example, measures the total length in the axial direction, the ratio between the thickness of the neck and the trunk, the degree of bending, and the like) from the direction of the gad 01a as described above. A determination result 5 is output, whereby the sorting of the agricultural products 01 is performed.

Instead of using the middle point of the window as the middle point, the projection array data PD is binarized with a predetermined threshold value to obtain the X-coordinate of the axial start and end points of the outer shape of the produce 01. The X coordinate of the middle point of the actual agricultural product 01 as the middle point between the start point and the end point may be used. In this case, it goes without saying that the detection accuracy of the jerk direction is further improved.

[0019]

According to the present invention, there is provided an automatic fruit sorting system having an approximately axisymmetric outer shape such as an eggplant and having a gaku at one end of the shaft for performing equal-grade sorting of agricultural products. Prior to shape measurement and class determination, a differential image obtained by differentiating a brightness image obtained from a camera image of the agricultural product, or a brightness image of the differential image having a predetermined brightness in order to remove pixels that appear to shine. An illuminated binarized image binarized by a threshold value and / or a hue image obtained from a camera image to remove a pixel of a predetermined color (in this case, yellow) is converted into a predetermined hue of a predetermined hue. An image obtained by performing a logical AND process with a predetermined color flaw removal binarized image binarized by a threshold value (that is, a differential image obtained by removing a luminous part and a predetermined color flaw part of the agricultural product); Almost on the axis of produce Projected to direction orthogonal obtains both sides of the sum of projection data for each of the substantially central point of the axial length of the agricultural products (X coordinate xm) a (PD _L and PD _R),
Since it is determined that there is gaku on the side with the larger sum, it is possible to accurately detect the gaku orientation without having to align the gaku orientation when placing the agricultural products on the conveyor bucket that transports the agricultural products to be sorted. Shape measurement can be performed stably, and labor for handling agricultural products can be reduced.

[Brief description of the drawings]

FIG. 1 is an explanatory diagram of a back direction detection process as an embodiment of the present invention.

FIG. 2 is a configuration diagram of hardware of an automatic fruit sorting system as one embodiment of the present invention.

[Explanation of symbols]

01 agricultural products (eggplant) 01a gaku 1 TV camera 2 image processing device 3 discharge control device 4 judgment start signal 5 judgment result 6 main conveyor 6a bucket 7 (71-73) discharge conveyor F1 RGB camera image F2 brightness image F3 hue image F4 illumination Binary removal binarized image F5 Differential processing multi-valued image F6 Yellow scratch removal binarized image F7 AND processed multi-valued image PD projection array data PD _L left projection array sum PD _R right projection array sum xs Start point of measurement processing window X coordinate xm X coordinate of middle point of measurement processing window xe X coordinate of end point of measurement processing window

Claims (5)

[Claims] An image of an agricultural product having a substantially axially symmetric outer shape and having gaku at one end of this axis is taken by a monochrome or color imaging camera, and the gaku direction of the agricultural product is detected from the image. In the direction detection device, a brightness image is created from the captured image of the camera, a differential image is created from the brightness image, and projection array data of the differential image in a direction substantially orthogonal to the axis of the agricultural product is provided. Calculating the sum of the projection array data for each of the two sides of the point substantially at the center of the axial length of the agricultural product, and determining that there is a gaku on the side of the larger sum. Detection device for agricultural products. 2. The illumination direction detecting apparatus according to claim 1, wherein said brightness image is binarized with a predetermined threshold value, and pixels having brightness equal to or higher than said threshold value are removed. An apparatus for detecting a gaku direction of agricultural products, wherein an image is created, and the projection array data is obtained from an image obtained by a logical product of the binarized image with illuminated removal and the differential image. 3. An apparatus according to claim 1, wherein the camera is a camera for color imaging, and an image of the hue is further created from an image taken by the camera, and the hue image is determined by a predetermined method. A predetermined threshold value of the hue of the image, and a predetermined color defect removal binarized image for removing pixels of the predetermined hue equal to or more than the threshold value is created. A gaku direction detection device for agricultural products, wherein the image of (a) is subjected to a logical product process with the predetermined color flaw removal binarized image in advance. 4. The gaku direction detection device for agricultural products according to claim 1, wherein the agricultural product is an eggplant. 5. The gaku direction detection device for agricultural products according to claim 4, wherein a predetermined hue for binarizing the hue image is yellow.