JP2006180862A - Seedling regulatory station - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a seedling regulatory station for enabling the automatization of a transplanting process of separating cultivated aseptic plants and transplanting them to rooting vessels in a place crowded with obstacles or in a narrow working space such as in vitro. <P>SOLUTION: The seedling regulatory station has a plurality of seedling regulatory systems. The system comprises a regulatory hand having a cutting blade and movable in the XYZ directions, a sensor unit capable of recognizing seedling shape, and a turnable separate chuck gripping a seedling with optimum grip force. The system works as follows: The shape of a seedling gripped by the separate chuck is measured by the sensor unit, a point of the seedling to be cut off is calculated based on the measurement data, and the unnecessary seedling point is cut off with the regulatory hand. For this station, different points are cut off in the respective regulatory stations. <P>COPYRIGHT: (C)2006,JPO&NCIPI

Description

  The present invention relates to a seedling preparation station that can mechanically adjust the lower leaves and the like of the buds when cultivating the buds in vitro, and in particular, in transplanting sterile clonal seedlings, dividing the cultured sterile seedlings, The present invention relates to a seedling adjustment station that makes it possible to automate operations such as adjusting lower leaves and transferring to rooting containers.

In addition to the production of seedlings and cuttings seedlings that have been used so far in the production of horticultural crops and tree seedlings in recent years, micropropagation allows mass propagation of sterilized clones of excellent individuals and rooting them into seedling products Has become widespread.
As a general technique, a growing tissue point is cut out from a plant individual in a sterile environment, provided with a culture environment in vitro (in a sterile transparent container), grown into multi-buds, and further proliferated in large quantities by subculture. After each grown bud was grown to an optimal size in vitro, the shoot was again separated in a sterile environment, transferred to a container for rooting, cultured again in vitro, and germinated clonal seedlings after completion of rooting. There is something to be a product. This is a technique that utilizes the fact that an infected seedling does not invade the growth point tissue that is the source of plant growth.

  As described above, the production process in the plant factory mainly includes a transplanting process for planting separated seedlings in vitro and a culture process in vitro. The seedling culture process is almost automated, but the transplanting process is still carried out by hand, and the need for mechanization has increased due to the need to secure a large number of excellent workers during the busy season. Yes.

  However, at the time of automation, in order to divide and transplant young shoots of an irregular shape, it is necessary to make a high-level and complicated recognition of which part of the stem is cut at which position. In addition, the young buds are soft and fragile, and when an excessive gripping force is applied, there is a problem that the stem duct is broken and the subsequent growth does not proceed smoothly. For this reason, advanced shape recognition and minute gripping force control are indispensable for the automatic transplantation of shoots.

  As a seedling division transplantation device, a device that recognizes the shape of a seedling by a recognition device using laser light and guides a gripping mechanism and a cutting mechanism to a predetermined height position of the seedling based on the recognition result is proposed. (Patent Documents 1 and 2).

In addition, the transplanted seedling transplanting system scans the stem shape of the seedlings on the open tray with a slit laser beam and PSD sensor, controls the gripping force by feedback with a strain sensor, and automatically produces the inserted seedlings with a manipulator. It has been proposed (Non-Patent Documents 1 and 2).
JP-A-3-228607 JP-A-5-3707 M. Takatsuji, Handbook of Plant Factory, Tokai University Press, pp. 123 {159, 1997. Tokai University Press: “Plant Factory Handbook”, Tokai University Press (1997), pp123-159 Takayama Kosaku CMC Publishing Seedling Production System (1992 First Edition 2002 Popular Edition pp180)

Most of the costs associated with growing seedlings are labor costs, and mechanization can greatly reduce seedling production costs. In addition, in the manual work, the yield decreases due to virus contamination, which also leads to an increase in production cost.
As described above, since the seedling culture process is almost automated, the present invention has an object to be solved by automating the in vitro transplantation process to the culture container. FIG. 1 is a flow chart of a clonal seedling growing process. The present invention aims to automate the steps 13 to 15 which are in vitro transplantation processes to a culture vessel.

  In the first place, the devices described above are devices that focus on the separation of buds, and it has been difficult to perform operations such as cutting the lower leaves of the buds and adjusting them to be suitable for in vitro culture.

  Moreover, in the above devices, since the opening and closing of the hand is directly controlled by a motor, it is difficult to make a weak adjustment of the force to grip the stem, and if used in the adjustment process, the stem may be damaged. was there.

Further, in the above devices, when there is an obstacle (leaf) in the vicinity of a gripping part such as a plant, the object cannot be gripped by avoiding the obstacle.
Furthermore, it was premised on work in an open tray and could not be operated in a work environment where seedlings in the container were dense. That is, when producing aseptic clone seedlings, since in vitro work is not possible, a clean room was required to achieve aseptic conditions. For this reason, there is a problem that a large work space and expensive equipment are required, and a large cost is required.

  Therefore, in view of the above problems, the present invention enables automation of a transplanting process in which a cultured aseptic plant is divided and transferred to a rooting container in a place where obstacles are concentrated or in a narrow work space such as in vitro. The aim is to provide a seedling adjustment station.

The gist of the present invention is the following seedling preparation apparatus (1) to (4).
(1) An adjustment hand having a cutting blade and movable in the XYZ directions, a sensor unit capable of recognizing the shape of the seedling, and a rotatable separate chuck for gripping the seedling with an optimum gripping force. An apparatus for adjusting a seedling that measures the shape of a seedling held by a separate chuck by a section, calculates a cutting location based on the measurement data, and cuts an unnecessary portion of the seedling with an adjusting hand.
(2) The seedling adjustment device according to (1), wherein the adjustment hand is rotatable about the engagement point of the cutting blade as a central axis.
(3) The seedling adjusting device according to (1) or (2), wherein the adjusting hand is movable in the XYZ directions by a three-axis linearly moving slider.
(4) The apparatus for adjusting a seedling according to any one of (1) to (3), wherein the calculation of the cut portion is performed based on measurement data at a plurality of positions obtained by rotating the separate chuck by a predetermined angle.

The gist of the present invention is the following seedling adjustment station (5) to (7).
(5) A seedling adjustment station including a plurality of seedling adjustment devices according to any one of (1) to (4), wherein each adjustment station cuts a different portion.
(6) The seedling adjustment station according to (5), wherein the one or more seedling adjustment devices have a grip portion in an adjustment hand.
(7) A cutting leaf removing device further comprising an adjustment brush having a cylindrical brush at the tip and movable in the XYZ directions, and a separate chuck for gripping the seedling with an optimum gripping force (5) or (6) Seedling adjustment station.

The gist of the present invention is the following (8) seedling automatic transplanting station.
(8) An automatic seedling transplanting station in which a general seedling separation station is provided in the preceding process of the seedling adjustment station in any one of (5) to (7) and a general seedling transplanting station is provided in the subsequent process .

According to the present invention, it is possible to automate the cutting operation of a lower leaf or the like necessary for obtaining a seedling suitable for in vitro culture, which has been conventionally performed manually.
In addition, the plant can be adjusted in various shapes to cut the lower leaves after performing shape recognition and the relative distance recognition of the cutting blade. For rosette-type plants (for example, spinach) Can also respond.

  In addition, work can be mechanized in a narrow work space such as a place where obstacles are densely packed or an in vitro environment. That is, since it is not necessary to provide a clean room and work on a clean bench is possible, the production cost of cloned seedlings can be reduced in terms of equipment.

The best mode for carrying out the present invention will be described with reference to the drawings.
1. System Configuration As shown in FIG. 2, the adjustment station 102 of the present invention is used with a separation station 101 disposed upstream thereof and an implantation station 103 disposed downstream thereof.
The separation station 101 delivers the shoots separated to the adjustment station 102, and the shoots adjusted by the adjustment station 102 are delivered to the implantation station 103, and are implanted in vitro at the implantation station 103.
Each component is connected to the integrated controller by a general-purpose communication method such as Ethernet (registered trademark) or RS-232C cable, and is controlled cooperatively.
Hereinafter, each component of the adjustment station 102 will be described.

The configuration of the adjustment station 102 includes a separate chuck 21, a three-dimensional sensor 22, and an adjustment hand 23 as main components (see FIG. 3).
(1) Separate chuck 21
The separate chuck 21 is a gripping chuck that can rotate clockwise and counterclockwise, and grips a stem portion of a seedling. The separate chuck 21 is rotatable about the seedling stalk, and the rotation of the separate chuck 21 can minimize the movement of the adjusting hand 23.

(2) Three-dimensional sensor 22
The three-dimensional sensor 22 includes a pair of stereo cameras and a slit projector, and can perform stereo imaging. In order to eliminate perspective distortion, it is preferable to use an optical system in the imaging optical system. That is, as shown in FIG. 4, while the two cameras are kept parallel to each other, the centers of the lenses 50a and 50b and the centers of the CCD image sensors 51a and 51b are shifted, and a pair of the CCD image sensor and the lens center are connected. By using a tilt optical system in which the intersection of the center lines of the imaging system is in the vicinity of the object to be measured, it is also possible to obtain an imaging image without perspective distortion on the CCD imaging device. The tilt optical system is an effective optical system for stereo image processing because the area that can be imaged in common by the left and right cameras is larger than that of a normal optical system. As a result, an accurate relative distance image can be obtained even for a target object located near the camera.
A slit projector that emits a laser diode light source in an oblique slit pattern is installed in the center of the apparatus. There are four slant slit patterns in total, and it is considered that the slit light hits no matter where the measurement target object is located in the imaging range.
In addition, it is preferable to arrange | position this light source for illumination, Preferably LED in the position where the light from an illumination light source reflects directly on a container in the peripheral part of a visual sensor, and does not enter into a camera.

Note that the sensor unit may be the above-described tilt optical system, but may be more versatile when relative distance measurement is unnecessary. For example, when separating buds planted in a lattice shape with a target position at the planting position, space recognition is performed according to the following procedure.
That is, the callus of seedlings planted in a transparent agar medium is binarized with an appropriate threshold from the image captured by a CCD camera placed under a transparent container, and image processing such as reduction / expansion is added to the image. The callus part of the seedling is extracted, the center of gravity of each callus is recognized, the pixel position of each seedling on the CCD image is calculated, and the XY coordinates of the callus positions of all the seedlings are obtained. Each seedling position is labeled according to a predetermined area. The order of labeling may be determined according to the movement path that the hand is unlikely to interfere with, and seedlings are taken out in ascending order of labeling.
When taking out the seedling, not only the stem portion but also the callus portion may be gripped. The taken out seedlings may be placed on a temporary storage place for seedlings outside the container so that the next adjustment process can be easily performed. If the adjustment station is in the next process, it may be transferred to a hand for delivery.

(3) Adjustment hand 23
The adjustment hand 23 includes a cutting blade 26, a motor 25 that opens and closes the cutting blade 26, and a plate 28 to which they are attached. The adjusting hand 23 is connected to the articulated robot by a plate 28 and is movable in the XYZ axis directions. Based on the measurement information of the three-dimensional sensor 22, the adjusting hand 23 performs cutting processing of the lower leaves of the seedlings while moving in the XYZ axis directions. Do. It is preferable to rotate the cutting blade 26 around the Y axis (referred to as the θ axis), and thereby the processing speed can be increased by minimizing the movement of the heavy adjustment hand 23 itself in the XYZ axes. .
Further, a finer adjustment operation can be performed by separately providing a gripping hand provided above or below the cutting blade of the adjusting hand and linking it with an adjusting hand without a gripping hand.
Note that the cutting blade is preferably made of a material that does not reflect when the shape of the plant is recognized, and a preferable material is ceramics.

2. Measurement of the spatial position and shape of a plant In the present invention, the shape and spatial position of a plant are measured by a combination of a light cutting method and a relative stereo method.
The light cutting method is a method of irradiating a measurement object with a planar laser beam (slit light) and measuring the distance using the principle of triangulation. The relative stereo method basically uses stereo vision with multiple cameras as a method of distance measurement, but it uses the distance information to the reference point in the screen obtained by a method other than stereo vision to create a stereo image. This is a method of measuring the relative height from the reference point by performing processing. High-speed image processing is possible, and there is an advantage that external parameters such as camera interval and mounting angle are not required.

When determining whether there is a stalk in the work space, identify the feature that the slit light reflects on the stalk to determine whether the stalk with the thickness of the set range is within the set three-dimensional space range. The determination is made by measuring the position by the light cutting method. When a stem is detected in the work space range, it is detected which position of the stem is suitable as a gripping position. Specifically, the branch point is determined by tracking from the light spot position of the stem, and a portion where no leaf exists at the upper part is set as a cutout point.
After the cut point is known, the relative distance between the cutting or gripping position and the hand can be calculated by moving the combined work hand to the vicinity by the relative stereo method. As a result, the remaining movement amount of the hand can be known.

3. Unnecessary part cutting process The cutting process of the lower leaf or the like is performed in the procedure shown in FIG. First, an image of the separated seedling is taken by the three-dimensional sensor 22 in order to determine the cut location of the separated seedling (STEP 21). Data obtained by imaging is subjected to image processing, and a portion to be cut (for example, a lower leaf in the case of eucalyptus and a root portion in the case of cymbidium) is recognized (STEP 23). At this time, although there is a part that cannot be determined as a cut part or a non-cut part depending on the imaged angle, it is treated as a gray point. The separated seedlings are imaged a plurality of times until a predetermined angle is reached (STEP 23). If the specified angle is not reached, the separated seedling is rotated (STEP 24), and imaging is performed from a different angle. As a preferable aspect, an imaging pattern with a constant accuracy and good working efficiency is exemplified by imaging three places in the range of 60 to 90 degrees. When the imaging at the designated angle is completed, the cutting location is calculated based on all the imaging data, and the cutting location is determined by performing superposition or the like (STEP 25). Cut unnecessary roots etc. (STEP26) and plant or deliver to planting station (STEP27). It should be noted that the leaves can be easily cut by making the swing angle of the scissors parallel to the inclination of the stem.

4). Recognition of leaf / stem boundary The processing of STEP22 will be described in detail.
To recognize the boundary between leaves and stems, the contour is extracted by applying contour tracking (chain code) to the binarized image data, and the edge of the leaf is searched to detect the end of the leaf. That's it. Recognition accuracy is enhanced by performing these operations on image data from a plurality of angles.

(1) Extraction of contour line The bottom leftmost point of the binarized processing target data is detected, and this point is set as a tracking start point P1. Next, as shown in FIG. 12A, the vicinity of P1 around P1 is detected clockwise as shown in FIG. 12A, and the first point P2 changed from the background pixel to the processing target pixel is detected as shown in FIG. 12B. . The pixel of interest is moved from P1 to P2, and P3 is obtained by performing a similar search for P2. This process is repeated until Pn = P1 or Pn is out of the effective search range. Also, the vector relationship between Pn and Pn-1 described by the code shown in FIG. 12C is called a chain code.
FIG. 13 shows the binarized image used in the above processing and the edge image obtained by the above processing.
(2) Leaf recognition algorithm A direction vector between certain pixels is obtained from the thinned edge image of the seedling, an algorithm for distinguishing leaves and stems from the inclination is created, and leaf cutting points are extracted. Here, it is assumed that the work is performed on the assumption that the seedling stalk extends to some degree in the vertical direction with respect to the x-axis.

B) Stem edge and search area Edge edge search is performed in the x-axis positive direction from the lower side of the image, and the edge coordinate value of the stem is set as the “stem initial position point” (see FIG. 14A). . The edge search is started from the obtained two stem initial position points. That is, the edge search is performed separately for the right edge and the left edge of the seedling. Further, the edge search is performed by determining a search area within a certain range. For example, the search area is 3 × 3 pixels.

B) The edge point adjacent to the initial stem position point obtained in the adjacent edge detection a) is detected, and the edge point search is sequentially repeated. At this time, the number of edge points in the search area is counted, and when the number is 4 or more, it is assumed that the edge is branched, and the center of the search area is stored as a branch point. Then, one edge point is searched. If the end of the edge does not reach an arbitrary height, it is determined that the edge is not an edge of the contour, and the process returns to the branch point to perform another edge point tracking. For example, as shown in FIG. 14B, the next edge point to be detected is an edge point that has been detected in the past in the search region centered on the i−1th edge point that is currently obtained, ie, i−1. Appears in pixels other than the i-th and i-2th edge point positions. Here, the i-th edge point is the detected edge point. FIG. 14C shows a case where there are a plurality of edge points in the search.

C) Direction vector calculation Obtain the edge direction vector of the stem. The direction vector at the i-th edge point is obtained as a difference between the i-th edge point coordinate value and the i-2th edge point coordinate value detected two times before (see FIG. 14D). Here, the inclination of the vector is an angle formed by the x axis and the vector. FIG. 14 (e) shows the direction vector.

D) When the inclination of the direction vector is within the range of 50 ° to 120 ° under the assumption that the leaf detection stem using the inclination of the direction vector is oriented to a certain degree of vertical with respect to the x-axis direction, It is assumed that it represents a direction vector, and if it has any other inclination, it is determined as a leaf direction vector. Here, since the leaf has a portion having an inclination of 50 ° to 120 °, the portion may be recognized as a stem (see “leaf candidate” in FIG. 14F). However, since the position of the leaf candidate is far away from the stem, the final point of the edge point recognized as the stem before the leaf candidate (point p in FIG. 14 (f)) and the starting point of the edge point of the leaf candidate ( The x-coordinate values of the point q) in FIG. 14 (f) are compared, and the edge point of the leaf candidate is recognized as the leaf edge point if it is other than a set threshold value. FIG. 14 (g) shows the cut points of the leaves and stems thus obtained.

E) Leaf detection The leaf end point is detected using the leaf region detected in step d), the detected region and the cut point of the leaf. Two cut points are output in one leaf region. The distance from the reference point to each edge point of the leaf is calculated from the reference point (point p in FIG. 14 (h)) that is large in the y direction, and then the edge point with the longest distance is the leaf end point (figure 14 (h) is output as point r).

F) Simulation results FIG. 14 shows the simulation results of the above image processing.
In FIG. 14, since a clear image is used as an input image, a preferable result is obtained. However, when binarized in an actual captured image, a blurred portion appears. Therefore, for image data obtained by imaging seedlings from multiple angles (for example, rotated every 120 °), pre-processing and leaf recognition programs are executed, and the accuracy of each image is improved by obtaining the border between stems and leaves. Yes. Furthermore, when measuring the shape data in the depth direction, the position of the leaf is recognized, and the relative distance measurement by tracing the stem position using the relative distance measurement is performed on each point of the outline of the stem on the stereo camera image. Measure the parallax. As a preferred method, there is a method described in Japanese Patent Application No. 2004-228743 proposed by the inventor. After integrating this data and the leaf cut-out point position, the cut-out coordinate position on the XYZ orthogonal axis corresponding to the index angle of the separate chuck is recognized.

  Hereinafter, details of the present invention will be described with reference to examples, but the present invention is not limited to the examples.

(Constitution)
As shown in FIGS. 6 and 7, the adjustment station of the present embodiment moves the adjustment hand 23 a and the three-axis linear motion slider 31 that allows the adjustment hand 23 a to move in the XYZ axis direction, and the adjustment hand 23 b and the XYZ axis direction. A triaxial linear motion slider 32 that can be moved, an adjustment brush 34, a triaxial linear motion slider 33 that can move the adjustment brush 34 in the XYZ axial directions, and separate chucks 21a to 21c are configured.

The adjustment hand 23a is disposed at a position opposite to the separate chuck 21a, and has a gripping hand 29 above the cutting blade 26 as shown in FIGS. 8 and 9, and the seedling is gripped by the gripping hand 29. The lower part of can be cut. The adjustment hand 23a can be rotated about the X axis (θ axis operation), and the processing speed is increased by minimizing the operation by the triaxial linear motion slider 31.
The adjusting hand 23b is a hand that is disposed at a position facing the separate chuck 21b and has only the cutting blade 26. The adjustment hand 23b can also be rotated about the X axis (θ axis operation), and the processing speed is increased by minimizing the operation by the triaxial linear motion slider 32.
The separate chucks 21a and 21b are L-shaped gripping chucks fixed on a rotary table that can rotate around a stem, and have a gripping width of about 3 mm and a gripping force of about 20 g. The L-shaped gripping chuck can be removed, and can be removed for high-temperature sterilization. The separate chucks 21a and 21b can grip a seedling stalk with an optimum gripping force by bending a pair of gripping claws.
The separate chuck 22c grips the upper part of the seedling when the lower leaf of the seedling cut by the adjustment brush 34 is excluded. In the present embodiment, the separate chuck 22c is configured separately, but the separate chuck 22b may also be used.
The three-dimensional sensors 22a and 22b recognize the shape of the seedling, calculate the posture and cutting point of the seedling, and calculate the amount of hand movement necessary for the adjustment work.
In addition, although the structure of a present Example presupposes adjusting the bud of Eucalyptus, it cannot be overemphasized that holding | grip width and holding | grip force are variable and it can apply also to another plant.

(Operation)
First, the seedling separated by the separation station is gripped by the separate chuck 21a. The shape of the stem of the seedling grasped by the three-dimensional sensor 22a is recognized, and a position of a predetermined length is cut from the tip while the stem is grasped by the adjusting hand 23a. At this time, if there is a lower leaf or the like at the cutting position, the separate chuck 21a is rotated as necessary.
The adjusting hand 23a holding the seedling of a predetermined length moves horizontally by the triaxial linearly moving slider 31, and delivers the seedling to the separate chuck 21b. The seedling whose shape has been recognized by the three-dimensional sensor 22b is subjected to cutting processing such as lower leaves by the adjusting hand 23b while rotating the separate chuck 21b.
The seedlings that have been adjusted, such as the lower leaves, are handed over in a form perpendicular to the implantation hand, and the adjustment is completed. At this time, if the final height of the seedling is grasped by the transplanting hand and then cut to a specified length, it can be expected that a new cut end is formed and rooting is improved. The seedling that has been cut off the lower leaves and the like is moved to the separate chuck 21c by the adjusting hand 23a, and when the lower leaves and the like are completely removed by the adjusting brush 34, it is again gripped by the adjusting hand 23a and delivered to the implantation station.

  In plant varieties with weak joints between leaves and stems, the upper part of the seedling where the lower leaves remain is gripped by the separate chuck 21c, the specified position is sandwiched between the left and right brushes of the adjustment brush 34 (see FIG. 16), Productivity can be improved if it is configured to drop leaves by moving. At this time, even after the processing by the adjustment brush 34, the seedlings may be gripped with the left and right brushes as they are, and the seedlings may be delivered to the separate chuck 21b, and the shape inspection may be performed by the sensor unit 22b and received by the implantation hand. . For opening and closing the brush, an actuator similar to that used for opening and closing the hand can be used.

As shown in FIG. 10, the seedling transplantation assisting station of the present invention includes a rotary table 41, a three-dimensional sensor 22, and a robot hand unit as main components. Each component is connected to the integrated controller 6 and controlled cooperatively by a general-purpose communication method such as Ethernet (registered trademark) or RS-232C cable.
In the configuration shown in FIG. 10, a series of operations can be performed from separation to planting. However, the separation station and the implantation station separately provided may be linked or a part thereof may be manually performed.
Below, each component of the seedling transplantation assistance station will be described.

(1) Rotary table 41
The rotary table 41 is a circular rotary table that can rotate clockwise and counterclockwise, and includes a large rotary table and a small rotary table. On the rotary table 41, an implantation container, a separation container, a sterilization container, and a cutting waste petri dish are placed. In each operation stage, the large rotary table is rotated to adjust the position of each container. Each container can be individually rotated by a small rotary table provided on the rotary table 41. By rotating each container individually in each work stage, the work range can be expanded and work accuracy can be improved. This is possible (see FIG. 11).

(2) Three-dimensional sensor 22
The three-dimensional sensor 22 includes a pair of stereo cameras and a slit projector, and can perform stereo imaging. In order to eliminate the perspective distortion, the imaging optical system uses an optical system, which is the same as the configuration shown in FIG.

(3) Robot Hand Unit The robot hand unit includes an adjustment hand 23, a separation hand 42, and an implantation hand 43. The robot hand unit does not necessarily include all of these, and some hands may be configured separately.
When the buds are separated by the separation hand 42 provided at the tip of the multi-joint robot 1, the adjustment hand 23 moves to a position opposite to the separation hand 42 and performs an unnecessary lower leaf cutting process. The transplanting hand 43 holds the young buds after the adjustment of the lower leaves and the like so as not to damage the conduit of the stem and performs the implantation.

It is a flowchart of the growth process of a clone seedling. It is a top view of the system (separation station, adjustment station, implantation station) for transplanting a clone seedling. It is a system block diagram of an adjustment station. It is explanatory drawing of a tilt optical system. It is a flowchart of the unnecessary part cutting operation | work by the adjustment station of this invention. FIG. 3 is a configuration perspective view of an adjustment station according to the first embodiment. FIG. 3 is a configuration plan view of an adjustment station according to the first embodiment. 3 is a configuration plan view of an adjustment hand with a gripping hand according to Embodiment 1. FIG. It is a structure side view of the adjustment hand with a holding hand of Example 1. It is a structure top view of the adjustment station of Example 2. FIG. It is explanatory drawing of a rotary table. It is explanatory drawing of an edge image creation procedure. It is the output of imaging binarized data and an edge image. It is explanatory drawing of a leaf recognition algorithm. It is an output result of leaf recognition simulation. It is explanatory drawing of the removal operation | work of the lower leaf by an adjustment brush.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Articulated robot 2 Robot controller 3 Sensor controller 6 Integrated controller 21 Separate chuck 22 Three-dimensional sensor 23 Adjustment hand 25 Cutting blade drive motor 26 Cutting blade 27 Chute 28 Plate 29 Gripping hands 31, 32, 33 Linear motion slider 34 Adjustment brush 41 Rotary Table 42 Separation Hand 43 Implantation Hand 50 Lens 51 CCD Image Sensor 91, 95 Carry-in Conveyor 92, 96 Separate Chuck 93, 97 Carry-out Conveyor 101 Separation Station 102 Adjustment Station 103 Implantation Station

Claims (8)

An adjustment hand having a cutting blade and movable in the XYZ directions, a sensor unit capable of recognizing the shape of the seedling, and a rotatable separate chuck for gripping the seedling with an optimum gripping force,
An apparatus for adjusting a seedling that measures the shape of a seedling held by a separate chuck by a sensor unit, calculates a cutting location based on the measurement data, and cuts an unnecessary portion of the seedling by an adjusting hand.   The seedling adjustment device according to claim 1, wherein the adjustment hand is rotatable about a meshing point of the cutting blade as a central axis.   The seedling adjustment device according to claim 1 or 2, wherein the adjustment hand is movable in the XYZ directions by a three-axis linear motion slider.   4. The seedling adjustment device according to claim 1, wherein the cutting location is calculated based on measurement data at a plurality of positions obtained by rotating the separate chuck by a predetermined angle. A plurality of seedling adjusting devices according to any one of claims 1 to 4,
A seedling adjustment station that cuts different points at each adjustment station.   6. The seedling adjustment station according to claim 5, wherein the one or more seedling adjustment devices have a grip portion in an adjustment hand.   The seedling according to claim 5 or 6, further comprising a cutting leaf removing device having a cylindrical brush at the tip, and an adjustment brush movable in XYZ directions, and a separate chuck for gripping the seedling with an optimum gripping force. Adjustment station. An automatic seedling transplanting station in which a general seedling separation station is arranged in a pre-process of the seedling adjustment station according to claim 5 and a general seedling transplanting station is arranged in a post-process.