JP2015146748A - Electric working vehicle - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an electric working vehicle which can remove grass between rows of crops by autonomous travel and can be produced at a low cost.SOLUTION: An electric working vehicle 1 includes a left-and-right pair of travel units 20, a left-and-right pair of travel motors 26 for driving the travel units 20 respectively, a weeder 30 for removing grass, and a control unit 70 in which an autonomous travel program for autonomous travel is stored and can remove grass by autonomous travel. The electric working vehicle 1 also includes an imaging device 50 capable of capturing images of the surroundings of the electric working vehicle 1. The control unit 70 identifies crops on the basis of color information in images captured by the imaging device 50 and travels autonomously between rows of crops and removes grass.

Description

  The present invention includes a pair of left and right traveling devices, a pair of left and right traveling motors that drive the respective traveling devices, a weeding device for weeding, and a control unit that stores an autonomous traveling program for autonomous traveling. And an electric work vehicle capable of weeding by autonomous traveling.

  As paddy field weeding methods, weeding with herbicides, weeding by workers, weeding with hand-held or riding-type weeding machines, weeding with animals such as Aigamo farming methods, and the like. In recent years, there has been a demand for herbicides that do not use herbicides for reasons such as reducing environmental burdens and supplying safe food.

  However, weeding by workers is heavy labor and the burden on workers is very heavy. In addition, weeding with a weeding machine is costly to purchase an expensive weeding machine. Further, since the worker performs weeding while operating the weeding machine, the worker is restrained by the weeding work, so the work efficiency is poor. Weeding with animals can cause unevenness in weeding, and increases the cost of breeding and work related to breeding.

  In order to solve the above problems, paddy field weeding robots have been proposed that perform weeding by autonomous running. For example, a main body having a space through which rice plants can pass, a pair or a plurality of wheels provided at the lower part of the main body, a plurality of driving units for driving the wheels, and a lower part of the main body. One or more pairs of floats, detection means provided at the front of the main body for detecting rice stock, and control of wheel rotation via the drive unit so that the rice stock passes through the center of the space Some have a control device (Patent Document 1). Here, in patent document 1, a rice strain is detected with a contact or non-contact sensor. In addition, there is one that detects the detection of rice stock from an image captured by a camera (Non-Patent Document 1).

JP 2011-120573 A

Katsuhiko Tabata and 4 others, “Development of small weeding robot for paddy field (Aigamo robot)”, Gifu Prefectural Institute of Information Technology Research Report No. 10

  However, in Patent Document 1 and Non-Patent Document 1, the paddy weeding robot has a reverse U-shape when viewed from the front, and has a configuration in which a seedling passes through the center of the machine body and travel devices are provided on both sides of the seedling. Therefore, the structure of these paddy field weeding robots is complicated and large, and there is a problem that the manufacturing cost is high. In addition, since the seedling passes through the center of the aircraft, there is a problem that when the seedling grows, the seedling and the paddy weeding robot come into contact with each other and damage the seedling.

  Accordingly, an object of the present invention is to provide an electric work vehicle that can perform weeding work between crop strips by autonomous traveling and can be manufactured at low cost.

In order to solve the above problems, an electric work vehicle according to the present invention includes:
A pair of left and right traveling devices;
A pair of left and right traveling motors for driving the respective traveling devices;
A weeding device for weeding,
A control unit storing an autonomous traveling program for autonomous traveling,
In an electric work vehicle that can be weeded autonomously,
An imaging device capable of imaging the periphery of the electric work vehicle,
The control unit identifies a crop based on color information of an image captured by the imaging device,
It is characterized by weeding by autonomously running between the crop streaks.

A pair of left and right traveling devices;
A pair of left and right traveling motors for driving the respective traveling devices;
A weeding device for weeding,
A control unit storing an autonomous traveling program for autonomous traveling,
In an electric work vehicle that can be weeded autonomously,
An imaging device capable of imaging the periphery of the electric work vehicle,
The control unit identifies a travelable area based on color information of an image captured by the imaging device,
Weeding is carried out autonomously in the travelable area.

Furthermore, a colorless and transparent protective cover for protecting the imaging device;
A removal device for removing deposits on the outer surface of the protective cover,
The imaging device is capable of imaging the periphery of the electric work vehicle through the protective cover,
The controller is configured to identify deposits on the outer surface of the protective cover based on the color information of the image and to operate the removing device.

Furthermore, the protective cover has a cylindrical shape, and the imaging device is disposed inside the protective cover.
The removing device includes a wiper that contacts the outer peripheral surface of the protective cover, and a rotating device that rotates the protective cover about a cylindrical axis.
The protective cover is rotated, and the adhered matter adhered to the outer peripheral surface of the protective cover is removed by the wiper.

  Further, the control unit identifies a farmland boundary based on the color information of the image, and autonomously travels within the boundary.

  Furthermore, an obstacle detection sensor for detecting an obstacle around the electric work vehicle is provided.

  Furthermore, the weeding device is a rake provided at the rear of the vehicle, and is characterized in that it can rotate about the horizontal direction as an axis.

  Furthermore, a load sensor that detects a load applied to the rake, and a rotation device that rotates the rake about the left-right direction are provided.

  Furthermore, an alarm device is provided.

  Furthermore, the communication apparatus is provided.

  According to the electric work vehicle of the present invention, a pair of left and right traveling devices, a pair of left and right traveling motors that drive the respective traveling devices, a weeding device for weeding, and an autonomous traveling program for autonomous traveling An electric working vehicle capable of weeding by autonomous traveling, and an imaging device capable of imaging the periphery of the electric working vehicle, wherein the control unit is configured to capture an image captured by the imaging device. A crop is identified based on the color information, and weeds by autonomously running between the strips of the crop.

  Thereby, the autonomous running between the strips of the crop is possible with a simple configuration and control. Therefore, it is possible to provide an electric working vehicle that can be weeded between crop strips by autonomous traveling and can be manufactured at low cost.

  According to the electric work vehicle of the present invention, a pair of left and right traveling devices, a pair of left and right traveling motors that drive the respective traveling devices, a weeding device for weeding, and an autonomous for traveling autonomously. An electric work vehicle capable of weeding by autonomous running, and having an imaging device capable of imaging the periphery of the electric work vehicle, the control unit being imaged by the imaging device A travelable area is identified based on the color information of the image, and autonomous travel is performed in the travelable area to perform weeding.

  Thereby, it is possible to autonomously travel in an area (a travelable area) free from crops and obstacles with a simple configuration and control. Therefore, it is possible to provide an electric working vehicle that can be weeded between crop strips by autonomous traveling and can be manufactured at low cost.

  Furthermore, according to the electric work vehicle of the present invention, it is provided with a colorless and transparent protective cover that protects the imaging device, and a removal device that removes deposits on the outer surface of the protective cover, and the imaging device includes the protection device. The periphery of the electric work vehicle can be imaged through the cover, and the control unit identifies the deposit on the outer surface of the protective cover based on the color information of the image and activates the removing device. Even if dirt or insects adhere to the outer surface of the cover, these deposits can be removed. Therefore, it is possible to continue the autonomous traveling while avoiding the imaging failure due to the attached matter (reflection of the attached matter to the image), and the efficiency of the weeding work is improved.

  Furthermore, according to the electric work vehicle of the present invention, the protective cover has a cylindrical shape, the imaging device is disposed inside the protective cover, and the removing device contacts the outer peripheral surface of the protective cover. It is composed of a wiper and a rotating device that rotates the protective cover about a cylindrical axis, and rotates the protective cover, so that the adhered matter adhered to the outer peripheral surface of the protective cover is removed by the wiper. Even if dirt or insects adhere to the outer surface of the protective cover, these deposits can be removed. Therefore, it is possible to continue the autonomous traveling while avoiding the imaging failure due to the attached matter (reflection of the attached matter to the image), and the efficiency of the weeding work is improved. Moreover, since the removal device has a simple configuration, it can be manufactured at a low cost.

  Furthermore, according to the electric work vehicle of the present invention, the control unit identifies the boundary of the farmland based on the color information of the image, and autonomously travels within the boundary. Therefore, the position information of the boundary of the farmland is required. Therefore, it is possible to run independently in the farmland.

  Furthermore, according to the electric work vehicle of the present invention, since the obstacle detection sensor for detecting an obstacle around the electric work vehicle is provided, a collision with the obstacle can be avoided and a failure of the electric work vehicle can be prevented. In addition, the collision between the electric work vehicle and the crop can be reliably avoided, and the crop can be prevented from being damaged.

  Furthermore, according to the electric work vehicle of the present invention, the weeding device is a rake provided at the rear part of the vehicle and is rotatable about the left and right direction, so that weeding is possible with a simple configuration. The manufacturing cost of the work vehicle can be reduced. Moreover, even if the terrain has unevenness, the rake rotates corresponding to the unevenness, so that weeding unevenness can be reduced and weeding can be performed more reliably.

  Furthermore, according to the electric work vehicle of the present invention, since the load sensor that detects the load applied to the rake and the rotation device that rotates the rake about the left-right direction as an axis, the rake is caught on the ground. It can be detected and can be removed by rotating the rake. Therefore, it is possible to prevent the autonomous traveling of the electric work vehicle from being stopped due to the rake being caught on the soil, and to continue the autonomous traveling, thereby improving the efficiency of the weeding work. Moreover, it can prevent that an excessive load is added to an electric work vehicle, and can reduce the failure of an electric work vehicle.

  Furthermore, according to the electric work vehicle of the present invention, since the alarm device is provided, it is possible to clearly notify the surroundings of a malfunction or the like of the electric work vehicle. Therefore, the malfunction of the electric work vehicle can be detected at an early stage, and the efficiency of the weeding work is improved.

  Furthermore, according to the electric work vehicle of the present invention, since the communication device is provided, the electric work vehicle can be remotely operated. Moreover, it becomes possible to confirm the state of the crop and the operating state of the electric work vehicle from a remote place. Therefore, the work amount of the worker can be reduced and the efficiency of the weeding work is improved.

It is a left view which shows the paddy field weeding robot as an example of the electric work vehicle which concerns on embodiment of this invention. It is a top view of FIG. It is a perspective view of a protective cover. It is a top view for demonstrating a protective cover and a removal apparatus. It is a figure for demonstrating an example of the processing flow of control. It is the schematic which shows an example of the image imaged by the imaging device. It is the schematic which shows the image which performed the image division process on the image in FIG. It is a figure for demonstrating the determination point in the image which performed the image division process. It is the schematic which shows the image which performed the image division process to an example of the image imaged just before the eyelid. It is a schematic diagram explaining the turning at the time of moving to the space | interval of another seedling. It is a figure for demonstrating another determination point in the image which performed the image division process. It is a figure for demonstrating an example of the processing flow of control. It is the schematic which shows an example of the image imaged by the imaging device. It is the schematic which shows the image which performed another image division process on the image in FIG. It is a top view which shows an example of the paddy field weeding robot which concerns on another embodiment of this invention.

  Hereinafter, the best mode for carrying out the present invention will be described in detail with reference to the drawings. In this specification, “front” refers to the forward direction of the electric work vehicle, “rear” refers to the reverse direction, “left / right” refers to “left / right”, “up / down”, respectively. Each means the “up and down” direction of the electric work vehicle. FIG. 1 is a left side view showing a paddy weeding robot as an example of an electric work vehicle according to an embodiment of the present invention. FIG. 2 is a plan view of FIG. FIG. 3 is a perspective view of the protective cover as viewed obliquely from above. FIG. 4 is a plan view for explaining the protective cover and the removing device.

  As shown in FIGS. 1 and 2, a paddy weeding robot (electric work vehicle) 1 includes a body 10, a pair of left and right traveling devices 20 (20R, 20L), and left and right driving the respective traveling devices 20R, 20L. A pair of traveling motors 26 (26R, 26L) and a weeding device 30 are provided. The paddy weeding robot 1 also includes an imaging device 50 that can capture the periphery of the paddy weeding robot 1, a battery 60, an alarm device 65, a control unit 70, and the like. The travel motor 26, the imaging device 50, the alarm device 65, and the like operate under the control of the control unit 70.

  The airframe 10 is mainly composed of a resin, and includes an airframe lower part 11 and an airframe upper part 12 disposed above the airframe lower part 11. The body lower part 11 is a substantially rectangular parallelepiped elongated in the front-rear direction, the front part has a tapered shape in side view, and the bottom part has a ship bottom shape. In the paddy weeding robot 1, the lower body 11 serves as a float, and gives buoyancy to the body 10 when traveling in the paddy field. The machine body upper part 12 is a substantially rectangular parallelepiped elongated in the front-rear direction. A front portion of the upper body 12 includes a colorless and transparent protective cover 40 that protects the imaging device 50, and the imaging device 50 can image the front of the aircraft 10 through the protective cover 40.

  The traveling device 20 (20R, 20L) is a crawler traveling device and has a symmetrical shape on the left and right. The traveling device 20 includes a drive wheel 21, two driven wheels 22, three auxiliary rollers 23, a crawler belt 24, an attachment frame 25, and the like. The crawler belt 24 is wound around the drive wheel 21, the two driven wheels 22, and the three auxiliary rollers 23. Then, the pair of left and right traveling motors 26R and 26L provided in the body 10 rotate the left and right drive wheels 21R and 21L independently. That is, the pair of left and right traveling motors 26R and 26L drive the left and right traveling devices 20R and 20L independently of each other. The traveling motor 26 is connected to the left and right traveling devices 20 via gears (not shown). Further, the electric power of the traveling motor 26 is supplied from a battery 60 provided in the machine body 10.

  By independently driving the left and right traveling devices 20R and 20L, the paddy field weeding robot 1 can be moved forward, backward, turned, and the like. During forward and reverse travel, the left and right traveling devices 20R and 20L are driven in the same direction and at the same speed. At the time of turning, the right traveling device 20R and the left traveling device 20L are driven at different speeds. In addition, a zero turn is possible by driving the right traveling device 20R and the left traveling device 20L in the reverse direction.

  The traveling device 20 is not limited to the above-described configuration, and may be any device that enables the paddy weeding robot 1 to travel (forward, reverse, turn, etc.). For example, it may be composed of four wheels on the front, rear, left and right. The front left and right wheels are driven wheels, and the rear left and right wheels are driven by two travel motors 26R and 26L, respectively. The structure which makes it a drive wheel may be sufficient. Moreover, it is good also as a structure which equips the outer periphery edge part of a wheel with a some fin and can scrape water with this fin.

  The weeding device 30 is a rake, and includes five arms 31, a rotating shaft 32, left and right mounting arms 33, and the like, and is provided at the rear portion of the body 10. The arm 31 is substantially L-shaped in a side view. One end of the arm 31 is tapered, and the other end is provided with a through hole. A rotation shaft 32 is inserted into the through holes of the five arms 31. The rotating shaft 32 is fixed to the upper ends of the left and right mounting arms 33. The lower ends of the left and right attachment arms 33 are attached to the rear ends of the airframe 10 (the airframe lower part 11). Accordingly, one end of each of the five arms 31 is supported by the body 10 so as to be rotatable about the left-right direction as an axis. The five arms 31 are aligned in the left-right direction of the body 10 at substantially equal intervals. Further, the tapered tips of the five arms 31 face downward. The tapered tip of the arm 31 pierces the paddy soil by the weight of the arm 31. And the weeding apparatus 30 digs up the soil of a paddy field and weeds by the paddy field weeding robot 1 moving forward.

  As described above, the weeding device 30 is a rake having a simple configuration and can be manufactured at low cost. In addition, since the arm 31 is supported by the body 10 so as to be rotatable about the left-right direction as an axis, even if there is unevenness on the ground, the arm 31 rotates corresponding to the unevenness, so that weeding unevenness can be reduced. And weeding more reliably.

  Here, the weeding device 30 is not limited to the above-described configuration, and any device that can be weeded as the paddy weeding robot 1 moves forward may be used. For example, the five arms 31 may be integrally supported by the body 10 so as to be rotatable about the left-right direction as an axis. One end of the arm 31 may be fixed to the rotating shaft 32, and the rotating shaft 32 may be rotatably supported by the mounting arm 33. Further, the number of arms 31 is not limited to five. The number and arrangement positions of the arms 31 are appropriately designed according to the spacing between the seedlings.

  As shown in FIG. 3, the protective cover 40 includes an upper part 41 on a circular plate and a cylindrical body part 42 depending from the periphery of the upper part 41. A pulley 43 and a rotation shaft 44 that are coaxial with the cylindrical axis of the body portion 42 are formed on the upper surface of the upper portion 41. The upper portion 41 has a plurality of through holes 45 penetrating in the vertical direction. The protective cover 40 is rotatably supported by the machine body 10 about a cylindrical axis by a rotation shaft 44.

  As shown in FIG. 4, an electric motor 46 that rotates the protective cover 40 about the cylindrical axis is provided inside the machine body 10. More specifically, a pulley 47 is fixed to the drive shaft of the electric motor 46, and an endless belt 48 is stretched between the pulley 43 and the pulley 47. The airframe 10 also includes two wipers 49R and 49L extending in the cylindrical axis direction that abuts on the outer peripheral surface of the body portion 42 of the protective cover 40 (see FIG. 3). The two wipers 49R and 49L are disposed in the left and right symmetrical positions in the front part of the airframe 10, and the protective cover 40 positioned between the two wipers 49R and 49L and the two wipers 49R and 49L. The outer peripheral surface of the body portion 42 is exposed to the outside of the body 10. A sealing member (not shown) is appropriately provided between the protective cover 40 and the machine body 10 so that water, soil, or the like does not enter the machine body 10 and the protective cover 40.

  The imaging device 50 is attached to the body 10 inside the protective cover 40. That is, even if the protective cover 40 rotates about the cylindrical axis, the imaging device 50 does not rotate. The imaging device 50 can image the front of the airframe 10 through the body 42 of the colorless and transparent protective cover 40. Since the protective cover 40 is colorless and transparent, even if the imaging device 50 captures an image through the body 42 of the protective cover 40, the captured image is not affected by the protective cover 40. The imaging direction of the imaging device 50 is the arrow direction in FIG. 4, and images the front of the body 10 from between the two wipers 49 </ b> R and 49 </ b> L.

  Here, since the outer peripheral surface of the trunk portion 42 of the protective cover 40 positioned in the imaging direction of the imaging device 50 is exposed to the outside of the machine body 10, soil, insects, or the like may adhere to the outer peripheral surface. When the imaging device 50 captures an image with dirt, insects, or the like attached thereto, the attached image may appear in the captured image, and the front of the machine body 10 may not be clearly imaged. In such a case, the protective cover 40 is rotated about the cylindrical shaft by driving the electric motor 46 described above. By rotating the protective cover 40, dirt and insects attached to the outer peripheral surface of the body portion 42 of the protective cover 40 can be removed by the wipers 49R and 49L. Therefore, the electric motor 46 and the wipers 49R and 49L serve as a removing device that removes deposits on the outer peripheral surface of the body portion 42 of the protective cover 40.

  Further, the through hole 45 formed in the protective cover 40 communicates the inside of the protective cover 40 and the inside of the machine body 10, and makes it difficult for air to stay inside the protective cover 40. The through hole 45 prevents condensation from forming on the inner surface of the protective cover 40. Therefore, the imaging device 50 can image the front of the machine body 10 without being affected by the condensation. In addition, what is necessary is just to be able to prevent that dew condensation generate | occur | produces on the inner surface of the protective cover 40, and the dew condensation prevention mechanism may be what kind. For example, instead of the through hole 45, a heating wire for preventing condensation may be provided on the body portion 42 of the protective cover 40. However, since the protective cover 40 rotates, the wiring of the heating wire becomes complicated. Therefore, it is preferable to prevent dew condensation by the through hole 45 having a simpler configuration.

  The alarm device 65 is arranged on the upper part of the machine body 10, and notifies the surroundings of an abnormality when the paddy weeding robot 1 is in an abnormal state (for example, when autonomous traveling becomes impossible). The alarm device 65 is not particularly limited as long as it can notify the surroundings of an abnormality. For example, an alarm sound generating device or a lamp blinking device that notifies an abnormality to the surroundings by sound or light may be used. The paddy weeding robot 1 may be configured not to include the alarm device 65, but the paddy weeding robot 1 travels between the streak of seedlings as crops. It may be difficult to do. Therefore, it is preferable to provide the alarm device 65, and it is preferable to notify the surroundings of abnormality by both sound and light. Also, by providing the alarm device 65, it can be recognized early that the paddy weeding robot 1 is in an abnormal state, the time during which the paddy weeding robot 1 is in an abnormal state can be shortened, and the efficiency of the weeding work can be reduced. Will improve.

  The control part 70 is arrange | positioned inside the body 10 and is comprised from a calculation part, a memory | storage part, etc. which are not shown in figure, and controls each apparatus. In addition, the control unit 70 includes an autonomous traveling program for autonomous traveling, an image processing program for performing image processing of an image captured by the imaging device 50 described later, and the like. Then, the control unit 70 controls the image processing and the traveling motor 26 to cause the paddy weeding robot 1 to autonomously travel.

  The paddy weeding robot 1 includes a communication device (not shown) connected to the control unit 70. The communication device transmits and receives information wirelessly, and transmits and receives information to and from an external operation device (not shown). The external operation device is for a worker to remotely operate the paddy weeding robot 1 and is capable of transmitting and receiving information to and from the paddy weeding robot 1.

  Note that the external operation device is configured to be able to confirm an image captured by the image capturing device 50 with a monitor, and an operator can confirm the state of the seedling even at a location away from the paddy field. Therefore, the worker does not need to move to the paddy field, and the amount of work for the worker can be reduced. Further, the external operation device is configured to notify the worker that the alarm device 65 has been activated, and can notify the worker who is away from the paddy field that the paddy weeding robot 1 is in an abnormal state. Therefore, the time during which the paddy weeding robot 1 is in an abnormal state can be shortened, and the efficiency of the weeding work is improved. Further, when the paddy weeding robot 1 is collected, the paddy weeding robot 1 hidden by a seedling or the like can be easily found by operating the alarm device 65 by an external operation device. Here, the paddy weeding robot 1 may be configured not to be operated by the external operation device, but for these reasons, the configuration of performing the operation by the external operation device is preferable.

  Next, the autonomous traveling of the paddy field weeding robot 1 will be described. The paddy field weeding robot 1 identifies a seedling as a crop based on an image captured by the imaging device 50, and autonomously travels between the identified seedling strips. In FIG. 5, the figure for demonstrating an example of the processing flow of control in the autonomous running of the paddy field weeding robot 1 is shown.

  First, the paddy weeding robot 1 is arranged between the seedling streaks so that the traveling direction is parallel to the seedling streak direction, and an external operation device is instructed to start autonomous traveling. The running is started (START).

  The paddy weeding robot 1 that has started autonomous traveling images the front of the paddy weeding robot 1 with the imaging device 50 (step S1). A schematic diagram of an example of a captured image is shown in FIG. Here, FIG. 6 shows a cocoon 81 as a boundary between the seedling 80 and a paddy field (farmland), and a straight line L1 is a center line in the horizontal direction of the image, and a straight line extending forward from the center in the horizontal direction of the paddy field weeding robot 1. Matches. That is, the straight line L1 indicates the traveling direction when the paddy weeding robot 1 moves forward.

  Next, the control unit 70 performs a known image division process for dividing the image into a plurality of regions based on the color (step S2). In the embodiment of the present invention, a flood fill process, which is an example of an image division process, is performed. The colors used for the division are the color of the seedling (green), the color of the straw that fills the paddy field (brown), and soil or insects attached to the outer peripheral surface of the protective cover 40. The reflected color (black) and other colors. Here, the adhering matter adhering to the outer peripheral surface of the protective cover 40 is imaged as black because the distance to the imaging device 50 is short. That is, the image division processing here is a region division based on only three types of seedling color, cocoon color, and attached color, and is relatively simple. Note that the image dividing process is not particularly limited as long as it is possible to extract and divide the areas of the seedling, the cocoon, and the attached matter. The color threshold is designed as appropriate.

  FIG. 7 is a schematic diagram showing an image obtained by performing this image division process on the image in FIG. In FIG. 7, for convenience of explanation, a region 82 in which a seedling is shown is hatched with an upward-sloping diagonal line, and a region 83 in which a cocoon is reflected is represented as a vertical line hatching. Note that FIG. 6 is an image taken with no deposit on the outer peripheral surface of the protective cover 40, and FIG. 7 does not have a black region.

  The paddy field weeding robot 1 autonomously travels using a region 84 other than the region 82 in which the seedlings identified by the image dividing process and the region 83 in which the cocoons are reflected as a travelable region. Here, when there is a deposit on the outer peripheral surface of the protective cover 40, a seedling or a cocoon is not clearly imaged due to the reflection of the deposit. Therefore, the presence / absence of deposits on the outer peripheral surface of the protective cover 40 is determined using an image (FIG. 7) that has been subjected to image division processing (step S3).

  The determination of the attached matter is made based on the size of the region where the attached matter is reflected, that is, the entire black region in the entire image subjected to the image division processing. When the entire black area is 5% or more in size with respect to the entire image, it is assumed that there is a deposit on the outer peripheral surface of the protective cover 40.

  Therefore, when the entire black area is 5% or more of the entire image, the electric motor 46 is driven, the protective cover 40 is rotated about the cylindrical axis, and the protective covers are provided by the wipers 49R and 49L. The deposits on the outer peripheral surface of 40 are removed (step 4). Here, the rotation direction and rotation angle of the protective cover 40 are appropriately set based on the size of the protective cover 40 and the like. It is desirable that the protective cover 40 is sequentially rotated in both the clockwise and counterclockwise directions in FIG. With such a configuration, the deposits are wiped off from different directions by the wipers 49R and 49L, so that the deposits can be more reliably removed. And the front of the paddy weeding robot 1 is imaged by the imaging device 50 (step S1), and step S2 and step S3 are repeated.

  In addition, when this step S4 is repeated continuously, the number of times of repetition is stored. When this number of times reaches a predetermined number, for example, 5 times, the alarm device 65 is activated. In step S4, the deposits on the outer peripheral surface of the protective cover 40 are removed. However, when step S4 is repeated continuously, the deposits cannot be removed. Therefore, the paddy weeding robot 1 cannot clearly capture the front, and informs the surroundings by the alarm device 65 that autonomous traveling is impossible.

  Here, the threshold value of the size of the entire black region used for the determination of the presence / absence of adhering matter on the outer peripheral surface of the protective cover 40 is not limited. However, it is preferable that the threshold value of the size of the entire black region is at least 1 percent and at most 10 percent with respect to the entire image. If it is less than 1 percent, even if it is a fine deposit, the deposit is removed, so that the autonomous traveling is frequently interrupted and the efficiency of the weeding work is reduced. If it exceeds 10 percent, a clear image cannot be taken, a problem may occur in the autonomous running control described later, and autonomous running between the seedling strips may not be possible.

  Moreover, the determination of the presence or absence of the deposit on the outer peripheral surface of the protective cover 40 is not limited to the above method. For example, the size of each of the plurality of black areas is calculated, and if any of the black areas is larger than a predetermined size, it is determined that there is an attachment on the outer peripheral surface of the protective cover 40 Good. Moreover, you may determine combining these.

  When the entire black area is less than 5% of the entire image, it is assumed that there is no deposit on the outer peripheral surface of the protective cover 40, and the autonomous running control step based on the image subjected to the image division processing Migrate to Here, since the paddy weeding robot 1 travels between the seedling strips, the seedling strips are positioned on the left and right sides of the paddy weeding robot 1. Therefore, using the image (FIG. 7) subjected to image division processing, the relationship between the paddy weeding robot 1 and the seedling strips located on the left and right sides of the paddy weeding robot 1 is recognized, and autonomous traveling is performed based on the relationship. Control to do.

  As shown in FIG. 8, a point that is on the straight line L1 of the center line in the horizontal direction of the image (FIG. 7) subjected to the image division process and slightly below the center in the vertical direction (position at a distance H from the lowest end of the image) The determination point is P. This determination point P corresponds to the traveling direction when the paddy weeding robot 1 moves forward. The vertical position in the image of the determination point P corresponds to the distance from the paddy weeding robot 1. That is, it is possible to determine the situation in front of the paddy weeding robot 1 depending on which region the determination point P is located in. The determination point P here corresponds to immediately before the paddy field weeding robot 1 (about 50 cm forward). Note that the position of the determination point P is not limited to the above-described position, and is set as appropriate.

  Therefore, it is determined in which region the determination point P is located with respect to the image subjected to the image division processing (step S5). When the determination point P is located in the region 84, it can be seen that the paddy weeding robot 1 is located in a place where there is no seedling or pod immediately before. Here, since the paddy weeding robot 1 travels between the seedling strips, it can be seen that the paddy weeding robot 1 is located between the seedling strips. In addition, when the determination point P is located in the area | region 82, it turns out that the paddy field weeding robot 1 is located just before a seedling. Moreover, when the determination point P is located in the area | region 83, it turns out that the paddy field weeding robot 1 is located just before the paddy. Moreover, when the determination point P is located in the black region, it cannot be determined what the situation is immediately before the paddy weeding robot 1.

  First, the case where the determination point P is located within the region 84 in step S5 described above will be described. At this time, since the paddy weeding robot 1 is located between the seedling strips, the paddy weeding robot 1 is caused to travel forward (step S6). Here, when the paddy weeding robot 1 travels forward, it needs to be advanced along the seedling strip. Therefore, before the paddy weeding robot 1 travels forward, the positional relationship between the paddy weeding robot 1 and seedlings (seedling strips) located on the left and right sides of the paddy weeding robot 1 is grasped from the image.

  In the image (FIG. 8) subjected to the image division process, a line passing through the determination point P and orthogonal to the straight line L1 is defined as a straight line L2. A point that is the intersection of the straight line L2 on the right side of the straight line L1 and the edge of the region 82 and is closest to the determination point P is defined as a point QR. The distance between the point P and the point QR is WR. A point that is the intersection of the straight line L2 on the left side of the straight line L1 and the edge of the region 82 and closest to the determination point P is defined as a point QL. The distance between the determination point P and the point QL is WL.

  Since the paddy weeding robot 1 is located between the seedling strips, the positional relationship between the paddy weeding robot 1 and the seedling strips on the left and right sides of the paddy weeding robot 1 can be known from the distances WR and WL. When the distance WR is smaller than the predetermined value, it can be seen that the paddy weeding robot 1 is located near the right seedling streak. When the distance WL is smaller than the predetermined value, it can be seen that the paddy weeding robot 1 is located near the left seedling streak. Note that the above-described predetermined values for the distances WR and WL vary depending on the width of the seedling streak and the imaging direction of the imaging device 50, and are set as appropriate.

  Therefore, when the distance WR and the distance WL are larger than the predetermined values, since there is a space between the paddy weeding robot 1 and the left and right side seedling strips, the paddy weeding robot 1 is caused to travel forward (step S61).

  When the distance WR is smaller than the predetermined value, the paddy weeding robot 1 is located near the right seedling streak, and therefore the paddy weeding robot 1 is turned left by a predetermined angle or is made to move forward after being zero-turned ( Step S62).

  When the distance WL is smaller than the predetermined value, the paddy weeding robot 1 is located near the left seedling streak, and therefore the paddy weeding robot 1 is turned forward by a predetermined angle to the right or is made to travel forward (zero turn) ( Step S63). That is, in steps S62 and S63, the direction of the paddy weeding robot 1 is changed and the vehicle is allowed to travel forward.

  When the distance WR and the distance WL are smaller than the predetermined values, the process proceeds to an abnormal mode to be described later (step S64).

  Here, the predetermined angle of turning or zero turn is not limited, but is preferably 5 ° or more and 30 ° or less. When the angle is less than 5 °, the orientation of the paddy weeding robot 1 is not changed much. Also, if the angle exceeds 30 °, the direction of the paddy weeding robot 1 is changed too much, and when it is run forward, the direction is changed (left side when step S62 is selected, when step S63 is selected) Will approach the seedling strip on the right).

  The predetermined angle may be based on the values of the distances WR and WL. That is, the predetermined angle may be changed according to the distance between the paddy weeding robot 1 and the left and right seedling strips of the paddy weeding robot 1. With this configuration, the paddy weeding robot 1 can be advanced more accurately along the seedling strip.

  When Steps S61, S62, and S63 are selected, after moving forward for a predetermined time, the front of the paddy weeding robot 1 is imaged by the imaging device 50 (Step S1), and Steps S2 to S5 are repeated. The predetermined time is not particularly limited and is, for example, 1 second. If the predetermined time is long, the possibility of coming into contact with the seedling at the time of advance increases, and if the predetermined time is short, the calculation load of the control unit 70 increases. In addition, it is good also as a structure which replaces with predetermined time and performs step S1 after advancing a predetermined distance.

  The abnormal mode described above represents a state in which the paddy weeding robot 1 cannot move forward. In step S64, when the distance WR and the distance WL are smaller than the predetermined values, the paddy weeding robot 1 is not located between the seedling strips, the seedling has fallen, or the width between the seedling strips is narrow. It turns out that it is. When the paddy weeding robot 1 is moved forward in such a state, the paddy weeding robot 1 may come into contact with the seedlings, and the paddy weeding robot 1 may be unable to travel (a state in which it cannot move forward, reverse, or turn). Further, the seedling may be damaged by the advance of the paddy weeding robot 1. Therefore, the paddy weeding robot 1 is zero-turned by a predetermined angle (step S7). By doing so, the orientation of the paddy weeding robot 1 is changed. And the front of the paddy weeding robot 1 is imaged by the imaging device 50 (step S1), and steps S2 to S5 are repeated.

  In addition, when this step S7 is repeated continuously, the number of times of repeating the same is stored. When this number of times reaches a predetermined number, for example, 5 times, the alarm device 65 is activated. In step S7, the paddy weeding robot 1 makes a zero turn and changes the direction. However, when step 5 is repeated continuously, the paddy weeding robot 1 cannot move forward even if the direction is changed. Therefore, the alarm device 65 informs the surroundings that such a paddy weeding robot 1 is in an incapable autonomous running.

  Here, the predetermined angle at the time of the above-described zero turn is not limited, but is preferably 10 ° or more and 50 ° or less. When the angle is less than 10 °, the orientation of the paddy weeding robot 1 is not changed so much and the situation is not changed much. Moreover, when an angle exceeds 50 degrees, the direction of the paddy field weeding robot 1 is changed too much, and the situation will change too much.

  Note that step S7 is for changing the direction of the paddy weeding robot 1 from a state where it cannot move forward to a state where it can move forward, and the change of direction is not limited to that by zero turn. For example, the configuration may be such that a predetermined angle is zero-turned after traveling backward for a predetermined time.

  Next, the case where the determination point P is located in the region 82 in step S5 described above will be described. At this time, the paddy weeding robot 1 is located immediately before the seedling. Therefore, since the paddy weeding robot 1 cannot move forward, the paddy weeding robot 1 is zero-turned by a predetermined angle as the abnormal mode described above (step S7). By doing so, the orientation of the paddy weeding robot 1 is changed. And the front of the paddy weeding robot 1 is imaged by the imaging device 50 (step S1), and steps S2 to S5 are repeated. In addition, when step S7 is repeated a predetermined number of times, the paddy weeding robot 1 is in a state where it cannot autonomously travel, so the alarm device 65 is activated to notify the surroundings.

  Next, the case where the determination point P is located in the region 83 in step S5 described above will be described. At this time, the paddy weeding robot 1 is located immediately before the paddy. That is, the paddy weeding robot 1 is located near the end of the seedling strip. Therefore, the paddy weeding robot 1 is moved from the space between the seedlings that have been traveling so far to another space between the seedlings (step S8).

  FIG. 9 shows a schematic diagram of an image obtained by subjecting the image taken when the paddy weeding robot 1 is just before the paddy to the above-described image division processing. In FIG. 9, the determination point P is located in the region 83. A point that is the intersection of the straight line L1 on the lower side of the straight line L2 and the edge of the region 83 and is closest to the determination point P is defined as a point QD. The distance between the determination point P and the point QD is HD.

  Here, as shown in FIG. 10, the paddy weeding robot 1 is located at a point R immediately before the paddy (the point where the image was captured), and the distance HD in the image (FIG. 9) subjected to the above-described image division processing. Based on this, the positional relationship (distance D) between the ridge 81 and the paddy field weeding robot 1 is known. Note that the arrows in FIG. 10 indicate the traveling direction of the paddy weeding robot 1. Further, the width WP between the seedling streaks is determined at the time of seedling planting and is approximately constant. Therefore, the paddy weeding robot 1 is moved forward by a predetermined distance D1 (step S81) and is positioned at the point S. Step S81 is for setting the distance from the ridge 81 to a predetermined distance D2 determined in advance when step S81 is completed. The predetermined distance D1 for moving the paddy weeding robot 1 forward is calculated from the distance HD described above.

  Next, the paddy weeding robot 1 is turned 180 ° left or right by a predetermined turning radius R1 (step S82) (left turning in FIG. 10) and is positioned at the point T. This predetermined turning radius R1 is half of the width WP between the seedling strips. Here, since the distance between the paddy weeding robot 1 and the ridge 81 is always a constant distance D2 at the end of step S81, it moves between adjacent seedlings by turning at a constant turning radius. It becomes possible. Then, after turning, the front of the paddy weeding robot 1 is imaged by the imaging device 50 (step S1), and steps S2 to S5 are repeated.

  Here, it is stored whether the left or right turn is made in step S82. Whether to turn left or right in step S82 depends on the stored turning direction of immediately preceding step S82, and is turned in the direction opposite to the turning direction of immediately preceding step S82. In this way, by sequentially turning the turning direction in the opposite direction, it is possible to move between different seedling strips sequentially. In FIG. 10, the right seedlings move sequentially from the right seedling streak to the left seedling streak.

  In addition, step S8 should just be able to move the paddy weeding robot 1 to the space | interval of another seedling from the space | interval of the seedling which was drive | working until now, and is not limited to the above-mentioned process flow. For example, instead of a right or left turn, a turn that turns twice at a right angle using a zero turn may be used.

  Next, a description will be given of the case where the determination point P is located in the black region in step S5 described above. At this time, it is impossible to determine the situation immediately before the paddy weeding robot 1 due to the deposits on the outer peripheral surface of the protective cover 40. Therefore, in order to remove the deposit on the outer peripheral surface of the protective cover 40, the process proceeds to step S4 described above (step S9), and the deposit is removed. And the front of the paddy weeding robot 1 is imaged by the imaging device 50 (step S1), and steps S2 to S5 are repeated. If Step S9 is repeated for a predetermined number of times, the paddy weeding robot 1 is in a state where it cannot autonomously travel, so the alarm device 65 is activated to notify the surroundings.

  When the autonomous traveling of the paddy weeding robot 1 is to be terminated, an instruction to terminate autonomous traveling is given by an external operation device, and the autonomous traveling of the paddy weeding robot 1 is terminated (END). In addition, in step S4, S7S9, it is good also as a structure which complete | finishes autonomous driving | running | working, after operating the alarm device 65 for the predetermined time.

  Based on the processing flow as described above, the paddy weeding robot 1 can identify seedlings based on the color information of the image captured by the imaging device 50, and autonomously travel between the streak strips to perform weeding. It becomes possible. Note that this autonomous traveling is simple control based only on an image without the need to detect the position of the paddy weeding robot 1 and perform autonomous traveling (for example, using a GPS system or a direction sensor). In addition, the autonomous running control is performed based on the distance between specific areas in the image subjected to the image division process, and a simple calculation and processing flow are obtained. Therefore, the paddy weeding robot 1 can be manufactured at low cost. Further, the paddy weeding robot 1 travels between the seedling streaks, and its lateral width is narrower and smaller than that of the streak, and the paddy weeding robot 1 and the seedlings even when the seedlings grow. Will not touch and will not damage the seedlings. Further, it is possible to identify the paddy that is the boundary of the paddy field based on the color information of the image and to autonomously travel in the paddy field. Therefore, since the position information of the paddy paddy that is autonomously driven is not required, the versatility is high. In addition, even if dirt or insects adhere to the outer surface of the protective cover, these deposits can be removed, imaging defects caused by the deposits can be avoided, autonomous driving can be continued, and the efficiency of weeding work can be continued. Will improve. Moreover, since the removal device has a simple configuration, it can be manufactured at a low cost.

  Note that the autonomous traveling (steps S2 to S9) of the paddy weeding robot 1 based on an image obtained by performing image division processing on the image captured by the imaging device 50 is not limited to the above-described configuration. A plurality of configurations may be provided. As shown in FIG. 11, new determination points P1 to P6 are provided around the above-described determination point P, and the positional relationship between the six determination points P1 to P6 and the region 82 where the seedling is shown is also taken into consideration. Then, it may be configured to autonomously travel.

  With such a configuration, it is possible to more reliably travel between seedling strips. In addition, even if the airframe 10 is inclined due to the unevenness of the paddy field, the airframe 10 is inclined due to the positional relationship with the area where the seedlings are reflected and the areas where the wrinkles are reflected at each of the determination points P1 to P6. Can be recognized. Therefore, it is possible to travel between the seedlings even in such a situation.

  Here, in the above description, the vehicle autonomously travels based on the image in which the seedling region is identified by the image division process, but the present invention is not limited to this. Any specific region may be identified by image segmentation processing and autonomously traveling based on an image identifying the specific region. For example, an area where the paddy weeding robot 1 can travel is identified by image division processing, and autonomous traveling may be performed based on an image identifying the travelable area.

  Next, an example will be described in which the paddy weeding robot 1 identifies a travelable region based on an image captured by the imaging device 50 and autonomously travels in this travelable region. Here, the travelable area is a paddy soil area, and the identified paddy soil area (runnable area) autonomously travels. In FIG. 12, the figure for demonstrating an example of the processing flow of control in the autonomous driving | running | working of the paddy field weeding robot 1 in this case is shown. In addition, description is abbreviate | omitted about the content similar to the autonomous running based on the image which identified the area | region of the seedling by the above-mentioned image division process.

  First, the paddy weeding robot 1 is arranged between the seedling streaks so that the traveling direction is parallel to the seedling streak direction, and an external operation device is instructed to start autonomous traveling. The running is started (START).

  The paddy weeding robot 1 that has started autonomous traveling images the front of the paddy weeding robot 1 with the imaging device 50 (step S21). A schematic diagram of an example of a captured image is shown in FIG. Here, FIG. 13 is the same as FIG. 6 described above, and the straight lines L3 and L4 are lines orthogonal to the straight line L1. Further, the straight line L3 and the straight line L4 are lines that divide the captured image into three regions A1, A2, and A3 in the vertical direction and are divided into approximately three equal parts.

  Next, the control unit 70 performs a known image division process for dividing the image into a plurality of regions based on the color (step S22). In the embodiment of the present invention, a flood fill process, which is an example of an image division process, is performed. The color when dividing is the color of the soil of the paddy field (dark brown), the color of the straw that fills the paddy field (brown), and when the soil or insects adhere to the outer peripheral surface of the protective cover 40 The color that appears in the image (black) and other colors. That is, the image division processing here is a region division based on only three types of soil color, paddy color, and deposit color of paddy fields, and is relatively simple. Note that the image dividing process is not particularly limited as long as it is possible to extract and divide the area of paddy soil, paddy, and attached matter. The areas to be subjected to image processing are an area A1 below the image and an area A2 at the center of the image. Therefore, since the image processing range is not the entire image, the calculation load of the control unit 70 in the image processing can be reduced.

  FIG. 14 is a schematic diagram showing an image obtained by performing this image division process on the image in FIG. In FIG. 14, for convenience of explanation, a region 85 (85A, 85B, 85C, 85D, 85E, 85F) in which the paddy soil is reflected is represented as hatching with a downward slanting line. Note that FIG. 13 was taken in a state where there was no deposit on the outer peripheral surface of the protective cover 40, and FIG. 14 does not have a black region.

  Next, based on the image subjected to the image division processing (FIG. 14), the presence / absence of adhered matter on the outer peripheral surface of the protective cover 40 is determined (step S23) in the same manner as in step S3 described above. If it is determined that there is an adhering substance on the outer peripheral surface of the protective cover 40, the adhering substance on the outer peripheral surface of the protective cover 40 is removed in the same manner as in step S4 described above (step 24), and the paddy field weeding robot 1 is moved forward. An image is taken (step S21), and step S22 and step S23 are repeated.

  On the other hand, when it is determined that there is no deposit on the outer peripheral surface of the protective cover 40, the process proceeds to a step of autonomous traveling control based on the image subjected to the image division processing. The paddy weeding robot 1 autonomously travels using the soil area of the paddy field ahead as a travelable area. Here, since the paddy weeding robot 1 travels between the seedling streaks, the seedling streak is located on the left and right sides of the paddy weeding robot 1, and the seedling is located just before the paddy weeding robot 1. Not. Accordingly, the relationship between the paddy weeding robot 1 and the soil (runnable area) of the paddy field in front of the paddy weeding robot 1 is recognized based on the image that has been subjected to image division processing (autonomous operation), and autonomously based on the relationship. Control to run.

  Regions A1, A2, and A3 divided in the vertical direction of the image (FIG. 14) subjected to the image division processing correspond to the distance from the paddy weeding robot 1. The situation immediately before the paddy weeding robot 1 is imaged in the area A1, the situation ahead of the area A1 is imaged in the area A2, and the situation ahead of the area A2 is imaged in the area A3. ing. Therefore, based on how the region 85 (travelable region) in which the soil of the paddy field is reflected in the regions A1 and A2 and the region 83 in which the paddy is reflected is distributed immediately before the paddy weeding robot 1 and its Determine the situation ahead. Note that the vertical division method of the image is not limited to the above-described three regions A1, A2, and A3, and is appropriately set based on the surrounding range imaged by the imaging device 50. For example, the area A1 may be further divided into three in the left-right direction, and the center area may be set as a new area A1.

  First, it is determined how the area 85 and the area 83 are distributed in the area A1 in which the situation immediately before the paddy field weeding robot 1 is imaged in the image that has been subjected to image division processing (step S25). The distribution of the region 85 and the region 83 is determined based on the size of the region 85 and the region 83 with respect to the region A1.

  If the size of the region 85 (85A in FIG. 14) where the soil of the paddy field in the region A1 is 80% or more of the region A1 in this step S25, the paddy weeding robot 1 is in the travelable region. It can be seen that it is a travelable area immediately before. Further, when the size of the area 85 is less than 80% of the area A1, it can be seen that there is no travelable area immediately before the paddy weeding robot 1. That is, it can be seen that the paddy weeding robot 1 is located immediately before the cocoon or just before the seedling or other obstacle.

  First, the case where the size of the region 85 is 80% or more with respect to the region A1 in step S25 will be described. At this time, since the paddy weeding robot 1 is located in the travelable area and immediately before it is the travelable area, the paddy weeding robot 1 is moved forward (step S26). Here, when the paddy weeding robot 1 travels forward, it is necessary to move forward in the travelable area. Therefore, before the paddy weeding robot 1 travels forward, the distribution of the travelable area in front of the paddy weeding robot 1 is grasped from the image.

  In the image A that has been subjected to the image division processing (FIG. 14), the size of each of the plurality of travelable regions 85B, 85C, 85D, 85E, and 85F in the region A2 in which the situation ahead of the region A1 is captured is shown. Operate and select the one with the largest area size. Here, in FIG. 14, the region 85B has the largest size. Next, the centroid G of the region 85B having the maximum size is calculated, and the shortest distance GL from the centroid G to the straight line L1 is calculated.

  Then, based on the distance GL, the paddy weeding robot 1 is turned or zero-turned so that the straight line L1 runs forward so that the straight line L1 coincides with the centroid G. That is, the paddy weeding robot 1 is moved for a predetermined time in the direction of the position corresponding to the centroid G (step S261). And the front of the paddy field weeding robot 1 is imaged (step S21), and steps S22 to S25 are repeated.

  Here, when the distance GL is larger than a predetermined value, when the above-mentioned maximum area is selected, the area 85C, 85D, 85E, and 85F (adjacent seedlings) are not used instead of the area 85B that is the forward travelable area. It is considered that the paddy soil between the two strips and the paddy soil seen from the gap between the seedlings is selected. Here, the paddy weeding robot 1 travels between the streak of seedlings, and the seedlings are located on the left and right sides of the paddy weeding robot 1. Therefore, in front of the paddy field weeding robot 1, there is a travelable area that is continuous with the travelable area that has been traveled so far, and such a situation does not normally occur. However, this rarely occurs due to the fall of the seedling or the orientation of the paddy weeding robot 1 relative to the seedling strip is not parallel.

  In such a state, when traveling toward the position corresponding to the centroid G of the selected region, the paddy weeding robot 1 may approach the seedlings located on the left and right sides. Therefore, when the distance GL is larger than the predetermined value, the process shifts to the abnormal mode (step 262). Here, this abnormal mode is the same as the above-described abnormal mode (step S7), and the paddy weeding robot 1 is zero-turned by a predetermined angle (step S27). And after changing the direction of the paddy weeding robot 1, the front of the paddy weeding robot 1 is imaged (step S21), and steps S22 to S25 are repeated.

  Next, a case where the size of the region 85 is less than 80% with respect to the region A1 and the size of the region 83 with the wrinkles is less than 20% with respect to the region A1 in step S25 will be described. . At this time, the area immediately before the paddy weeding robot 1 is not a travelable area, and the paddy is not positioned immediately before. That is, it can be seen that seedlings, obstacles, and the like are located immediately before the paddy weeding robot 1 and the paddy weeding robot 1 cannot move forward. Therefore, as an abnormal mode, the paddy weeding robot 1 is zero-turned by a predetermined angle (step S27). And after changing the direction of the paddy weeding robot 1, the front of the paddy weeding robot 1 is imaged (step S21), and steps S22 to S25 are repeated.

  Next, the case where the size of the region 85 is less than 80% of the region A1 and the size of the region 83 in which the wrinkle is reflected is 20% or more of the region A1 in step S25 will be described. . At this time, the paddy weeding robot 1 is located immediately before the paddy. That is, the paddy weeding robot 1 is located near the end of the seedling strip. Therefore, similarly to step S8 described above, the paddy weeding robot 1 is moved from the seedling streak that has been traveling to another seedling streak (step S28).

  This step S28 is the same as step S8 described above, and the paddy weeding robot 1 is moved using the distance between the lowermost end of the region 83 and the lowermost end of the image (the lowermost end of the region A1) instead of the distance HD. Move forward, turn or zero turn and move between adjacent seedling strips. And the front of the paddy field weeding robot 1 is imaged (step S21), and steps S22 to S25 are repeated.

  It should be noted that the above-described threshold value of the area size, the predetermined distance, the time, and the like are designed as appropriate, and are not limited to the above-described values.

  Based on the processing flow as described above, the paddy weeding robot 1 identifies the soil of the paddy field as the travelable area based on the color information of the image captured by the imaging device 50, and autonomously travels in the travelable area. Can be weeded. In addition, this autonomous traveling will not travel except in the soil area of the paddy field. Therefore, it is possible to autonomously travel so as to avoid seedlings and other stones as obstacles. Further, the image division processing is a part of the image, and the processing flow of the autonomous running is simple. Therefore, the control of autonomous traveling becomes a simple calculation and processing flow, and the paddy weeding robot 1 can be manufactured at low cost.

Further, the configuration of the paddy field weeding robot 1 is not limited to the above-described configuration. For example, it is good also as a structure provided with the inclination sensor which detects the inclination of the body 10, for example, a gyro sensor. The inclination of the airframe 10 when an image is captured by this inclination sensor can be detected.
Since correction of the detected inclination can be added to the image and autonomous running based on the corrected image can be performed, it is possible to more autonomously run between the seedling streaks.

  Moreover, it is good also as a structure provided with the detection sensor which detects the load added to the arm 31 of the weeding apparatus 30, and the rotation apparatus which rotates this arm 31 centering on the left-right direction. In the above-described weeding device 30, one end of the arm 31 is fixed to the rotating shaft 32, and the rotating shaft 32 is rotatably supported by the mounting arm 33. And as shown in FIG. 10, the motor 35 and gear box 36 as a rotation apparatus are provided. The rotation shaft of the motor 35 is connected to the rotation shaft 32 via a gear (not shown) in the gear box 36. Therefore, the arm 31 can be rotated about the left-right direction by the motor 35.

  In such a configuration, when a value detected by a load sensor (not shown) exceeds a predetermined value, the arm 31 is rotated upward about the left-right direction and then moved forward for a predetermined time. Thereafter, the arm 31 is rotated downward about the left-right direction, and the tip of the arm 31 is brought into contact with the ground. When the paddy weeding robot 1 is operated as described above, when the arm 31 is caught on the soil, the catch can be removed. Therefore, stop of the autonomous traveling of the paddy field weeding robot 1 due to the arm 31 being caught on the soil can be prevented and the autonomous traveling can be continued, and the efficiency of the weeding work is improved. Moreover, it is possible to prevent an excessive load from being applied to the paddy weeding robot 1 and to reduce the malfunction of the paddy weeding robot.

  Note that the operation of the paddy field weeding robot 1 when the value detected by the load sensor exceeds a predetermined value is not limited to the above-described operation. Any operation that enables removal of the arm 31 from the soil may be used. For example, the arm 31 may be rotated upward with the left-right direction as an axis and may be moved backward at the same time. By setting it as such an operation | movement, it becomes possible to remove more reliably the hook of the arm 31 to the soil.

  The paddy weeding robot 1 may be configured to include an obstacle detection sensor that detects obstacles around the paddy weeding robot 1. The obstacle detection sensor only needs to be able to detect an obstacle around the paddy field weeding robot 1, and is, for example, an infrared sensor or an ultrasonic sensor. The paddy weeding robot 1 is provided with an infrared sensor, and the infrared sensor irradiates the infrared ray forward and detects the reflected infrared ray.

  By adopting such a configuration, it becomes possible to detect obstacles ahead of the traveling direction, such as workers, seedlings, and cocoons. Then, when an obstacle is detected, the paddy weeding robot 1 can perform an avoidance operation (stop, reverse, turn, etc.). Therefore, it is possible to prevent the paddy weeding robot 1 from being damaged or the seedling being damaged. Moreover, when moving from the above-mentioned seedling streak to another seedling streak, the distance between the paddy and the paddy weeding robot 1 can be detected more accurately by this infrared sensor. Can be moved more reliably.

  Here, it is desirable that the obstacle detection sensor is attached to the body 10 inside the protective cover 40 similarly to the imaging device 50 described above. With this configuration, it is possible to remove dirt, insects, and the like attached to the outer peripheral surface of the body portion 42 of the protective cover 40, and it is possible to detect an obstacle without being affected by the attached matter. The obstacle detection accuracy is improved.

  The paddy weeding robot 1 may be configured to include a battery sensor that detects the amount of electricity stored in the battery 60. When the value detected by the battery sensor falls below a predetermined value, the alarm device 65 is activated. With this configuration, it is possible to prevent the autonomous traveling of the paddy field weeding robot 1 from being stopped due to insufficient power of the battery 60 and to continue the autonomous traveling, thereby improving the efficiency of the weeding work.

  The paddy weeding robot 1 may be configured to detect the position of the paddy weeding robot 1 using a GPS system, a direction sensor, or the like. By adopting such a configuration, it is possible to efficiently autonomously travel within the paddy field. However, since autonomous running is performed by control based on images and positions, the control system becomes complicated and the manufacturing cost may increase, which is not preferable.

  Moreover, the structure (protective cover 40, electric motor 46, wiper 49R, 49L) which removes the deposit | attachment of the outer peripheral surface of the above-mentioned protective cover 40 is not specifically limited. Any structure may be used as long as it can protect the imaging device 50 with the protective cover and can remove the deposits attached to the outer surface of the protective cover, which are obstacles to imaging.

  For example, the above-described protective cover 40 may be fixed to the airframe 10 and the wiper 49 may rotate about the cylindrical axis while contacting the outer peripheral surface of the body portion 42 of the protective cover 40. With such a configuration, the portion of the protective cover 40 located inside the body 10 is unnecessary, and the protective cover 40 can be made small. Further, only the wiper provided on the outside operates, and the seal between the protective cover 40 and the airframe 10 can be easily performed.

  Further, the above-described protective cover 40 is fixed to the airframe 10, and a wiper extending in the circumferential direction is provided along the outer peripheral surface of the body portion 42 of the protective cover 40. It may be configured to be moved. With such a configuration, the portion of the protective cover 40 located inside the body 10 is unnecessary, and the protective cover 40 can be made small. Further, only the wiper provided on the outside operates, and the seal between the protective cover 40 and the airframe 10 can be easily performed.

  Moreover, as shown in FIG. 15, the structure provided with the injection nozzle 90 which injects water to the outer peripheral surface of the protective cover 40 and removes adhesion may be sufficient. By setting it as such a structure, the deposit | attachment of the outer peripheral surface of the protective cover 40 can be removed more reliably.

  The paddy weeding robot 1 may have a configuration that does not include such a removing device that removes deposits on the outer surface of the protective cover. However, the paddy weeding robot 1 autonomously travels in the paddy field and is a small robot that can travel through the streak of seedlings. Therefore, the crawler-type traveling device 20 makes it easy for the soil to jump, and the protective cover 40 is located close to the ground. Therefore, the protective cover 40 tends to adhere to the soil, and a configuration including such a removing device is preferable. .

  The imaging device 50 is not particularly limited as long as it can capture at least the front of the paddy weeding robot 1. For example, the imaging device 50 may be configured to change the imaging direction vertically and horizontally, and may be configured to be able to change the imaging direction using an electric motor or the like. In addition, the imaging device 50 may be configured to be attached to the body 10 via a support member that is slidable in the vertical direction, and is configured to be able to image the front of the paddy weeding robot 1 from above the body 10. May be. With such a configuration, it becomes easy to image seedlings from various directions without changing the direction of the paddy weeding robot 1.

  Furthermore, the imaging device 50 is not limited to one. For example, the imaging device 50 may be configured by four imaging devices so that the front, rear, and left and right sides of the paddy field weeding robot 1 can be imaged. It is possible to autonomously travel based on images captured by the four imaging devices (front, rear, and left and right side images of the paddy weeding robot 1). Therefore, it is possible to autonomously travel between the seedling strips more accurately. In addition, it is possible to easily image seedlings from a plurality of directions without changing the direction of the paddy weeding robot 1, and the state of the seedlings can be confirmed in more detail.

  Here, the protective cover provided with the removing device for removing the adhering matter on the outer surface as described above can be applied to the protection of any device that images and detects the external situation through the protective cover, and its use. Is not limited to a paddy weeding robot as an electric work vehicle. For example, the present invention can be applied to any vehicle such as an agricultural work vehicle such as a tractor, a rice transplanter, or a combiner, or an automobile.

  Further, the present invention is not limited to the above-described examples, and can take various forms within a range not departing from the gist of the invention.

  The paddy field weeding robot of the present invention is not limited to the paddy field weeding, and can be applied to an electric work vehicle for weeding any farmland in which crops are planted in a plurality of rows.

1 Paddy weeding robot (electric work vehicle)
DESCRIPTION OF SYMBOLS 10 Airframe 20 Traveling device 26 Traveling motor 30 Weeding device 40 Protective cover 50 Imaging device 60 Battery 65 Alarm device 70 Control part 80 Seedling (crop)

Claims (10)

A pair of left and right traveling devices;
A pair of left and right traveling motors for driving the respective traveling devices;
A weeding device for weeding,
A control unit storing an autonomous traveling program for autonomous traveling,
In an electric work vehicle that can be weeded autonomously,
An imaging device capable of imaging the periphery of the electric work vehicle,
The control unit identifies a crop based on color information of an image captured by the imaging device,
An electric work vehicle characterized in that weeds are autonomously driven between the crop streaks. A pair of left and right traveling devices;
A pair of left and right traveling motors for driving the respective traveling devices;
A weeding device for weeding,
A control unit storing an autonomous traveling program for autonomous traveling,
In an electric work vehicle that can be weeded autonomously,
An imaging device capable of imaging the periphery of the electric work vehicle,
The control unit identifies a travelable area based on color information of an image captured by the imaging device,
An electric work vehicle characterized in that weeding is carried out autonomously in the travelable area. A colorless and transparent protective cover for protecting the imaging device;
A removal device for removing deposits on the outer surface of the protective cover,
The imaging device is capable of imaging the periphery of the electric work vehicle through the protective cover,
3. The electric work vehicle according to claim 1, wherein the control unit identifies deposits on an outer surface of the protective cover based on color information of the image and activates the removing device. 4. The protective cover has a cylindrical shape, and the imaging device is disposed inside the protective cover.
The removing device includes a wiper that contacts the outer peripheral surface of the protective cover, and a rotating device that rotates the protective cover about a cylindrical axis.
The electric work vehicle according to claim 3, wherein the protective cover is rotated, and the adhered matter adhered to the outer peripheral surface of the protective cover is removed by the wiper.   The electric work vehicle according to any one of claims 1 to 4, wherein the control unit identifies a farmland boundary based on color information of the image and autonomously travels within the boundary.   The electric work vehicle according to any one of claims 1 to 5, further comprising an obstacle detection sensor that detects an obstacle around the electric work vehicle.   The electric working vehicle according to any one of claims 1 to 6, wherein the weeding device is a rake provided at a rear portion of the vehicle and is rotatable about a left-right direction as an axis. A load sensor for detecting a load applied to the rake;
The electric work vehicle according to claim 7, further comprising: a rotation device that rotates the rake about the left-right direction as an axis.   The electric work vehicle according to claim 1, further comprising an alarm device.   The electric work vehicle according to claim 1, further comprising a communication device.

Images