JP6586496B1 - Output control device and output control method - Google Patents

Abstract

An object of the present invention is to make it easy to see reflected light when reflected light reflected from a pallet by an irradiation device is difficult to see from a forklift driver. An output control device (1) includes a detection unit (103) for detecting a predetermined facial expression or motion from an image obtained by photographing a driver performing position alignment of a forklift using a learned model. And an output control unit (104) that causes the irradiation device (2) to change the mode of light irradiation when a predetermined facial expression or motion is detected. [Selection] Figure 1

Description

  The present invention relates to an output control device for controlling an irradiation device that irradiates light to a pallet from a forklift.

  In a cargo handling operation using a forklift, a forklift driver performs a series of operations of inserting a fork into a fork pocket of a pallet loaded with a load, lifting the load together with the pallet, and transporting the load. Of these series of operations, the difficulty of inserting a fork into the fork pocket is particularly high, and depending on the driver, it may take time to insert the fork or cause the fork to collide with the pallet.

  As a technique for facilitating the fork insertion operation, for example, Patent Document 1 is cited. In Patent Document 1, a pallet formed with a light reflecting surface that reflects a laser beam is irradiated with a laser beam having the same height as the fork, and the height alignment of the fork with respect to the pallet is performed using the reflected light as a mark. A technique for performing the above is disclosed.

JP 2005-298000 A

  According to the above technique, the height alignment of the fork with respect to the pallet becomes easy. However, it may be difficult to align the height of a pallet on which no light reflecting surface is formed. Further, when the pallet is arranged at a high place, it is assumed that the laser beam is difficult to see even if the light reflecting surface is formed.

  One aspect of the present invention provides an output control device and the like that makes it easy to see reflected light from a pallet when reflected light reflected from the pallet is difficult to see from a forklift driver. With the goal.

  In order to solve the above-described problem, an output control device according to an aspect of the present invention is configured such that irradiation light irradiated by an irradiation device installed on a forklift is applied to the pallet on which a load handling object of the forklift is placed. The forklift driver's facial expression and operation when the forklift is positioned with the reflected light reflected by the forklift as a mark, and when the reflected light is difficult to see Using a learned model generated by machine learning using an image in which the image is taken as teacher data, at least one of the facial expression and the motion is obtained from an image obtained by photographing the driver who wants to perform the height alignment. And detecting the detection unit and controlling the irradiation device when the detection unit detects at least one of the facial expression and the movement. It includes an output control section for changing the manner of light irradiation, the.

  In order to solve the above-described problem, an output control method according to an aspect of the present invention is an output control method using an output control device, in which a forklift is attached to a pallet on which a forklift object to be loaded is placed. This is an image of a situation in which the fork lift of the forklift is positioned using the reflected light reflected by the pallet as a mark, and the reflected light is difficult to see. Using the learned model generated by machine learning using teacher images as images of at least one of the facial expressions and movements of the forklift driver, we photographed the driver trying to perform the height alignment. A detection step of detecting at least one of the facial expression and the motion from the captured image, and the facial expression and When detecting at least one of the serial operation, including controlling the irradiation device and an output control step of changing the manner of light irradiation, the.

  According to one aspect of the present invention, when the reflected light reflected from the irradiating device by the pallet is difficult to see from the driver of the forklift, the reflected light from the pallet can be easily seen.

It is a block diagram which shows an example of a principal part structure of the output control apparatus which concerns on Embodiment 1 of this invention. It is a figure which shows the outline | summary of the output control by the said output control apparatus. It is a figure which shows an example of a normal irradiation aspect. It is a figure which shows the example which projected the image | video used as the mark of the fork alignment. It is a figure which shows the other example which projected the image | video used as the mark of the fork alignment. It is a flowchart which shows an example of the process which the said output control apparatus performs.

Embodiment 1
(Outline of output control)
An outline of output control by the output control apparatus of the present embodiment will be described with reference to FIG. FIG. 2 is a diagram showing an outline of output control by the output control device 1. As shown in FIG. 2, the side surface of the pallet P is irradiated with laser light (irradiation light) L from the irradiation device 2 installed on the forklift 3.

  The output control device 1 controls the irradiation device 2 to control the irradiation of the laser light L when the reflected light reflected by the side surface of the pallet P is difficult to see from the driver A driving the forklift 3. It is an apparatus that changes the irradiation mode.

  The irradiation device 2 is fixed to the forklift 3 at the same height as the tips of the claw portions of the pair of forks included in the forklift 3 and at an intermediate position between the claw portions (equal distance from the two claw portions). . Moreover, the irradiation apparatus 2 is being fixed so that it may raise / lower with the raising / lowering of a fork. The irradiation device 2 emits the laser beam L so as to spread horizontally toward the front of the forklift 3 and horizontally in a fan shape. That is, the irradiation apparatus 2 irradiates the laser beam L that spreads horizontally and linearly on the extension line of the claw portion (see the line LN in FIG. 3).

  In the example of FIG. 2, the driver A who operates the forklift 3 tries to align the height of the fork of the forklift 3 using the reflected light reflected by the pallet P as the laser beam L as a mark. The imaging device 4 captures the state of the driver A, and the captured image is input to the output control device 1. Note that the image to be captured may be a still image or a moving image.

  The output control device 1 acquires the image from the photographing device 4. When the driver A detects at least one of a predetermined facial expression and a predetermined operation indicating that the position of the line of the laser light L is difficult to see from the image, the line of the laser light L becomes easy to see from the driver A. Thus, the irradiation mode of the laser beam L is changed. Thereby, when the reflected light which the laser beam L from the irradiation apparatus 2 reflected on the palette P is hard to see from the driver | operator A, this reflected light can be made easy to see.

  In addition, the output control apparatus 1 may be comprised integrally with the irradiation apparatus 2, may be mounted in the forklift 3, and may be installed in the position away from the forklift 3. Moreover, the output control apparatus 1 is good also considering the several irradiation apparatus 2 as a control object. In this case, the output control device 1 acquires an image photographed with respect to the forklift 3 on which each irradiation device 2 is mounted, detects the above-mentioned predetermined facial expression and predetermined motion, and each irradiation device 2 according to the detection result. To control.

  The imaging device 4 may be provided exclusively for output control of the output control device 1, or an imaging device provided for other purposes may be used as the imaging device 4. For example, a monitoring camera for monitoring the inside of a facility where work by the forklift 3 is performed, and a camera for a drive recorder mounted on the forklift 3 may be used as the imaging device 4.

(Main components of the output control device)
The principal part structure of the output control apparatus 1 is demonstrated based on FIG. FIG. 1 is a block diagram illustrating an example of a main configuration of the output control device 1. As shown in the figure, the output control device 1 controls the control unit 10 that controls each part of the output control device 1, the storage unit 20 that stores various data used by the output control device 1, and the output control device 1. An input unit 30 that receives an input operation is provided. The control unit 10 includes a photographing control unit 101, an input data generation unit 102, a detection unit 103, and an output control unit 104. FIG. 1 also shows the irradiation device 2 and the photographing device 4 together.

  The photographing control unit 101 instructs the photographing device 4 to start photographing and finish photographing. As described above, the object to be photographed is the state of the driver A who is trying to perform the height alignment of the fork using the reflected light reflected by the pallet P as the laser beam L, and more specifically, the height alignment. This is at least one of the facial expression and the motion of the driver A who is going to perform. In the following, for the sake of simplicity, an example in which the facial expression (that is, the face) of the driver A is photographed will be described.

  The input data generation unit 102 generates input data for a learned model, which will be described later, from an image obtained by capturing at least one of the facial expression and movement of the driver A who is trying to align the height of the fork. As described above, since it is assumed here that the face of the driver A is photographed, the input data generation unit 102 generates input data from an image obtained by photographing the face of the driver A.

  The detection unit 103 detects at least one of a predetermined facial expression and a predetermined action from an image obtained by photographing the state of the driver A who is going to perform height alignment using the learned model. Predetermined motion is when the driver (driver A or other driver) of the forklift (forklift 3 or other forklift) tried to align the height of the fork in the past, but the reflected light was difficult to see The operation performed by the driver. Specific examples of the predetermined operation include an operation of changing the angle at which the reflected light of the laser beam is looked up (or looking down), an operation of repeatedly raising and lowering the fork, and an operation of holding a hand over the eyes. Similarly, the predetermined expression is the expression of the driver when the driver of the forklift tried to align the height of the fork in the past, but the reflected light was difficult to see. Specific examples of the predetermined facial expression include, for example, an expression that narrows the eyes so that it is difficult to see, an irritated expression, and the like. In the present embodiment, an example will be described in which the detection unit 103 detects that the driver A has the predetermined facial expression.

  When the detection unit 103 detects the predetermined facial expression, the output control unit 104 controls the irradiation device 2 to change the irradiation mode of the laser light L. Of course, the output control unit 104 performs the same processing when the detection unit 103 detects the predetermined operation. An example of the irradiation mode of the laser light L will be described later.

(About detection of learned model and detector)
The learned model used by the detection unit 103 is a learned model generated by machine learning using teacher data. As this teacher data, for example, the driver of the forklift tried to align the height of the fork in the past, but the driver's facial expression when the reflected light was difficult to see (predetermined facial expression described above) was photographed. Images can be used. In addition, the forklift 3 shown in FIG. 2 may be sufficient as the said forklift, and another forklift may be sufficient as it. Further, the driver may be the driver A shown in FIG. 2 or another driver.

  By using the teacher data as described above, for example, by optimizing the parameters of a convolutional neural network (CNN), a learned model for detecting the predetermined facial expression from an image can be generated. . The learned model is not limited to CNN. However, when a predetermined facial expression is detected from an image as in the present embodiment, CNN that can realize high detection accuracy is preferable.

  When input data generated from an image of the driver A is input to the learned data generated in this manner, output data indicating the probability that the driver A in the image is the predetermined facial expression is displayed. Is output. If the probability indicated in the output data output from the learned model is equal to or greater than a predetermined threshold, the detection unit 103 detects that the driver A is the predetermined facial expression.

  As the teacher data, for example, the driver's operation when the forklift driver tried to align the height of the fork in the past but the reflected light was difficult to see (predetermined operation described above). A photographed image can also be used. As a result, a learned model for detecting the predetermined motion can be generated from the image. In this case, a moving image obtained by photographing the predetermined operation may be used as teacher data, and a still image (frame image) extracted from the moving image may be used as teacher data.

  When detecting the predetermined operation, it is preferable to use, for example, a learned model that combines CNN and RNN (Recurrent Neural Network) or CNN and LSTM (Long Short-Term Memory) rather than using CNN alone. . By combining a model suitable for handling time-series data such as RNN and LSTM and CNN having high image recognition performance, highly reliable output data can be obtained.

(About input data)
The input data generation unit 102 generates input data from an image obtained by photographing how the driver A aligns the fork of the forklift 3. For example, when the image photographed by the photographing device 4 is a moving image, the input data generation unit 102 may extract a frame image from the moving image at a predetermined time interval, and may use each extracted frame image as input data.

  The image used for generating the input data is preferably a photograph taken under the same photographing conditions as the image used as the teacher data in the machine learning of the learned model. The photographing conditions include, for example, the positional relationship between the photographing device 4 and the driver A who is the subject, the background of the driver A, the ambient brightness, and the like.

  The input data generation unit 102 may use the image as input data as it is, or may use an image obtained by performing a predetermined process on the image. The predetermined process may be any process that can reduce information unrelated to the characteristic without losing the characteristic of the predetermined facial expression. For example, if the image is a color image, the input data may be converted to gray scale.

  Furthermore, the input data generation unit 102 may detect the driver A and its face portion from, for example, an object shown in the image, and generate input data from the detected face portion. Thereby, since the influence of the background of the driver A can be eliminated, the detection accuracy can be increased. For detection of the driver A, for example, a known image analysis technique, a learned model such as CNN, or the like can be used. However, the input data generation unit 102 needs to generate input data in the same format as the teacher data.

(Example of laser light irradiation mode)
The example of the irradiation mode of the laser beam L by the irradiation apparatus 2 is demonstrated based on FIGS. FIG. 3 is a diagram illustrating an example of a normal irradiation mode. FIG. 4 is a diagram illustrating an example of the irradiation mode after the change. FIG. 5 is a diagram illustrating another example of the irradiation mode after the change.

  As described above, the irradiation device 2 emits the laser light L on the extension line of the claw portion of the forklift 3 so as to spread horizontally in a fan shape. Therefore, in a state where the pallet P is positioned on the extended line of the claw portion of the forklift 3, as shown in FIG. 3, the laser beam L is reflected by the side surface of the pallet P, and the reflected light line LN is formed on the side surface. Appears. In addition, the said side surface is a side surface with which the forklift 3 is facing directly among the side surfaces of the pallet P, and is a side surface in which one set of fork pockets FP which is an opening part for inserting a fork is opened. .

  The height of the line LN is equal to the height of the claw portion of the forklift 3. For this reason, the driver A raises and lowers the fork so that the position of the line LN is a position that passes through the center of the fork pocket FP in the height direction. Thereby, the fork can be aligned with the fork pocket FP.

  The output control unit 104 changes the irradiation mode of the irradiation device 2 from the normal irradiation mode as shown in FIG. 3 to another irradiation mode. As a result, it is possible to improve the situation in which it is difficult for the driver A to visually recognize the line LN. The irradiation mode after the change may be a mode that can improve the above situation. For example, the output control unit 104 performs irradiation control by increasing the output of the irradiation device 2 to increase the brightness of the line LN, causing the irradiation device 2 to perform light irradiation intermittently, and blinking the line LN. You may change an aspect. In addition, for example, the output control unit 104 inserts an optical member such as a lens in the optical path of the laser light L so that the laser light L spreads in the height direction, and thereby the height direction of the line LN. The width of may be widened.

  In addition to the laser beam L, or instead of the laser beam L, an image that serves as a mark for fork alignment is projected onto the pallet P or the like, thereby improving the situation in which it is difficult for the driver A to visually recognize the line LN. Also good. 4 and 5 are diagrams illustrating an example in which an image serving as a mark for fork alignment is projected. Note that the irradiation apparatus 2 also performs control to end irradiation in the normal mode by the irradiation apparatus 2 and project an image instead, and control to further project another image while continuing irradiation in the normal mode by the irradiation apparatus 2. It is an example of the control which changes the aspect of light irradiation by.

  In the example of FIG. 4, a line LN appears on the pallet P and a rectangular frame FM centered on the line LN is projected as shown in FIG. As shown in FIGS. 4A to 4D, the size of the rectangular frame FM is reduced in time series. Accordingly, the driver A visually recognizes the rectangular frame FM so as to converge toward the line LN. After the state (d), the state may be returned to the state (a) again, and the size of the rectangular frame FM may be reduced in time series, or the projection of the rectangular frame FM may be terminated.

  Since the rectangular frame FM is projected in a wider range than the line LN, it is easily visible to the driver A, and since the rectangular frame FM converges toward the line LN, the rectangular frame FM is projected from the projected rectangular frame FM to the line LN. It is also easy for the driver to recognize the position of. Therefore, by performing the light irradiation as shown in FIG. 4, the driver A can easily grasp the position of the line LN and smoothly align the fork. Further, the same effect can be obtained by changing the thickness of the line LN in time series. The frame line to be projected is not limited to a rectangle, and may be a frame of another shape such as an ellipse, or may not be a complete frame (that is, a part of the frame may be missing). .

  On the other hand, in the example of FIG. 5, the image of the character is projected instead of the line LN to indicate the height of the fork. In the state of FIG. 5A, a face image FC, which is an example of a character image, is projected onto a load placed on the pallet P. The face image FC includes three parts, a right eye FC1, a left eye FC2, and a mouth FC3, and the center positions of the right eye FC1 and the left eye FC2 in the height direction correspond to the height of the tip of the fork.

  That is, in the state of FIG. 5 (a), the tip of the fork is located above the fork pocket FP, and the driver A can see the tip of the fork from the fork pocket FP from the face image FC projected on the luggage. Can also be recognized as being located above. Therefore, the driver A operates the forklift 3 to lower the fork, and as shown in FIG. 5B, the driver A moves the fork so that the right eye FC1 and the left eye FC2 of the face image FC overlap the position of the fork pocket FP. The position is adjusted, thereby completing the fork alignment.

  Since the face image FC is projected in a wider range than the line LN, it is easily visible to the driver A. Further, if the driver A recognizes that the projected image is the face image FC, the driver A can recognize from the position if the driver A can visually recognize at least a part of the face image FC (for example, the part of the mouth FC3). The positions of the right eye FC1 and the left eye FC2 (that is, the height of the tip of the fork) can be estimated. Therefore, by performing the light irradiation as shown in FIG. 5, the driver A can easily grasp the height of the tip of the fork and can smoothly align the fork. Furthermore, the driver A can also confirm from the face image FC whether the fork is displaced in the left-right direction with respect to the fork pocket FP.

  In addition, when projecting an image as shown in FIGS. 4 and 5, for example, an image projection apparatus may be mounted on a forklift as another irradiation apparatus, separately from the irradiation apparatus 2. Then, the output control unit 104 may change the irradiation mode using only the irradiation device 2 to the irradiation mode using the video projection device by controlling the video projection device. In addition, the irradiation apparatus 2 may be configured to project various images, and the above-described images may be projected.

  Of course, the above-described mode of light irradiation is only an example. For example, different colors of light may be projected onto the upper region and the lower region of the fork tip. Further, the color to be projected may be changed stepwise or continuously according to the distance from the height of the tip of the fork. By irradiating light in such a manner, the driver A can easily recognize the projected image, easily grasp the height of the tip of the fork, and smoothly align the fork. It can be performed.

(Process flow)
The flow of processing executed by the output control device 1 will be described with reference to FIG. FIG. 6 is a flowchart illustrating an example of processing (output control method) executed by the output control device 1.

  In S <b> 1, the imaging control unit 101 determines whether the output control unit 104 has caused the irradiation apparatus 2 to start irradiation with the laser light L. If it is determined that the irradiation is started (YES in S1), the process proceeds to S2. If it is determined that the irradiation is not started (NO in S1), the determination in S1 is repeated.

  In S <b> 2, the photographing control unit 101 instructs the photographing device 4 to start photographing. Then, the input data generation unit 102 acquires an image from the imaging device 4 that has started imaging in accordance with the above instruction, and generates input data from the image. As described above, here, the imaging device 4 captures a moving image, and the input data generation unit 102 uses a frame image extracted from the captured moving image as input data. Note that the process of generating input data is continuously performed until it is determined in S5, which will be described later, that photographing has been completed.

  In S3 (detection step), the detection unit 103 detects that the driver A has a predetermined facial expression using the learned model. Specifically, the detection unit 103 inputs the input data generated in S2 to the learned model, and if the probability indicated in the output data output from the learned model is equal to or greater than a predetermined threshold, the driver A Is detected as the predetermined expression. If a predetermined facial expression is detected (YES in S3), the process proceeds to S6, and if not detected (NO in S3), the process proceeds to S4.

  In S <b> 4, the imaging control unit 101 determines whether the output control unit 104 has finished irradiating the irradiation apparatus 2 with the laser light L. If it is determined that the irradiation has ended (YES in S4), the process proceeds to S5. If it is determined that the irradiation has not ended (NO in S4), the process returns to S3. When the process returns to S3, the process of S3 is performed using the input data generated next to the input data used in the previous S3.

  In S <b> 5, the shooting control unit 101 instructs the shooting device 4 to end shooting. Thereby, the generation of the input data in S2 is also ended, and the processing in FIG. 6 is ended.

  In S6 (output control step), the output control unit 104 causes the irradiation device 2 to change the irradiation mode of the laser light L. After the end of S6, the process returns to S3. If YES in S3, the process of S6 is performed again. However, in the second process of S6, it is preferable to change the irradiation mode after the first change to a different irradiation mode. Thus, even when the irradiation mode is changed, when the driver A is difficult to recognize the reflected light of the laser light L, the irradiation mode can be further changed to make it easier to recognize the reflected light.

  Regarding S1 and S4 described above, the trigger for the output control unit 104 to start the irradiation of the laser light L in the irradiation device 2 and the trigger to end the irradiation are not particularly limited. For example, the irradiation may be started when an operation for starting the irradiation of the laser beam L is performed on the input unit 30. The same applies to the end of irradiation. In addition, for example, if the fork tilt changes horizontally from a forward tilt (a state where the tip is higher than the base), the fork may be inserted immediately after that, so that the change is detected. Irradiation may be started as an opportunity. Further, since the irradiation is no longer necessary when the operation for inserting the fork into the fork pocket FP is started, the irradiation may be terminated when the start of the operation is detected.

(Summary of Embodiment 1)
As described above, in the learned model used in the present embodiment, the irradiation light irradiated by the irradiation device 2 installed on the forklift 3 is applied to the pallet P on which the cargo handling object of the forklift 3 is placed. FIG. 6 is an image of a situation where the fork height of the forklift 3 is aligned using the reflected light reflected by the forklift 3 as a mark, and at least the expression and operation of the driver A of the forklift 3 when the reflected light is difficult to see It is generated by machine learning using an image in which any of the images is captured as teacher data.

  In addition, the detection unit 103 detects at least one of the facial expression and the motion from an image obtained by photographing the state of the driver A who wants to perform the height alignment using the learned model. And the output control part 104 controls the irradiation apparatus 2 and changes the aspect of light irradiation, when the detection part 103 detects at least any one of the said facial expression and the said operation | movement.

  As described above, according to the output control device 1, when at least one of the facial expression and the operation of the driver A when the reflected light from the pallet P is difficult to see, the light irradiation mode is changed. Therefore, when it is difficult for the driver A to see the reflected light from the pallet P, the reflected light from the pallet P is automatically made easy to see, and the driver A can perform the height alignment operation smoothly.

[Embodiment 2]
Embodiment 2 of the present invention will be described below. For convenience of explanation, members having the same functions as those described in the first embodiment are given the same reference numerals, and the description thereof will not be repeated.

  In the output control device 1 of the present embodiment, the detection unit 103 sets a plurality of types of detection targets as at least one of the above-described predetermined action and the above-described predetermined facial expression. And the output control part 104 controls the irradiation apparatus 2 so that light irradiation may be performed in the aspect according to the said type which the detection part 103 detected.

  For example, when a plurality of types of predetermined facial expressions are to be detected, a learned model generated by machine learning using teacher data for each facial expression to be detected may be used. The detection unit 103 can detect each type of facial expression by inputting the input data generated by the input data generation unit 102 to the learned model. The same applies to a case where a plurality of types of predetermined operations are to be detected.

  Further, the predetermined facial expression and the predetermined motion may be detected respectively. In this case, a learned model for detecting a predetermined facial expression and a learned model for detecting a predetermined motion may be prepared. Then, the input data generation unit 102 may generate input data to be input to the learned model for detecting a predetermined facial expression and input data to be input to the learned model for detecting a predetermined action. Thereby, the detection part 103 can each detect a predetermined facial expression and a predetermined operation | movement based on the output data of each learned model.

  A plurality of patterns are conceivable in a situation in which the reflected light from the palette P is difficult to be seen by the driver A with respect to the detected facial expression or the mode according to the type of action. As this pattern, for example, a pattern in which the driver A recognizes the approximate position of the line LN of reflected light but is difficult to visually recognize the line LN due to the color and brightness of the line LN (hereinafter referred to as a pattern a). Described). In addition, as another pattern, for example, a pattern in which the driver A cannot recognize even the position of the line LN of reflected light (hereinafter referred to as a pattern b) can be cited.

  Examples of facial expressions corresponding to the pattern a include facial expressions that make it difficult to see laser light such as narrowing the eyes. Examples of the operation corresponding to the pattern a include an operation of holding a hand over the eye and an operation of repeating blinking. As the operation corresponding to the pattern b, for example, there is an operation of changing the angle (face direction) at which the line LN is looked up or looked down.

  The output control unit 104 may control the irradiation apparatus 2 so that light irradiation is performed in a manner according to such a pattern. For example, the output control unit 104 may control the irradiation device 2 so as to increase the visibility of the line LN when an expression or motion of the pattern a is detected. Specifically, the output control unit 104 may execute at least one of control for increasing the luminance of the laser light L output from the irradiation device 2 and control for changing the color of the laser light L. On the other hand, the output control unit 104 detects, for example, the width of the laser light L output from the irradiation device 2 so that the driver A can easily recognize the position of the line LN when the expression or action of the pattern b is detected. Control may be performed to widen the range of light irradiation such as widening.

(Summary of Embodiment 2)
As described above, since there are various patterns in the difficulty of seeing the reflected light from the palette P, it is preferable to perform the light irradiation in a manner according to how difficult it is to see. Therefore, as described above, the detection unit 103 according to the present embodiment uses a plurality of types of at least one of facial expressions and actions as detection targets. And the output control part 104 controls the irradiation apparatus 2 so that light irradiation may be performed in the aspect according to the kind which the detection part 103 detected. Thereby, the light irradiation can be performed in a manner in which the driver A can more easily visually recognize the reflected light from the pallet according to the pattern of difficulty in seeing the reflected light from the pallet P.

[About distributed processing]
A part of the processing executed by the output control device 1 described in each of the above embodiments may be executed by one or a plurality of devices connected to the output control device 1 in communication. For example, among the processes executed by the detection unit 103, an arithmetic process using a learned model may be executed by an AI server that is connected to the output control apparatus 1 in communication. In this case, the output control device 1 generates input data, transmits it to the AI server, receives the output data from the AI server, and detects at least one of the predetermined facial expression and motion of the driver A.

[Example of software implementation]
The control block of the output control apparatus 1 (particularly each unit included in the control unit 10) may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or may be realized by software. Good.

  In the latter case, the output control apparatus 1 includes a computer that executes instructions of a program that is software that realizes each function. The computer includes, for example, one or more processors and a computer-readable recording medium storing the program. In the computer, the processor reads the program from the recording medium and executes the program, thereby achieving the object of the present invention. For example, a CPU (Central Processing Unit) or a GPU (Graphics Processing Unit) can be used as the processor. As the recording medium, a “non-temporary tangible medium” such as a ROM (Read Only Memory), a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like can be used. Further, a RAM (Random Access Memory) for expanding the program may be further provided. The program may be supplied to the computer via an arbitrary transmission medium (such as a communication network or a broadcast wave) that can transmit the program. Note that one embodiment of the present invention can also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

  The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope shown in the claims, and embodiments obtained by appropriately combining technical means disclosed in different embodiments. Is also included in the technical scope of the present invention.

1 Output Control Device 2 Irradiation Device 3 Forklift A Driver L Laser Light (Irradiation Light)
P Pallet 103 Detection unit 104 Output control unit

Claims (2)

A state in which the fork height of the forklift is aligned using the reflected light reflected from the pallet of the irradiation light applied to the pallet on which the object to be handled by the forklift is placed. Using a learned model generated by machine learning using a captured image, which is an image showing at least one of the expression and movement of the forklift driver when the reflected light is difficult to see A detection unit for detecting at least one of the facial expression and the movement from an image obtained by photographing a driver who wants to perform the height alignment;
An output control unit that controls the irradiation device to change the mode of light irradiation when the detection unit detects at least one of the facial expression and the movement ; and
The detection unit has at least one of the facial expression and the motion as a plurality of types of detection targets,
The output control unit, the output control device, characterized that you control the illumination device so that the light irradiation is performed in a manner that depends on the type of the detection unit has detected. An output control method by an output control device,
A state in which the fork height of the forklift is aligned using the reflected light reflected from the pallet of the irradiation light applied to the pallet on which the object to be handled by the forklift is placed. Using a learned model generated by machine learning using a captured image, which is an image showing at least one of the expression and movement of the forklift driver when the reflected light is difficult to see A detection step of detecting at least one of the facial expression and the movement from an image obtained by photographing a driver who wants to perform the height alignment;
When detecting at least one of the expression and the operation in the detecting step, seen including an output control step of changing the manner of light irradiation by controlling the irradiation device, and
In the detection step, a plurality of types of at least one of the facial expression and the motion are detected,
In the output control step, the output control method is characterized in that the irradiation device is controlled so that light irradiation is performed in a manner corresponding to the type detected in the detection step .

Images