Person recognition device and garbage collection vehicle

Abstract

To provide a person recognition device that can more reliably determine the intrusion of a person into the area near the garbage collecting vehicle.SOLUTION: A person recognition device comprises an image processor for recognizing a person in the vicinity of a garbage inlet 4. The image processor includes person determination means capable of performing: an upper body determination process of comparing an object image in an image with feature data of a head portion to determine whether or not the object image represents the head portion; and a lower body determination process of comparing the object image in the image with feature data of a leg portion to determine whether or not the object image represents the leg portion. The person determination means also includes intrusion determination means that determines whether or not the object image determined by the person determination means as representing the head portion or leg portion is in a predetermined area near the garbage inlet 4. The person determination mean is configured to: when the upper body determination process is executed to result in that the object image is determined not to represent the head portion, execute the lower body determination process; or otherwise, when the lower body determination process is executed to result in that the object image is determined not to represent the leg portion, execute the upper body determination process.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a person recognition device provided in a garbage collection vehicle for recognizing a person near a garbage inlet.

Background Art

[0002] Conventionally, in a garbage collection vehicle, a garbage loading device is equipped in a garbage inlet box provided at the rear of the vehicle body, and the garbage introduced into the garbage inlet box from the garbage inlet of the garbage inlet box is scraped by a rotating plate, a loading plate, etc. and loaded into the garbage storage box. In addition, from the viewpoint of safety of preventing an operator or the like who throws garbage from being accidentally卷入 into the garbage loading device, a person recognition device for determining the presence or absence of a person's intrusion into a predetermined area near the garbage inlet has been proposed (for example, Patent Document 1).

[0003] As in the case of the person recognition device of Patent Document 1, in order to prevent problems such as sunlight entering the imaging lens and reducing the contrast of the image, or rain or snow adhering to the imaging lens and blocking the field of view, the camera is installed above the garbage inlet of the garbage collection vehicle, and the optical axis of the imaging lens is set downward. Therefore, the person recognition device performs person recognition based on an image of the head of a person that is easy to photograph during the garbage loading operation in relation to the installation position of the camera.

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0005] By the way, in situations such as loading refuse onto a refuse collection truck from a raised platform at a logistics hub, where the height of the truck bed is set to match the height of the refuse bed, the position where the workers loading the refuse stand is not necessarily the same as the ground contact surface of the refuse collection truck.

[0006] Therefore, with conventional person recognition devices, there was a problem in that when loading garbage while standing on a raised platform higher than the ground contact surface of the garbage truck, the person's head would sometimes extend beyond the camera's field of view due to the camera's position, making it impossible to recognize them as a person.

[0007] The present invention has been made in view of the above problems, and aims to provide a person recognition device and a garbage collection vehicle equipped therewith that enable more reliable recognition of people and thereby more reliable determination of whether or not a person has entered a predetermined area near the garbage collection port. [Means for solving the problem]

[0008] A person recognition device according to a first aspect of the present invention is a person recognition device comprising a camera installed above the garbage inlet of a garbage collection vehicle and an image processing device that recognizes a person near the garbage inlet based on image data captured by the camera, wherein the image processing device includes a person determination means capable of performing an upper body determination process, which compares an object image contained in the image with predetermined upper body feature data of a person to determine whether the object image represents the upper body, and a lower body determination process, which compares an object image contained in the image with predetermined lower body feature data of a person to determine whether the object image represents the lower body, and an intrusion determination means that determines whether an object image determined by the person determination means to represent the upper body or the lower body is in a predetermined area near the garbage inlet, wherein the person determination means performs the lower body determination process when it has performed the upper body determination process and determined that the object image does not represent the upper body, or performs the upper body determination process when it has performed the lower body determination process and determined that the object image does not represent the lower body.

[0009] The terms "upper body" and "lower body" are relative. When two or more regions of the body are identified, the part located relatively above is the "upper body," and the part located relatively below is the "lower body." The same applies hereafter.

[0010] In the garbage loading operation, which is performed while standing on a predetermined step that is higher than the ground contact surface of the garbage truck, the upper part of a person's body (for example, their head) may extend beyond the camera's field of view due to the camera's mounting position. Even in such cases, the person recognition device according to the first embodiment can determine whether or not the object image represents a person by comparing the object image contained in the image with predetermined feature data of the lower part of a person's body (for example, their legs) using the person determination means of the image processing device.

[0011] On the other hand, in the loading of refuse while standing on the road surface or the ground where the refuse collection vehicle is planted, the lower part of a person's body (for example, their legs) may be hidden by the upper part of their body or the shadow of the refuse collection vehicle due to the camera's mounting position. Even in such cases, the person recognition device according to the first embodiment can determine whether the object image represents a person or not by comparing the object image included in the image with predetermined feature data of the upper part of a person's body (for example, their head) using the person determination means of the image processing device.

[0012] Of course, even if both the upper and lower parts of a person's body are photographed, it is still possible to determine whether or not the object image represents a person. An object image determined to represent either the upper or lower part of a body is used by the intrusion detection means of the image processing device to determine whether or not it has entered a predetermined area near the waste inlet.

[0013] Therefore, the person recognition device according to the first embodiment performs person recognition more reliably compared to conventional person recognition devices that perform person recognition based on an image of the person's head, thereby making it more reliable to determine whether or not a person has entered a predetermined area near the waste disposal opening.

[0014] A person recognition device according to a second aspect of the present invention is a person recognition device comprising a camera installed above the refuse loading port of a refuse collection vehicle, and an image processing device that recognizes a person near the refuse loading port based on image data captured by the camera, wherein the image processing device includes: work determination means for determining whether or not a step-up operation is being performed, in which refuse is loaded from a predetermined step-up platform higher than the ground surface of the refuse collection vehicle; if the work determination means determines that the step-up operation is not being performed, a person determination means for determining whether or not the object image included in the image represents the upper part of the body by comparing it with predetermined upper body feature data of a person; if the work determination means determines that the step-up operation is being performed, a person determination means for determining whether or not the object image included in the image represents the lower part of the body by comparing it with predetermined lower body feature data of a person; and intrusion determination means for determining whether or not the object image determined by the person determination means to represent the upper part of the body or the lower part of the body is in a predetermined area near the refuse loading port.

[0015] In the person recognition device according to the second embodiment, the work determination means of the person recognition device determines whether or not a step-up operation is being performed, in which the garbage is loaded from a predetermined step that is higher than the ground surface of the garbage collection vehicle. If the work determination means determines that no step-up operation is being performed, that is, in situations where the upper part of a person's body (for example, the head) is likely to be photographed in relation to the camera's installation position, the person determination means of the image processing device compares the object image included in the image with predetermined feature data of the upper part of a person's body (for example, the head) to determine whether or not the object image represents the upper part of the body (for example, the head).

[0016] On the other hand, if the work determination means of the person recognition device determines that work is being performed on steps, that is, in situations where the upper part of a person's body (for example, the head) is relatively often outside the camera's shooting range in relation to the camera's installation position, the person determination means of the image processing device compares the object image included in the image with predetermined feature data of the lower part of a person's body (for example, the legs) to determine whether or not the object image represents the lower part of the body (for example, the legs).

[0017] The object image determined to represent the upper or lower part of the body is then used by the intrusion detection means of the image processing device to determine whether or not it has entered a predetermined area near the waste inlet.

[0018] Therefore, the person recognition device according to the second embodiment performs person recognition more reliably compared to conventional person recognition devices that perform person recognition based on an image of the person's head, thereby making it more reliable to determine whether or not a person has entered a predetermined area near the waste disposal opening. Furthermore, since the feature data of the object image included in the image only needs to be compared with either predetermined feature data of the upper body or feature data of the lower body of the person, the processing load required for person recognition is also reduced compared to the person recognition device according to the first embodiment.

[0019] A person recognition device according to a third aspect of the present invention is characterized in that, in a person recognition device according to a first or second aspect, the feature data of the upper body includes the size, position, and shape of an object attached to the upper body, and the feature data of the lower body includes the size, position, and shape of an object attached to the lower body.

[0020] The person position determination unit according to the fourth aspect of the present invention is characterized in that, in the person position determination unit according to the third aspect, the upper body feature data is obtained by machine learning of a plurality of predetermined images of the upper body of a person standing on the ground surface of the garbage collection vehicle, and the lower body feature data is obtained by machine learning of a plurality of predetermined images of the lower body of a person standing at a height position where the upper body is not photographed.

[0021] A fifth aspect of the present invention is a person recognition device according to the fourth aspect, characterized in that the upper part of the body is the head and the lower part of the body is the legs.

[0022] A person recognition device according to a sixth aspect of the present invention is a person recognition device according to a second aspect, wherein the work determination means compares an object image included in the image with the feature data of the step area to determine whether or not the object image represents the step area, and if it is determined that the object image does not represent the step area, it is determined that the step work has not been performed, and if it is determined that the object image represents the step area, it is determined that the step work has been performed.

[0023] A person recognition device according to a seventh aspect of the present invention further comprises, in a person recognition device according to a second aspect, an approach signal transmitter provided on the step area that transmits a predetermined approach signal, and an approach signal detection device provided on the garbage collection vehicle that detects the approach signal when the distance to the approach signal transmitter is within a predetermined value, wherein the work determination means of the image processing device determines that the step work is not being performed if the approach signal detection device does not detect the approach signal, and determines that the step work is being performed if the approach signal detection device has received the approach signal.

[0024] The person recognition device according to the eighth aspect of the present invention is the person recognition device according to the second aspect, further comprising a position information acquisition device that acquires the position information of the garbage collection vehicle using a satellite positioning system. The work determination means of the image processing device refers to the position information acquired by the position information acquisition device, and determines that the step work is not being performed when the garbage collection vehicle is not at a predetermined position, and determines that the step work is being performed when the garbage collection vehicle is at a predetermined position. This is the gist of the invention.

[0025] The ninth aspect of the present invention is a garbage collection vehicle equipped with the person recognition device according to the first or second aspect.

Effects of the Invention

[0026] According to the person recognition device according to the present invention or the garbage collection vehicle equipped with the same, the recognition of a person is more reliably performed than in the past, and thus the determination of the presence or absence of a person's intrusion into a predetermined area near the garbage inlet is more reliably performed.

Brief Description of the Drawings

[0027] [Figure 1] It is a side view of a garbage collection vehicle to which the person recognition device according to the first embodiment of the present invention is applied. [Figure 2] It is a rear view of the garbage inlet box portion of the garbage collection vehicle in FIG. 1. [Figure 3] It is an operation explanatory diagram of the garbage loading device provided in the garbage collection vehicle in FIG. 1. [Figure 4] It is a front view of the camera unit provided in the garbage collection vehicle in FIG. 1 (viewed in the direction of arrow X1 in FIG. 3). [Figure 5] It is a diagram showing the hydraulic circuit of the garbage loading device provided in the garbage collection vehicle in FIG. 1. [Figure 6] It is a diagram showing a part of the electric circuit used for various controls of the garbage collection vehicle in FIG. 1. [Figure 7] It is a view of an operator loading garbage (garbage bags) into the garbage inlet box on the road as seen from the side. [Figure 8] It is a diagram showing an example of an image taken by the camera unit in FIG. 4. [Figure 9] This is a side view of a worker loading garbage bags into a garbage disposal bin on a platform. [Figure 10] This figure shows another example of an image captured by the camera unit shown in Figure 4. [Figure 11] This is a flowchart illustrating the processing flow performed by the control device installed in the refuse collection vehicle shown in Figure 1. [Figure 12] This is a flowchart illustrating the processing flow performed by the person recognition device according to the first embodiment of the present invention. [Figure 13] Figure 12 is a flowchart showing the flow of the first person recognition process. [Figure 14] Figure 12 is a flowchart showing the flow of the second person recognition process. [Figure 15] This is a flowchart illustrating the processing flow performed by the person recognition device according to the second embodiment of the present invention. [Figure 16] Figure 15 is a flowchart showing the workflow for determining the work location. [Figure 17] Figure 15 is a flowchart showing the flow of the third person recognition process. [Figure 18] Figure 15 is a flowchart showing the flow of the fourth person recognition process. [Figure 19] This flowchart shows the workflow for determining a work location, which is performed by the person recognition device according to the third embodiment of the present invention. [Figure 20] This flowchart shows the workflow for determining a work location, which is performed by the person recognition device according to the fourth embodiment of the present invention. [Modes for carrying out the invention]

[0028] Hereinafter, a person recognition device according to the first embodiment of the present invention will be described with reference to the attached drawings. The person recognition device according to the first embodiment is applied to the garbage collection vehicle 100 shown in Figure 1. In the following description, front, back, left, and right are based on the view from the direction of travel of the garbage collection vehicle.

[0029] As shown in Figure 1, the refuse collection vehicle 100 is equipped with a refuse storage box 2 and a refuse input box 3 on a chassis 1. The rear opening of the refuse storage box 2 and the front opening of the refuse input box 3 are in communication with each other. The refuse input box 3 is pivotally supported relative to the refuse storage box 2 by a left-right hinge 3a provided on its upper part, and is tiltable by a pair of left and right tilting cylinders (not shown).

[0030] A roughly rectangular refuse inlet 4 for loading refuse is provided at the lower part of the rear of the refuse inlet 3, and the refuse inlet 4 is opened and closed by a vertically movable tailgate 5. A switch box 6 for operating the refuse loading device 31 (Figure 3) is provided to the left of the refuse inlet 4. A camera unit 7 is installed above the refuse inlet 4 to photograph the refuse inlet 4 and the surrounding area. The camera unit 7 is fixed to the top of the refuse inlet 3 via a support member 80 that protrudes from the front to the rear of the inclined rear surface (curved rear surface) 3b of the refuse inlet 3, which is set above the refuse inlet 4.

[0031] To explain in more detail, the support member 80 is frame-shaped with base ends at both the left and right ends of the inclined rear surface portion 3b, as shown in Figures 2 to 4. The camera unit 7 is fixed to the rear end and center in the left-right direction of the support member 80 and includes a camera 71, an operating lamp 72, and a detection lamp 73.

[0032] The camera 71 is mounted via a mounting bracket 75 in a recess 74 located in the left-right center of the camera unit 7, and its orientation is set so that the optical axis of the imaging lens 71a points diagonally downward. The camera 71 is rotatable up and down around the horizontal axis via the mounting bracket 75. On both the left and right sides of the recess 74 of the camera unit 7, there are fixed surfaces 7a that are angled upward and backward relative to the horizontal plane (tilting upward as it approaches the rear of the vehicle). An operating lamp 72 is mounted facing downward on the left fixed surface 7a. A detection lamp 73 is mounted facing downward on the right fixed surface 7a.

[0033] As shown in Figure 3, the waste input box 3 is equipped with a waste loading device 31 that loads the waste G that is input into the waste storage box 2. This waste loading device 31 pushes the waste that is input into the waste input box 3 into the waste storage box 2, which is connected to the front of the waste input box 3. The waste storage box 2 is equipped with a waste discharge device (not shown) that discharges the contained waste.

[0034] The refuse loading device 31 is a so-called rotary plate type refuse loading device that scoops up refuse placed in the refuse input box 3 by the rotation of the rotary plate 10 and pushes it into the refuse storage box 2 by the pushing plate 20. A rotating shaft 11 is installed in the lower part of the refuse input box 3 so as to extend in the width direction, and the base end of the rotary plate 10 is fixed to it. In the illustrated example, a hydraulic motor 13 capable of forward and reverse rotation is connected to the end of the rotating shaft 11 via a reduction mechanism 12. The rotation of the hydraulic motor 13 is torque-up by the reduction mechanism 12 and transmitted to the rotating shaft 11, and the rotary plate 10 rotates together with the rotating shaft 11. As the rotary plate 10 rotates, the tip of the rotary plate 10 moves in the front-rear direction along the bottom wall of the refuse input box 3, which is formed in a roughly semi-circular cross-section.

[0035] On the other hand, the push plate 20 is provided above the rotating plate 10, extending across the entire width of the waste collection box 3, and is supported so as to be able to swing back and forth around a left-right swing axis 21 located at its upper part. A support pin 23 is provided inside the waste collection box 3, above the swing axis 21 and on the front side. The push plate 20 is provided with an extension portion 22 that extends above the swing axis 21, and a push cylinder 24 is installed between this extension portion 22 and the support pin 23. By extending and retracting this push cylinder 24, the push plate 20 is made to swing back and forth.

[0036] The operation of the refuse loading device 31 will now be explained. As shown by the solid line in Figure 3, when the pushing plate 20 swings to the position closest to the refuse container 2 (forward limit position), the rotating plate 10 rotates upward without interfering with the pushing plate 20, and the pushing plate 20 swings toward the refuse input port 4 with a delay. After the pushing plate 20 swings to the position closest to the refuse input port 4 and reaches the position shown by the dashed line in Figure 3 (retraction limit position), the rotation of the rotating plate 10 continues.

[0037] As the rotating plate 10 moves, it scoops the debris towards the debris collection box 2 and stops at a set stop position that extends forward towards the debris collection box 2, as shown by the solid line in Figure 3. Then, the pushing plate 20 swings towards the debris collection box 2, pushing the debris on the rotating plate 10 into the debris collection box 2. When the pushing plate 20 reaches its forward limit position again, the rotating plate 10 begins to rotate upward again.

[0038] In this way, the rotation of the rotating plate 10 and the oscillation of the pushing plate 20 are repeated in sync with each other, thereby performing a refuse loading operation in which refuse put into the refuse input box 3 is continuously loaded into the refuse storage box 2. The configuration of the hydraulic circuit and control system for operating the rotating plate 10 and the pushing plate 20 will be described later.

[0039] Inside the waste collection box 3, switches LS1 to LS4 are provided for detecting the positions of the rotating plate 10 and the push plate 20. Specifically, as shown in Figure 3, switches LS1 and LS2 are provided, respectively, that turn on when the push plate 20 is at its forward limit position or backward limit position, switch LS3 is provided, that turns on when the rotating plate 10 is at its set stop position, and switch LS4 is provided, that turns on when the rotating plate 10 rotates a predetermined angle in the positive direction (clockwise in Figure 1) from its set stop position, and turns off when it rotates another predetermined angle.

[0040] Switches LS1 and LS2 are designed to detect a dog (not shown) located at the end of the pivot axis 21 of the push plate 20, while switches LS3 and LS4 are designed to detect a dog (not shown) located at the end of the rotation axis 11 of the rotating plate 10. These switches LS1 to LS4 can be, for example, limit switches, photoelectric switches, or proximity switches. Switch LS4, as shown by hatching in Figure 3, detects the angular range Z from directly below the front edge 4a of the waste inlet 4 to the rear edge 4b ​​of the waste inlet 4 as the rotating plate 10 rotates and descends towards the rear. It is a sensor for detecting that the rotating plate 10 is operating in the vicinity of the waste inlet 4.

[0041] Furthermore, as shown in Figures 1 and 3, emergency stop buttons 60 and 61, and an emergency stop plate 62 are provided near the refuse inlet 4 to stop the operation of the refuse loading device 31. As shown in Figure 1, the emergency stop button 60 (corresponding to switch SW1 in Figure 6) is located on the side of the switch box 6, which is located to the left of the refuse inlet 4. Also, as shown by the dashed line in Figure 3, the emergency stop button 61 (corresponding to switch SW3 in Figure 6) is located to the right of the refuse inlet 4. The emergency stop plate 62 is positioned below the refuse inlet 4 to turn switch SW2 on and off. In addition, as shown in Figure 2, a stop release switch 63 is located on the rear of the switch box 6. This stop release switch 63 is positioned to turn momentary switch SW4 on and off, for example. The stop release switch 63 is operated to temporarily release the emergency stop of the waste loading operation of the waste loading device 31, which is triggered by image recognition, as will be described later.

[0042] Next, the hydraulic circuit C (Figure 5), which constitutes the control system of the refuse loading device 31, and the electrical circuit E (Figure 6), which includes the control device PLC (Programmable Logic Controller) and the image processing unit 9, will be described. The control device PLC controls not only the operation of the refuse loading device 31 but also the operation of the refuse discharge device.

[0043] As shown in Figure 5, the hydraulic circuit C includes a hydraulic pump P, an oil reservoir T, a first electromagnetic control valve V1 for controlling the push cylinder 24, and a second electromagnetic control valve V2 for controlling the hydraulic motor 13. The hydraulic pump P is driven by a PTO (power take-off) which takes power from an engine (not shown).

[0044] The first and second electromagnetic control valves V1 and V2 are both electromagnetic directional control valves with 6 ports and 3 positions. When the solenoid SOLa of the first electromagnetic control valve V1 is energized by the control device PLC, it switches to the first communication position (upper position in Figure 5) and supplies hydraulic fluid from the hydraulic pump P to the rod-side oil chamber of the pair of push cylinders 24. On the other hand, when the solenoid SOLb of the first electromagnetic control valve V1 is energized by the control device PLC, it switches to the second communication position (lower position in Figure 5) and supplies hydraulic fluid to the head-side oil chamber.

[0045] When hydraulic fluid is supplied from the first electromagnetic control valve V1 to the head-side oil chamber, the pair of push cylinders 24 extend, causing the push plate 20 to swing forward. Conversely, when hydraulic fluid is supplied to the rod-side oil chamber, the pair of push cylinders 24 retract, causing the push plate 20 to swing backward. When neither solenoid SOLa nor SOLb is energized, the first electromagnetic control valve V1 returns to the neutral position (center position in Figure 5).

[0046] When the solenoid SOLc is energized, the second electromagnetic control valve V2 switches to the first communication position (lower position in Figure 5), supplying hydraulic fluid to the forward-rotating oil chamber of the hydraulic motor 13, causing the hydraulic motor 13 to rotate in the forward direction, and also allowing the push cylinder 24 to extend and retract. On the other hand, when the solenoid SOLd is energized, the second electromagnetic control valve V2 switches to the second communication position (upper position in Figure 5), supplying hydraulic fluid to the reverse-rotating oil chamber of the hydraulic motor 13, causing the hydraulic motor 13 to rotate in the reverse direction.

[0047] Furthermore, when neither solenoid SOLc nor SOLd is energized, the second electromagnetic control valve V2 returns to the neutral position (center position in Figure 5). When both the first and second electromagnetic control valves V1 and V2 are in the neutral position, the hydraulic fluid returns to the oil reservoir T. The hydraulic circuit C is also equipped with a check valve V3 to restrict the direction of hydraulic fluid flow and a relief valve V4 to adjust the upper limit of the discharge pressure of the hydraulic pump P.

[0048] Next, the signal flow in the control device PLC and the image processing unit 9 will be explained with reference to Figure 6. Power is supplied to the control device PLC by battery BT. The energizing line K2, which extends from the positive terminal of battery BT to the right side of Figure 6 and leads to the ground line K1, has the ignition switch SWK, PTO switch SWP, relay coil R1, etc. of the refuse collection vehicle 100 interposed therein.

[0049] Furthermore, the upstream end of the power supply line K3, which branches off from the power supply line K2, is connected between the ignition switch SWK and the battery BT, and the contact (relay switch) r1 of the relay coil R1 is interposed on the upstream side (closer to the battery BT). A power lamp L is interposed in this power supply line K3, and when the relay coil R1 is energized and contact r1 is closed, power is supplied to the power supply line K3, causing the power lamp L to light up.

[0050] Furthermore, the upstream end of the power supply line K4, which branches off from the power supply line K3, is connected between the contact r1 of the relay coil R1 and the power lamp L, thereby supplying power to the signal power supply unit (not shown) of the control device PLC. In other words, when contact r1 is closed, power is supplied to the control device PLC via power supply lines K3 and K4.

[0051] Furthermore, a power supply line K5, which branches off from the power supply line K4, ensures that the control device PLC is always energized while the refuse loading device 31 is in the process of loading refuse. In other words, power supply line K5 is the line that inputs the so-called loading continuation signal, and switches SW1 to SW3, which open and close in response to the operation of the emergency stop buttons 60 and 61 and the emergency stop plate 62 mentioned above, are installed here. When the power supply (i.e., the input of the loading continuation signal) is cut off by these switches SW1 to SW3, the control device PLC turns off the output of the control signal for exciting the solenoids SOLa to SOLd of the first and second electromagnetic control valves V1 and V2. As a result, the first and second electromagnetic control valves V1 and V2 return to the neutral position, and the operation of the refuse loading device 31 stops.

[0052] Furthermore, the power supply line K4 is connected to multiple branch lines that diverge from it, and each of these branch lines is fitted with the aforementioned switches LS1 to LS4. Signals from switches LS1 to LS4 are input to the control device PLC, and based on these signals, the position, i.e., the operating status, of the rotating plate 10 and the pushing plate 20 of the waste loading device 31 is detected.

[0053] Furthermore, in addition to switches LS1 to LS4, the input side to the control device PLC also has a loading switch (loading button) SW5 for operating the refuse loading device 31, a changeover switch SW6 for switching between the refuse loading operation of the refuse loading device 31 and the refuse discharge operation of the refuse discharge device, and a switch SW7 for tilting and opening the refuse input box 3. In addition, a selector switch for single-actuate or continuous refuse loading operation (not shown), and switches for independently operating the rotating plate 10 and the push plate 20 (not shown) are also electrically connected. Changeover switch SW6 is located at the branching point where the power supply line K5 branches off from the power supply line K4, and the power supply line K5 is connected to the loading side of changeover switch SW6.

[0054] As described above, various switches are connected to the input side of the control device PLC, while solenoids SOLa~SOLd for the first and second electromagnetic control valves V1 and V2 are connected to the output side of the control device PLC. The control device PLC is programmed to output to the corresponding solenoids SOLa~SOLd in order to operate the hydraulic motor 13, push cylinder 24, etc., according to a pre-set procedure based on the signals input from switches SW1~SW3, LS1~LS4, etc.

[0055] Specifically, when the waste loading device 31 is loading waste, the relay coil R1 is energized when both the ignition switch SWK and the PTO switch SWP on the power supply line K2 are turned on. This closes the contact r1 of the relay coil R1, so power is supplied to the control device PLC by the power supply lines K3 and K4. This makes the control device PLC operational and capable of outputting control signals to the solenoids SOLa~SOLd.

[0056] Upon receiving the control signal, the solenoids SOLa~SOLd are energized, the communication positions of the first and second electromagnetic control valves V1 and V2 are switched as appropriate, and hydraulic pressure is supplied to the hydraulic motor 13 and the push cylinder 24 as appropriate. As a result, the hydraulic motor 13 and the push cylinder 24 operate, and the rotation of the rotating plate 10 and the oscillation of the push plate 20 are repeated in sync with each other, as described above.

[0057] More specifically, as shown by the solid line in Figure 3, when the push plate 20 is in the forward limit position and an ON signal is output from switch LS1, and the rotating plate 10 is in the set stop position and an ON signal is also output from switch LS3, a control signal is output from the control device PLC, which receives the signal from the loading switch SW5. Then, the second electromagnetic control valve V2 is switched to the first communication position, the hydraulic motor 13 starts to rotate forward, and the rotating plate 10 starts to rotate upward.

[0058] After a predetermined period, a control signal is output from the control device PLC to the solenoid SOLa. This switches the first electromagnetic control valve V1 to the first communication position, the push cylinder 24 begins to retract, and the push plate 20 swings toward the rear waste inlet 4. When the push plate 20 reaches its retraction limit position, an ON signal is output from the switch LS2.

[0059] When the control device PLC receives an ON signal from switch LS2, it stops outputting the control signal to solenoid SOLa. This causes the first electromagnetic control valve V1 to return to the neutral position and stops the oscillation of the push plate 20. While the push plate 20 is oscillating, the rotating plate 10 continues to rotate, scooping the debris into the debris collection box 2. When the rotating plate 10 reaches the set stop position, an ON signal is output from switch LS3.

[0060] When the control device PLC receives an ON signal from switch LS3, it stops outputting a control signal to solenoid SOLc. This causes the second electromagnetic control valve V2 to return to the neutral position and stops the rotation of the hydraulic motor 13. The control device PLC also outputs a control signal to solenoid SOLb. This switches the first electromagnetic control valve V1 to the second communication position, causing the push cylinder 24 to extend and the push plate 20 to start swinging forward.

[0061] In this way, the pushing plate 20, which swings toward the forward waste collection box 2, pushes the waste on the rotating plate 10 into the waste collection box 2. When it reaches the forward limit position, an ON signal is output from switch LS1. In response, the control device PLC stops outputting the control signal to solenoid SOLb, causing the first electromagnetic control valve V1 to return to the neutral position, stopping the extension operation of the pushing cylinder 24, that is, the forward swinging of the pushing plate 20, and ending the series of operations.

[0062] As shown in Figure 6, the upstream end of the power supply line K6 is connected between the ignition switch SWK and the PTO switch SWP. The image processing unit 9 is connected to this power supply line K6. When the ignition switch SWK is turned on, power is supplied to the image processing unit 9 via the power supply line K6. The image processing unit 9 is connected to the camera 71 of the camera unit 7 via the power supply line K7. The image processing unit 9 is also configured to receive a reverse signal based on the vehicle's driving operation via the power supply line K8.

[0063] The image processing unit 9 includes a central processing unit CPU that executes a predetermined program to perform various controls, an image processing unit DSP that acquires image data from the camera 71 and performs image processing, a memory M that stores data used by the central processing unit CPU and the image processing unit DSP, and an image output unit VOP that receives commands from the central processing unit CPU and displays the results of image processing on a monitoring monitor 93 (see Figure 1). The image processing unit 9 also includes a position information acquisition device GR that acquires position information of the garbage collection vehicle 100 using a satellite positioning system such as GNSS (Global Navigation Satellite System), a proximity signal detection device CR that detects proximity signals transmitted from proximity signal transmitters installed at predetermined garbage loading locations, a timing unit C that measures the time of the data log, and a camera control unit (not shown) that controls the camera 71. The image processing unit DSP is an integrated circuit that performs image processing logic at high speed and executes the person recognition process shown in the flowchart of Figure 13.

[0064] Furthermore, the upstream end of the power supply line K9, which branches off from the power supply line K3, is connected upstream of the contact r1 of the relay coil R1, and power is supplied to the timing unit C via this power supply line K9. Therefore, even when the ignition switch SWK is off, power is always supplied to the timing unit C.

[0065] Here, the central processing unit CPU and the image processing unit DSP both constitute the image processing apparatus of the present invention. This image processing apparatus and the aforementioned camera 71 constitute the person recognition apparatus of the present invention.

[0066] Furthermore, the image processing unit 9 is connected to a power supply line K10 that branches off from the power supply line K5. This power supply line K10 is related to the person recognition processing performed by the image processing unit 9, and does not operate during the waste discharge operation of the waste loading device 31, but is a loading signal line that activates during the waste loading operation of the waste loading device 31. The power supply line K10 is connected to the loading side of the changeover switch SW6, and the loading line signal is input to the image processing unit 9 when the PTO switch SWP is ON and the changeover switch SW6 is switched to the loading side.

[0067] An operating switch SWS is interposed between the image processing unit 9 and the control device PLC. The image processing unit 9 is provided with a human safety signal output port, and this output port is connected to the operating switch SWS by a power supply line K16. The image processing unit 9 is also provided with a data log start signal input port, and this input port is connected to the operating switch SWS by a power supply line K17. Furthermore, a power supply line K15, which is branched from the power supply line K10, is connected to the operating switch SWS.

[0068] When the activation switch SWS is switched to the ON position, the output port of the person safety signal of the image processing unit 9 and the control device PLC are connected via the power lines K16 and K12.

[0069] Here, the person safety signal output from the image processing unit 9 is input to the control device PLC via the power lines K16 and K12. However, if a person enters the restricted area (the second detection area Y12 in Figures 7 to 10, described later) during the waste loading operation of the waste loading device 31, the person safety signal is not output from the image processing unit 9, and the person safety signal is not input to the control device PLC. The control device PLC is configured to perform an emergency stop of the waste loading operation of the waste loading device 31, with the absence of a person safety signal input being one of its conditions.

[0070] Furthermore, when the activation switch SWS is switched to the ON position, the power supply line K11, which branches off from the power supply line K6, is connected to the power supply line K17. In this case, when the ignition switch SWK is turned ON, power is supplied to the power supply line K6, and a data log start signal is sent to the image processing unit 9 through the power supply lines K11 and K17. This then triggers the execution of the data log.

[0071] On the other hand, when the activation switch SWS is switched to the off position, the power line K10 and the control device PLC are connected via the power lines K15 and K12. In this case, when the loading line signal is input to the image processing unit 9, the voltage from the power line K10 is input to the control device PLC instead of the person safety signal. Therefore, regardless of whether the image processing unit 9 has detected a person in the restricted area or not, the control device PLC always receives the person safety signal. Note that when the activation switch SWS is switched to the off position, power is not supplied to the power line K17, and data logging is not started.

[0072] Furthermore, a switch SW4 corresponding to the stop release switch 63 is interposed between the image processing unit 9 and the control device PLC. Switch SW4 is installed in the power supply line K13, which branches off from the power supply line K10. When the stop release switch 63 is not turned ON, switch SW4 is in the OFF state, and no signal is input to the control device PLC. On the other hand, when the stop release switch 63 is turned ON, switch SW4 is turned ON, and when the loading line signal is input to the image processing unit 9, the voltage from the power supply line K10 is input to the control device PLC as an ON signal for stop release. Note that when the waste loading device 31 is discharging waste, the loading line signal is not input to the image processing unit 9, so the function of the stop release switch 63 is disabled.

[0073] When the control device PLC recognizes the ON signal of the stop release switch 63, it is configured not to perform an emergency stop of the waste loading operation of the waste loading device 31, even if the input of a person safety signal from the image processing unit 9 ceases.

[0074] The image processing unit 9 is electrically connected to pilot lamps (not shown) located around the driver's seat, and to an operating lamp 72 and a detection lamp 73 located near the waste disposal port 4. The illumination of these lamps is controlled by the central processing unit CPU. The image processing unit 9 is also connected to a buzzer 91 that outputs a reverse warning sound, and this is also controlled by the central processing unit CPU. The low-potential sides of the operating lamp 72, detection lamp 73, and buzzer 91 are connected to the ground line K1 via the power supply line K14.

[0075] The operating lamp 72 is, in principle, turned on (lit) when a loading line signal is input to the image processing unit 9. The detection lamp 73 is, in principle, turned on (lit) when a loading line signal is input to the image processing unit 9 and it is determined by the person recognition process described later that a person has entered the person detection area near the waste inlet 4.

[0076] On the other hand, the operating lamp 72 and the detection lamp 73 may not light up as an exception, even if the image processing unit 9 determines that a person is in the person detection area near the waste inlet 4 while the loading line signal is input to the image processing unit 9. This exception occurs when the stop release switch 63 is turned on and the stop release ON signal is input to the control device PLC. In this case, the control device PLC energizes the relay coil R2, causing the relay switch r2 to cut off the output from the image processing unit 9 to the operating lamp 72 and the detection lamp 73, thereby forcibly turning off the operating lamp 72 and the detection lamp 73. This ensures that the operator is reliably informed that the person recognition processing function of the image processing unit 9 has been disabled.

[0077] Furthermore, the operating lamp 72 and detection lamp 73 are illuminated only when a loading line signal is input to the image processing unit 9. Therefore, when performing a waste discharge operation or a vehicle reverse operation without such input, the operating lamp 72 and detection lamp 73 are turned off. In other words, the control device PLC enables the operation of the operating lamp 72 and detection lamp 73 (allows them to be illuminated) when the changeover switch SW6 is switched to the waste loading operation (loading side), while disabling the operation of the operating lamp 72 and detection lamp 73 (keeps them off at all times) when the changeover switch SW6 is switched to the waste discharge operation (discharge side).

[0078] Furthermore, as shown in Figure 1, a camera 71 is positioned on the upper rear of the waste disposal box 3, in other words, above the waste disposal opening 4, with its imaging lens 71a (see Figure 4) pointed diagonally downwards and to the rear. Figure 7 shows a person H (worker) standing near the waste disposal opening 4 on a road, extending both arms forward to load waste G (garbage bags in Figure 8). Figure 9 also shows a person H standing near the waste disposal opening 4 on a platform, loading waste G (garbage bags in Figure 10).

[0079] The optical axis of the imaging lens 71a of the camera 71 is angled slightly backward from a vertically downward direction in order to photograph the vicinity of the waste inlet 4, so as to photograph the person H and waste G near the waste inlet 4 from above.

[0080] The image processing unit 9 determines, based on the image data captured by the camera 71 and input to the image processing unit 9, whether or not a person H, such as a worker, is in the waste inlet 4 and the person detection area behind it (the area enclosed by dashed lines and the area enclosed by solid lines in Figures 8 and 10). The control device PLC then determines, if the image processing unit 9 determines that a person H is in the person detection area, whether or not to stop the operation of the waste loading device 31.

[0081] The area for detecting people during loading is divided into a first detection area Y11 (shown by a dashed line in Figures 8 and 10) and a second detection area Y12 (shown by a solid line in Figures 8 and 10). The second detection area Y12 is located closer to the waste disposal port 4 than the first detection area Y11 and is set as a no-entry zone.

[0082] The image processing unit 9 determines whether a person H is in the first detection area Y11 or the second detection area Y12, and if it determines that a person H is in the second detection area Y12, which is a no-entry area, it sends a stop signal to the control device PLC. For example, the first detection area Y11 is the normal area when a worker is loading rubbish, and is set as a rectangular area near the rubbish inlet 4. The second detection area Y12 is set as a rectangular area between the front edge 4a and the rear edge 4b ​​of the rubbish inlet 4. This second detection area Y12 is set to include at least a part of the area enclosed by the front edge 4a, rear edge 4b, left edge 4c, and right edge 4d of the rubbish inlet 4.

[0083] In addition, when the garbage truck 100 is reversing, the image processing unit 9 determines whether or not person H is in the third detection area Y2 (the area shown by a dashed line in Figures 7 to 10) behind the garbage inlet 4. If it is determined that person H is in the third detection area Y2, the image processing unit 9 notifies the driver to alert them. The third detection area Y2 is set as a rectangular area that includes, for example, the second detection area Y12, the first detection area Y11, and the area behind them. On the monitor 93, along with the image from the camera 71, lines corresponding to the outer edge of the first detection area Y11 (shown by a dashed line in Figures 8 and 10), lines corresponding to the outer edge of the second detection area Y12 (shown by a solid line in Figures 8 and 10), which is a no-entry area, and lines corresponding to the outer edge of the third detection area Y2 (shown by a dashed line in Figures 8 and 10) may be displayed.

[0084] Next, the flow of control processing performed by the control device PLC will be explained with reference to Figure 11. Of the steps in Figure 11, all except steps S1 and S2 are performed by the control device PLC.

[0085] In the flowchart of Figure 11, steps S1 and S2 are steps that mark the start of control processing in the control device PLC. Step S1 is the step in which the ignition switch is switched from off to on, and step S2 is the step in which the power to the control device PLC is turned on when the PTO switch SWP is switched from off to on.

[0086] Step S3 follows step S2 and resets the stop release flag. In other words, when the PTO switch SWP is turned ON, the stop release flag is set to 0.

[0087] Step S4 is a step in which it is determined whether the changeover switch SW6 has been switched to the loading side (YES / NO) following step S3. If it is determined to be YES in step S4, the process proceeds to step S5; if it is determined to be NO, the process proceeds to step S14.

[0088] Step S5 is a step to determine whether the stop release flag is 1 or not (YES / NO). If NO is determined in step S5, the process proceeds to step S6; if YES is determined, the process proceeds to step S8.

[0089] Step S6 is a step to determine whether or not the stop release switch 63 has been switched from off to on (YES / NO). If the result in step S6 is YES, the process proceeds to step S7; if the result is NO, the process proceeds to step S10.

[0090] Step S7 is the step of setting the stop release flag to 1 as the stop release switch 63 is operated from off to on.

[0091] Step S8 is performed by the control unit PLC and is a step that energizes the relay coil R2 to switch the relay switch r2 to the OFF position.

[0092] Step S9 is the step of turning off the detection lamp 73 and the operating lamp 72.

[0093] Step S10 is a step in which the loading switch SW5 is determined to be turned on or not (YES / NO). If the control device PLC determines NO in step S10, it waits until the loading switch SW5 is turned on (YES). If it determines YES in step S10, it proceeds to step S11.

[0094] Step S11 is the step of performing the refuse loading operation of the refuse loading device 31. Step S12 is the step of determining whether (YES / NO) the emergency stop buttons 60, 61 or the emergency stop plate 62 were turned on following step S11.

[0095] If the result in step S12 is YES, proceed to step S13; if the result is NO, proceed to step S15.

[0096] Step S13 is the step of performing an emergency stop of the refuse loading operation of the refuse loading device 31.

[0097] Step S14 is executed following step S13 or step 19 (described later), or if NO is determined in step S4, and determines whether the power supply of the control device PLC has been switched from on to off (YES / NO). If it is determined in step S14 that the power supply of the control device PLC has been turned off (YES), the processing by the control device PLC is terminated. On the other hand, if it is determined in step S14 that the power supply of the control device PLC remains on (NO), the process returns to step S4 and the processing from step S4 onwards is executed.

[0098] Step S15 is a step to determine whether the rotating plate 10 of the waste loading device 31 is in a dangerous angle range Z (see Figure 3) (YES / NO). This determination is made based on the presence or absence of a signal from switch LS4. If the determination in step S15 is YES, the process proceeds to step S16; if the determination is NO, the process proceeds to step S18.

[0099] Step S16 is a step in which it is determined whether or not a person safety signal was input from the image processing unit 9 (YES / NO). If it is determined to be YES in step S16, the process proceeds to step S17; if it is determined to be NO, the process proceeds to step S18.

[0100] The person safety signal is input from the image processing unit 9, which will be described later, in accordance with the processing performed by the image processing unit 9. Specifically, if the activation switch SWS is ON and a person enters the first detection area Y11 during the refuse loading operation of the refuse loading device 31, the image processing unit 9 outputs a person safety signal, and the person safety signal is input to the control device PLC (step S119 in Figure 12). If a person enters the second detection area Y12, the image processing unit 9 does not output a person safety signal, and the person safety signal is not input to the control device PLC (step S114 in Figure 12). On the other hand, if the activation switch SWS is OFF, the person safety signal is input to the control device PLC regardless of whether a person is detected in the first detection area Y11 or the second detection area Y12 during the refuse loading operation of the refuse loading device 31 (step S117 in Figure 12). Furthermore, if the image processing unit 9 suddenly malfunctions during the waste loading operation of the waste loading device 31, and the image processing unit 9 stops outputting a person safety signal, the person safety signal will no longer be input to the control device PLC.

[0101] Step S17 is a step to determine whether the stop release flag is 0 or not (YES / NO). If it is determined to be YES in step S17, the process proceeds to step S13 to emergency stop the waste loading operation of the waste loading device 31; if it is determined to be NO, the process proceeds to step S18.

[0102] Step S18 is a step to determine whether the rotating plate 10 of the waste loading device 31 is at the stopping position (1 cycle stopping position) when the waste loading operation is stopped (YES / NO). If NO is determined in step S18, the process returns to step S11 and the processing from step S11 onwards is executed again. On the other hand, if YES is determined in step S18, the process proceeds to step S19.

[0103] Step S19 is a step in which the refuse loading operation of the refuse loading device 31 is stopped under normal circumstances (stopped at the 1-cycle stop position).

[0104] Next, the flow of control processing performed in the image processing unit 9 will be explained with reference to Figure 12. All processes except for steps S101, S108, and S124 in Figure 12 are performed by the central processing unit CPU of the image processing unit 9.

[0105] In Figure 12, step S101 is the step of turning on the power supply of the image processing unit 9 from off when the ignition switch SWK is switched from off to on.

[0106] Step S102 is a step in which the image processing unit 9 is initialized (started up). Until the initialization process of the image processing unit 9 is completed, the image processing unit 9 will not output the person safety signal described later, nor will it output an ON signal to the operating lamp 72 and the detection lamp 73.

[0107] If the operator turns on the PTO switch SWP during the initialization process of the image processing unit 9, the control device PLC will recognize that there is no human safety signal input from the image processing unit 9. In this case, even if the operator presses the loading button 64 immediately after turning on the PTO switch SWP, the control device PLC will, based on the lack of a human safety signal input, perform control to prevent the waste loading device 31 from being driven.

[0108] Step S103 is a step to determine whether the image processing unit 9 is operating normally (YES / NO). If the result in step S103 is YES, the process proceeds to step S104; if the result is NO, the process proceeds to step S120.

[0109] In step S120, the image processing unit 9 performs a process to prevent the operation lamp 72 from lighting up, and in S121, it lights up an error indicator light provided on the circuit board, and then proceeds to the process in step S115.

[0110] Step S104 is a step to determine whether or not a reverse signal based on the vehicle's driving operation has been input (YES / NO). A reverse signal is input to the image processing unit 9 via the power line K8 when the vehicle is reversed, and the determination in step S104 is made based on the presence or absence of this reverse signal. If the determination in step S104 is YES, the process proceeds to step S105; if the determination is NO, the process proceeds to step S122.

[0111] Step S105 is a step to determine whether or not a loading line signal has been input (YES / NO). The loading line signal is input to the image processing unit 9 via the power line K10 when the PTO switch SWP is ON and the changeover switch SW6 is switched to the loading side. The determination in step S105 is made based on the presence or absence of this loading line signal. If the determination in step S105 is YES, the process proceeds to step S106; if the determination is NO, the process returns to step S104.

[0112] Step S106 is the step of illuminating the operating lamp 72.

[0113] Step S107 is performed following step S106 and involves setting the feature data (dictionary data) used for the first person recognition process as the feature data for loading, and setting the detection area used for person detection determination as the person detection area for loading.

[0114] The feature data for loading is referenced during the first person recognition process (step S108), which will be described later, and is stored in the memory M of the image processing unit 9. This feature data for loading includes feature data for loading related to the head, which is part of the upper body of the person, and feature data for loading related to the legs, which is part of the lower body.

[0115] To elaborate, in the refuse collection vehicle 100, as shown in Figure 1, a camera 71 is installed above the refuse input port 4, and is positioned to photograph the area near the refuse input port 4 from almost directly above. Therefore, as shown in Figure 7, in refuse loading operations on a road R where the position where person H stands is at the same height as the ground surface of the refuse collection vehicle 100, the image from the camera 71 mainly shows the head of person H, as well as part of their face, as illustrated in Figure 8. On the other hand, as shown in Figure 9, this is not the case in refuse loading operations where person H stands on a platform or other elevated surface S, which is higher than the ground surface of the refuse collection vehicle 100. In operations on a platform or other elevated surface S, as illustrated in Figure 10, the upper part of person H's body, including their head, often extends beyond the camera 71's field of view, and only a certain range of the lower part of their body is displayed. For this reason, as mentioned above, feature data for loading is prepared for both the head and the legs.

[0116] The feature data for loading human heads is obtained by taking images of many people's heads in advance and extracting features such as size and shape. However, as shown in the example in Figure 8, the focus is on the heads of people like person H standing at the same height as the ground contact surface of the garbage collection vehicle 100, and the features of such heads as viewed from above are the main focus. Similarly, the feature data for loading human legs is obtained by taking images of many people's legs in advance and extracting features such as size and shape. However, as shown in the example in Figure 10, the focus is on the legs of people like person H standing at a height where their heads are not captured, and the features of such legs as viewed from above are the main focus.

[0117] Loading feature data for a person's head includes, for example, the size, position, and shape of hats, helmets, and other headwear. Loading feature data for a person's legs includes the size, position, and shape of work clothes, shoes, and other legwear.

[0118] The area for detecting people during loading is used when determining whether a person is detected and is stored in the memory M of the image processing unit 9. As mentioned above, the area for detecting people during loading is divided into a first detection area Y11, which is set as the normal area when a worker is loading rubbish, and a second detection area Y12, which is a no-entry area, and each area is stored in the memory M.

[0119] Step S108 is the step in which the central processing unit CPU of the image processing unit 9 instructs the image processing unit DSP to execute the first person recognition process, following step S107. The first person recognition process executed by the image processing unit DSP will be explained with reference to Figure 13.

[0120] In Figure 13, step S181 is a step in which a binarization process is performed on the image data captured by the camera 71 and input to the image processing unit 9. This binarization process, for example, sets the brightness value of each pixel in the image data to the maximum brightness value if it is above a preset threshold, and to the minimum brightness value if it is below the threshold. The resulting binarized image data has much of the noise and the effects of light intensity changes removed.

[0121] Step S182 is a step in which a first labeling process is performed. In this first labeling process, a labeling process and a head determination process are performed.

[0122] The labeling process in the first labeling step involves dividing each adjacent pixel in the binarized image data into regions. For example, multiple pixels belonging to the same brightness value and closely located within a predetermined distance are considered as a single region. The labeling process is performed on the entire image plane, and each of these regions is recognized as an object image.

[0123] The head detection process in the first labeling process refers to pre-prepared feature data for loading human heads and extracts object images from among the object images recognized in the labeling process that satisfy the conditions of the feature data for loading human heads. Information indicating the presence or absence of such object images is stored in the memory M of the image processing unit 9.

[0124] Step S183 is executed following step S182 and involves referring to the result of the first labeling process stored in memory M in step S182 to determine whether or not there is an object image that satisfies the conditions for the loading feature data related to a person's head (YES / NO). If the result in step S183 is YES, the process proceeds to step S184; if the result is NO, the process proceeds to step S190.

[0125] Step S184 is the step of setting an object image that satisfies the conditions for the loading feature data related to the head extracted in step S182 as a head identification target. If there are multiple object images that satisfy these conditions, each of these object images is set as a head identification target.

[0126] Step S185 is a step in which the head identification target set in step S184 is identified as a head or not. In this step S185, the image processing unit DSP detects changes in the position of the object image by performing difference processing (processing to find the difference in brightness value for each pixel) of multiple image data input in time series from the camera 71. The movement speed of the object image is calculated by dividing this change in position, that is, the distance the object image has moved, by the time required (the time interval for acquiring image data).

[0127] Step S186 determines whether the movement speed calculated in step S185 is below a predetermined threshold (YES / NO). This threshold is set in advance through experiments to distinguish between the movement speed of the worker's head when putting waste into the waste inlet 4 and the movement speed of the waste being put in.

[0128] If the result in step S186 is YES, it is determined that the object image targeted for head identification represents a person's head, and the process proceeds to step S187. On the other hand, if the result in step S186 is NO, it is determined that the object image targeted for head identification is not the head of person H but is rubbish because it is moving too fast, and the process proceeds to step S190. The processes in steps S185 and S186 are provided to distinguish between a person's head and something similar in shape and color, such as a black garbage bag.

[0129] Step S187 is a step in which a person detection signal is output. This person detection signal is referenced by the central processing unit CPU of the image processing unit 9 in step S111 in Figure 12, which will be described later, and is the signal that becomes the output condition for the person detection lamp 73.

[0130] Step S188 is performed following step S187 and determines whether (YES / NO) the object image determined to be a human head is located within the second detection area Y12. The determination in step S188 is made, for example, by determining whether at least a portion of the position coordinates of the pixels that form the outline of the region forming the object image are included in the second detection area Y12. If the result in step S188 is YES, the process proceeds to step S189; if the result is NO, the process ends.

[0131] Step S189 is the step in which a stop signal is output. This stop signal is referenced by the central processing unit CPU of the image processing unit 9, as described later, and becomes the condition for executing the process of not outputting a person safety signal (steps S112 to S114 in Figure 12). If a person safety signal is not output, the process follows the flow of steps S16 to S13 in Figure 11, and the operation of the waste loading device 31 stops. When a stop signal is output in step S189, the first person recognition process is terminated.

[0132] If, in step S183, it is determined that there is no object image that satisfies the conditions for the feature data for loading related to the head (NO), or if, in step S186, it is determined that the calculated movement speed is higher than the threshold (NO), the process proceeds to step S190.

[0133] Step S190 is a step in which a second labeling process is performed. In this second labeling process, a labeling process and a leg determination process are performed.

[0134] The labeling process in the second labeling process, similar to the first labeling process, involves defining each adjacent pixel in the binarized image data generated in step S181 as a region. For example, multiple pixels that belong to the same brightness value and are closely located within a predetermined distance are considered as a single region. The labeling process is performed on the entire image plane, and each of these resulting regions is recognized as an object image.

[0135] The leg detection process in the second labeling process refers to pre-prepared feature data for loading human legs and extracts object images from each object image recognized in the labeling process that satisfy the conditions of the feature data for loading human legs. Information indicating the presence or absence of such object images is stored in the memory M of the image processing unit 9.

[0136] Step S191 is executed following step S190 and involves referring to the result of the second labeling process stored in memory M in step S190 to determine whether or not there is an object image that satisfies the conditions for loading feature data related to a person's legs (YES / NO). If YES is determined in step S191, the process proceeds to step S192; if NO is determined, the process ends.

[0137] Step S192 is the step in which, if it was determined in step S191 that there is an object image that satisfies the conditions for loading feature data related to a person's legs (YES), that object image is determined to be a person's leg. In step S192, unlike the determination of the head, the object is determined to be a person's leg without performing the processing related to movement speed corresponding to steps S185 to S186. This is because the shape of the legs often differs greatly from that of rubble, so there is little need for further identification from rubble based on movement speed.

[0138] Return to the flowchart in Figure 12. Step S109 is a step to determine whether the operating switch SWS is ON or OFF (YES / NO). If it is determined to be YES in step S109, proceed to step S110; if it is determined to be NO, proceed to step S116.

[0139] Step S110 is the step of saving the processing result (log) of the first person recognition process in step S108 described above to memory M. In this step S110, whether or not a person detection signal was output (step S187) and whether or not a stop signal was output (step S189) are saved to memory M of the image processing unit 9.

[0140] Step S111 is executed following step S110 and determines whether (YES / NO) a person was detected in the first detection area Y11 or the second detection area Y12 during the first person recognition process in step S108. At this time, the central control unit CPU determines whether or not a person detection signal was output by referring to the memory M of the image processing unit 9. If YES is determined in step S111, the process proceeds to step S112; if NO is determined, the process proceeds to step S116.

[0141] Step S112 is a step in which it is determined whether or not (YES / NO) a person has entered the second detection area Y12, which is designated as a no-entry area, in the first person recognition process of step S108. At this time, the central processing unit CPU determines whether or not a stop signal has been output by referring to the memory M of the image processing unit 9. If it is determined to be YES in step S112, the process proceeds to step S113; if it is determined to be NO, the process proceeds to step S118.

[0142] Step S113 is a step in which an ON signal is output for the detection lamp 73, and the following step S114 is a step in which the person safety signal is not output to the control device PLC.

[0143] Step S116 is a step in which the ON signal for the detection lamp 73 is not output, and the following step S117 is a step in which a person safety signal is output to the control device PLC.

[0144] Step S118 is the step of outputting an ON signal for the detection lamp 73, and the following step S119 is the step of outputting a person safety signal to the control device PLC. After the processes in steps S114, S117, and S119 are performed, the process proceeds to step S115.

[0145] Step S115 is a step to determine whether the power supply of the image processing unit 9 has been switched from on to off (YES / NO). If the result in step S115 is YES, the process is terminated. On the other hand, if the result in step S115 is NO, the process returns to step S103 and the process from step S103 onwards is repeated.

[0146] If the determination in step S104 above indicates that there is a reverse signal input (NO), the process proceeds to step S122, where the central processing unit CPU illuminates the operating lamp 72. Then, the processes from step S123 onward are executed.

[0147] Step S123 is a step in which the feature data (dictionary data) used for person recognition processing is set as the feature data for backward movement, and the detection area used for person detection determination is set as the person detection area for backward movement.

[0148] The feature data for receding includes feature data for the head of a person and feature data for the legs of a person. The feature data for the head of a person is obtained by taking images of many people's heads in advance and extracting features such as size and shape, similar to the feature data for loading. However, the features extracted as this feature data for receding mainly represent the shape of the head when viewed from a more forward and oblique angle above than the feature data for loading mentioned above.

[0149] Furthermore, the feature data for the legs of people used for reversing, like the feature data for loading, was obtained by taking images of the legs of many people in advance and extracting features such as their size and shape. However, the features extracted for this reversing feature data mainly consist of the shape of the legs when viewed from a more forward, oblique angle above than the feature data for loading mentioned above.

[0150] The person detection area for backward movement is used when determining whether a person is detected while moving backward, and is stored in the memory M of the image processing unit 9. In this embodiment, the person detection area for backward movement is set in the second detection area Y12, the first detection area Y11, and the third detection area Y2 which includes the area behind them.

[0151] Step S124 is executed by the image processing unit DSP of the image processing unit 9 and is a step in which the second person recognition process is performed. The flow of the second person recognition process is shown in Figure 14. This second person recognition process is slightly different from the first person recognition process in step S108 (see Figure 13) described above. Specifically, in the second person recognition process in step S124, the feature data for receding movement related to the head and the feature data for receding movement related to the legs, which were set in step S123, are used. Then, an object image that satisfies the conditions of each receding feature data is determined to be the head or legs of a person. Also, since the rubble loading device 31 is in a stopped state from the beginning when receding, the processing of steps S188 to S189 in the first person recognition process in Figure 13 is omitted, as shown in Figure 14. The processes in steps S181-S187 and S190-S192 in the second person recognition process are the same as the processes in steps S181-S187 and S190-S192 in the first person recognition process shown in Figure 13.

[0152] Step S125 is the step of saving the processing result (log) of the second person recognition process from step S124 to memory M.

[0153] Step S126 is a step in which it is determined whether or not a person has been detected (YES / NO) within the third detection area Y2 by the second person recognition process in step S124. The processing in step S126 is performed by referring to the memory M of the image processing unit 9 to determine whether or not a person detection signal has been output. If it is determined to be YES in step S126, the process proceeds to step S127; if it is determined to be NO, the process proceeds to step S128.

[0154] Step S127 is performed when a person is detected within the third detection area Y2, and is the step of activating the buzzer 91 and outputting a reverse warning sound.

[0155] Step S128 is performed when no person is detected within the third detection area Y2, and it is a step that prevents the buzzer 91 from being activated. After each of steps 127 and S128 has been performed, the process proceeds to step S115.

[0156] As described above, the person recognition device of the first embodiment can determine whether a person has entered the restricted area (second detection area Y12) not only when working on a road R where the person's head can be photographed by the camera 71 (see Figure 7), but also when working on a stepped area S such as a platform where the person's head extends beyond the camera 71's field of view and only their legs are photographed (see Figure 9). Therefore, the person recognition device of this embodiment can determine whether a person has entered the restricted area more reliably than conventional person recognition devices that determine whether a person has entered the restricted area based only on an image of the person's head. As a result, the safety of the garbage collection vehicle 100 is further enhanced.

[0157] Next, a person recognition device according to a second embodiment of the present invention will be described. The person recognition device according to the second embodiment is an example in which the processing flow executed in the image processing unit of the person recognition device according to the first embodiment has been modified. Components similar to those in the first embodiment are denoted by the same reference numerals and their description is omitted.

[0158] In the person recognition device according to the second embodiment, as shown in Figure 15, if the result in step S105 is determined to be YES, the process in step S201 is executed. On the other hand, if the result in step S105 is determined to be NO, the process returns to step S104, similar to the person recognition device according to the first embodiment.

[0159] Step S201 is executed by the central processing unit CPU of the image processing unit 9 and determines whether the loading line signal has changed from off to on (YES / NO). Specifically, in step S201, it is determined whether the PTO switch SWP has been turned on while the changeover switch SW6 shown in Figure 6 is in the loading position, or whether the changeover switch SW6 has been switched from the discharge side to the loading side while the PTO switch SWP is on. If the result in step S201 is YES, the process proceeds to step S202; if the result is NO, the process proceeds to step S106.

[0160] Step S202 is performed by the image processing unit (DSP) and determines the location of the waste loading work currently underway. The work location determination process in step S202 will be explained with reference to Figure 16.

[0161] In Figure 16, step S301 is a step in which a binarization process is performed on the image data captured by the camera 71 and input to the image processing unit 9. The binarization process is the same as the binarization process (step S181) in the first person recognition process shown in Figure 13.

[0162] Step S302 is a step in which a third labeling process is performed. In this third labeling process, labeling and work location identification are performed.

[0163] The labeling process in the third labeling process is the same as the labeling process in the first person recognition process shown in Figure 13. The work location identification process in the third labeling process refers to the feature data of a predetermined work location with steps that has been selected in advance, and extracts object images that satisfy the conditions of that feature data from among the object images recognized in the above labeling process. Information indicating the presence or absence of such object images is stored in the memory M of the image processing unit 9.

[0164] Step S303 is executed following step S302 and involves referring to the result of the third labeling process stored in memory M in step S302 to determine whether or not there is an object image that satisfies the conditions of the work location feature data (YES / NO). If YES is determined in step S303, the process proceeds to step S304; if NO is determined, the process proceeds to step S305.

[0165] Step S304 is the step of determining that the current loading operation is being carried out at a predetermined work area with a step, which has been selected in advance. Step S305 is the step of determining that the current loading operation is not being carried out at a predetermined work area with a step, which has been selected in advance. In steps S304 and S305, information indicating whether or not the work is being carried out at the predetermined work area is stored in the memory M of the image processing unit 9.

[0166] In Figure 15, step S203 is the step in which the central processing unit CPU of the image processing unit 9 causes the image processing unit DSP to execute the third person recognition process, and corresponds to step S108 in Figure 12 in the first embodiment. The third person recognition process executed by the image processing unit DSP will be described with reference to Figure 17.

[0167] In Figure 17, step S401 is a step to determine whether or not work is being performed at a predetermined work location (YES / NO). This determination is made by referring to information stored in memory M in steps S304 and S305 of the work location determination process shown in Figure 16, which indicates whether or not work is being performed at a predetermined work location with steps. If the result in step S401 is YES, the process proceeds to step S402; if the result is NO, the process proceeds to step S403.

[0168] Step S402 is a step that performs the same processing as the binarization process (step S181) in the first person recognition process shown in Figure 13. Step S402 is executed on the condition that it was determined in step S401 that no work is being performed at a predetermined work area with steps (YES).

[0169] After the processing in step S402, the same processing as the first person recognition processing shown in Figure 13 is performed. That is, processing is performed to determine whether or not there is an object image that satisfies the conditions of the loading feature data related to a person's head, and as a result, a stop signal may be output to stop the operation of the waste loading device 31. However, for the determination processing in steps S404 and S405, which correspond to steps S183 and S186 in the first person recognition processing, the destination of the processing depending on the result of the determination is different.

[0170] If the result in step S404 is YES, the same process as the first person recognition process shown in Figure 13 is executed. However, if the result is NO, the third person recognition process is terminated. Unlike the first person recognition process, the third person recognition process in this embodiment is performed on the condition that the work is being done in a situation where the person's head is easily photographed (YES in step S401). Therefore, if the result in step S404 is NO, the process is terminated without determining whether or not there is an object image that satisfies the conditions for the loading feature data related to the person's legs.

[0171] Similarly, if the result in step S405 is YES, the same process as the first person recognition process shown in Figure 13 is executed, while if the result is NO, the third person recognition process is terminated. In other words, for the same reasons as in step S404, if the result in step S405 is NO, the process is terminated without determining whether or not there is an object image that satisfies the conditions for the loading feature data related to the legs of a person.

[0172] Step S403 is a step that performs the same processing as the binarization process (step S181) in the first person recognition process shown in Figure 13. Step S403 is performed on the condition that it was determined in step S401 that the work is being done in a predetermined work area with steps (NO). In other words, step S403 is performed when the work is being done in a situation where the person's head is likely to be outside the shooting range of the camera 71, while the person's legs are likely to be photographed.

[0173] After the processing in step S403, the same processing as the first person recognition processing shown in Figure 13 is performed. That is, a determination is made as to whether or not there is an object image that satisfies the conditions of the loading feature data concerning the legs of a person, and as a result, a stop signal may be output to stop the operation of the waste loading device 31.

[0174] In Figure 15, step S204 is the step in which the central processing unit CPU of the image processing unit 9 causes the image processing unit DSP to execute the fourth person recognition process, and corresponds to step S124 in the first embodiment. The fourth person recognition process will be described with reference to Figure 18.

[0175] The fourth person recognition process does not include a step of comparing the object image in the image with feature data for backward movement related to legs to determine whether or not it is a person's legs. Specifically, as shown in Figure 18, if the result in step S501 is YES, the same process as the second person recognition process shown in Figure 14 is executed, while if the result is NO, the fourth person recognition process is terminated without executing the process shown in steps S190 to S192 in Figure 14. This is because, when the garbage truck 100 is backing up, it is not necessary to recognize the legs of a person standing on a platform or other elevated surface.

[0176] Furthermore, since the waste loading device 31 is in a stopped state from the beginning when reversing, the processes corresponding to steps S188 to S189 in Figure 13 are omitted.

[0177] In summary, the person recognition device of the second embodiment can make a more reliable determination of whether or not a person has entered a restricted area based solely on an image of the person's head, similar to the person recognition device of the first embodiment.

[0178] Furthermore, in the person recognition device of the second embodiment, prior to the third person recognition process (step 203 in Figure 15), the work location determination process (step S202 in Figure 15) determines whether or not loading work is being performed at a predetermined work location with steps. If it is determined that loading work is not being performed at a work location with steps (processing from step S303 to step S305 in Figure 16), in other words, in situations where a person's head is likely to be photographed in relation to the installation position of the camera 71, the process proceeds to determining whether or not the object image in the image represents a person's head by comparing it with the characteristic data of a person's head (step S401 to step S402 in Figure 17). On the other hand, if the work location determination process (step S202 in Figure 15) determines that loading work is being carried out at the above work location (processing from step S303 to step S304 in Figure 16), in other words, in situations where a person's head is relatively often outside the shooting range of the camera 71 due to its position, the process proceeds to determine whether the object image in the image represents a person's legs by comparing it with the characteristic data of a person's legs (steps S401 to S403 in Figure 17).

[0179] In short, according to the person recognition device of the second embodiment, it is preferable to perform person recognition processing by comparing with the characteristic data of the person's head, or by comparing with the characteristic data of the person's legs, and this is automatically selected as appropriate depending on the work location for loading rubbish, and only one of the processes is executed. Therefore, compared to the person recognition device of the first embodiment, the processing load in the person recognition process can be reduced.

[0180] Next, a person recognition device according to the third embodiment of the present invention will be described. The person recognition device according to the third embodiment is an example in which the work location determination process (step S202 in Figure 15) of the person recognition device according to the second embodiment has been modified. The other configurations are the same as in the second embodiment, so their description will be omitted.

[0181] The work location determination process of the third embodiment will be described with reference to Figure 19. In Figure 19, step S601 is executed, for example, by the central processing unit CPU of the image processing unit 9, and is a step that determines whether or not (YES / NO) the proximity signal detection device CR shown in Figure 6 has detected a predetermined proximity signal transmitted by a proximity signal transmitter installed on the step area. If YES is determined in step S601, the process moves to step S602, and if NO is determined, the process moves to step S603.

[0182] Here, the proximity signal detection device CR detects the proximity signal transmitted by the proximity signal transmitter when the distance to the proximity signal transmitter falls within a predetermined value. The proximity signal transmitter may transmit the proximity signal continuously or under predetermined conditions. As an example of transmitting the proximity signal under predetermined conditions, a vehicle detection sensor may be installed on the side wall surface forming the step or on the road surface near the step, so that the vehicle detection sensor detects the garbage truck when the distance between the garbage truck and the step falls within a predetermined distance. The proximity signal transmitter may then transmit the proximity signal while the vehicle detection sensor is detecting the garbage truck. The vehicle detection sensor may be embedded in the side wall surface forming the step or on the road surface near the step.

[0183] Step S602 is executed, for example, by the central processing unit CPU of the image processing unit 9, and determines that the current loading operation is being carried out at the aforementioned pre-selected work location. Step S603 is determined that the current loading operation is not being carried out at the pre-selected predetermined work location. In steps S602 and S603, information indicating whether or not work is being carried out at the predetermined work location is stored in the memory M of the image processing unit 9.

[0184] After the processing in steps S602 and S603 is completed, the same processing as the person recognition device of the second embodiment shown in Figure 15 is performed. However, the determination process in step S401 in Figure 17 is performed by referring to the information stored in memory M by the determination process in steps S602 and S603 in Figure 19, which indicates whether or not work is being performed at a predetermined work location.

[0185] Next, a person recognition device according to the fourth embodiment of the present invention will be described. The person recognition device according to the fourth embodiment is an example in which the work location determination process (step S202 in Figure 15) in the person recognition device according to the second embodiment has been modified. The other configurations are the same as in the second embodiment, so their description will be omitted.

[0186] The work location determination process of the fourth embodiment will be described with reference to Figure 20. In Figure 20, step S701 is a step in which the central processing unit CPU of the image processing unit 9 refers to the location information acquired by the location information acquisition device GR shown in Figure 6.

[0187] Step S702 is executed by the central processing unit CPU and determines whether (YES / NO) whether the garbage collection vehicle 100 is located within a predetermined area, including a predetermined work area with a step, based on the location information referenced in step S701. If the result in step S702 is YES, the process proceeds to step S703; if the result is NO, the process proceeds to step S704.

[0188] Step S703 is executed by the central processing unit CPU and determines that the current loading operation is being performed at a predetermined work location with a step. Step S704 is executed by determining that the current loading operation is not being performed at a predetermined work location with a step. In steps S703 and S704, information indicating whether or not work is being performed at a predetermined work location is stored in the memory M of the image processing unit 9.

[0189] After the processing in steps S703 and S704 is completed, the same processing as the person recognition device of the second embodiment shown in Figure 15 is performed. However, the determination process in step S401 in Figure 17 is performed by referring to the information stored in memory M by the determination process in steps S703 and S704 in Figure 20, which indicates whether or not work is being performed at a predetermined work location.

[0190] The person recognition device and garbage collection vehicle according to the present invention have been described above based on the first to fourth embodiments, but the specific configuration is not limited to these embodiments. The above-mentioned problems can also be solved by changing or adding to the configuration within the scope that does not depart from the gist of the invention as described in the claims.

[0191] For example, we described an instance where feature data relating to the head, part of the upper body of a person, is used as one of the feature data for loading or unloading, but the feature data is not limited to this. For example, the feature data for the upper body may relate to a wider range of parts, with or without the head.

[0192] Furthermore, while we have described an example where feature data concerning the legs of a person's lower body is used as one of the feature data for loading or reversing, the feature data for the lower body is not limited to this. For example, the feature data for the lower body may include or exclude the legs, or may cover a wider range of body parts.

[0193] Furthermore, the upper body feature data may be limited to a narrower area, specifically a portion of the head. Similarly, the lower body feature data may be limited to a narrower area, specifically the feet.

[0194] Furthermore, with respect to the first person recognition process of the first embodiment shown in Figure 13, in the first labeling process of step S182, if the object image in the image is compared with feature data relating to the legs of a person to extract the legs of a person, and it is determined that there are no leg shapes, then in the second labeling process of step S190, the object image in the image may be compared with feature data relating to the head of a person to extract the head of a person. In this case, steps S184 to S186 are performed after the execution of the second labeling process of step S190, and steps S191 and S192 are performed after the execution of the first labeling process of step S182.

[0195] Furthermore, the central processing unit CPU and the image processing unit DSP may be a single, inseparable processing unit. In addition, each of these processes may be executed by a processing unit different from the central processing unit CPU and the image processing unit DSP. [Industrial applicability]

[0196] This invention can be widely applied to any refuse collection vehicle. [Explanation of Symbols]

[0197] 4 Garbage inlet 71. Camera (part of a person recognition system) 100 Refuse Collection Vehicles CPU (Central Processing Unit - part of the image processing unit) CR proximity signal detection device DSP Image Processing Unit (part of the image processing unit) GR location information acquisition device R step field

Claims

1. A person recognition device comprising a camera installed above the refuse inlet of a refuse collection vehicle, and an image processing device that recognizes a person near the refuse inlet based on image data captured by the camera, The aforementioned image processing device is A person determination means capable of performing the following: upper body determination process, which compares an object image contained in the image with predetermined upper body feature data of a person standing on the ground surface of the garbage collection vehicle to determine whether the object image represents the upper body; and lower body determination process, which compares an object image contained in the image with predetermined lower body feature data of a person standing on a predetermined step higher than the ground surface, where the upper body extends beyond the camera's shooting range, to determine whether the object image represents the lower body; The system includes an intrusion determination means that determines whether an object image determined by the person determination means to represent the upper or lower part of a body is inside a predetermined area near the waste disposal opening, The person recognition device is characterized in that the person determination means performs the lower body determination process when it is determined that the object image does not represent the upper body after performing the upper body determination process, or performs the upper body determination process when it is determined that the object image does not represent the lower body after performing the lower body determination process.

2. A person recognition device comprising a camera installed above the refuse inlet of a refuse collection vehicle, and an image processing device that recognizes a person near the refuse inlet based on image data captured by the camera, The aforementioned image processing device is A work determination means for determining whether or not a step-level operation is being performed in which refuse is loaded from a predetermined step-level that is higher than the ground surface of the refuse collection vehicle, If the work determination means determines that the step work has not been performed, the person determination means compares the object image contained in the image with predetermined upper body feature data of a person to determine whether the object image represents the upper body; if the work determination means determines that the step work has been performed, the person determination means compares the object image contained in the image with predetermined lower body feature data of a person to determine whether the object image represents the lower body; An intrusion determination means determines whether or not an object image determined by the person determination means to represent the upper or lower part of a body is in a predetermined area near the waste disposal opening, A person recognition device characterized by having the following.

3. In the person recognition device according to claim 1 or 2, The aforementioned upper body feature data includes the size, position, and shape of the items attached to the upper body. A person recognition device characterized in that the characteristic data of the lower body includes the size, position, and shape of an object attached to the lower body.

4. In the person recognition device according to claim 3, A person recognition device characterized in that the upper part of the body is the head and the lower part of the body is the legs.

5. In the person recognition device according to claim 2, The person recognition device is characterized in that the work determination means compares the object image contained in the image with the feature data of the step area to determine whether or not the object image represents the step area, determines that the step work has not been performed if it is determined that the object image does not represent the step area, and determines that the step work has been performed if it is determined that the object image represents the step area.

6. In the person recognition device according to claim 2, An approach signal transmitter installed on the step area that transmits a predetermined approach signal, An approach signal detection device installed on the garbage collection vehicle detects the approach signal when the distance to the approach signal transmitter is within a predetermined value, Furthermore, A person recognition device characterized in that the work determination means of the image processing device determines that the step work is not being performed if the proximity signal detection device does not detect the proximity signal, and determines that the step work is being performed if the proximity signal detection device has received the proximity signal.

7. In the person recognition device according to claim 2, The system further includes a location information acquisition device that acquires location information of the garbage collection vehicle using a satellite positioning system, The person recognition device is characterized in that the work determination means of the image processing device refers to the location information acquired by the location information acquisition device and determines that the step work is not being performed if the garbage collection vehicle is not in the predetermined location, and determines that the step work is being performed if the garbage collection vehicle is in the predetermined location.

8. A garbage collection vehicle equipped with a person recognition device according to any one of claims 1, 2, 5 to 7.

Images