KR102591602B1 - Fixed Approach Alarm Apparatus - Google Patents

Abstract

The present invention provides a stationary approach alarming device which increases the recognition rate for moving objects and workers by incorporating a camera and artificial intelligence so as to safely protect workers from collision accidents. The stationary approach alarming device of the present invention comprises: a camera module which is installed on a wall that is spared apart from an area of interest at risk of a collision by a predetermined distance, recognizes a worker approaching the area of interest, and captures images of the worker; a main body which is installed to be spaced apart from the location where the camera module is installed and analyzes image data and sensor data transmitted from the camera module to detect the worker and the risk of collision; an alarm which is installed to be spaced apart from the camera module, receives a control signal from the main body when the worker enters the area of interest, and notifies the worker of a dangerous situation through an alarm sound; and a warning light which is installed on one side of the alarm, receives the control signal from the main body when the worker enters the area of interest, and notifies the worker of a dangerous situation by lighting a warning light at the same time as the alarm.

Description

Fixed Approach Alarm Apparatus}

The present invention relates to a fixed approach warning device, and more specifically, to a fixed approach warning device that detects a worker recognized by a camera module installed in a preset area of interest using artificial intelligence and notifies the risk of collision between a moving object and the worker with an alarm and warning light. It's about alarm devices.

Figure 1(a) shows the status of fatal accidents in the construction industry by major disaster types in 2022, Figure 1(b) shows the status of fatal accidents in the construction industry by major disaster types in 2022, and Figure 1(c) shows the status of fatal accidents in the construction industry by major disaster types in 2022. Indicates the status of fatal accidents in other industries by accident type. As shown in Figures 1(a) to 1(c), various types of fatal accidents occur in industrial sites. Among them, there are a total of 63 collision-related accidents (hereinafter referred to as collision accidents), which is the third most common accident, accounting for 10.3% of all accidents. Additionally, collision accidents are the second most common accident in other industries (excluding construction and manufacturing).

Various types of crashes occur in industrial settings. Not only do collisions occur with various equipment or machines, but collision accidents involving collisions with buildings, structures, materials, or parts also occur, so appropriate safety measures are needed for each type. The types of crashes that frequently occur in industrial settings are as follows. 1) Collisions with various equipment and machinery such as forklifts, cranes, excavators, mobile trucks, cargo vehicles, etc., 2) Collisions with various buildings and structures such as H-beams, doors, etc., 3) Various materials such as coiled steel plates, pipes, etc. There are collisions with parts, 4) collisions with portable tools such as hammers, hammers, etc.

In particular, among the types of collision accidents in industrial sites, collisions with various equipment and machinery are prevented by attaching sensor devices, but the recognition rate of sensors for moving objects and workers is low, so improvement is needed.

Republic of Korea Patent Publication No. 10-2018-0101665 Republic of Korea Patent Publication No. 10-2020-0030802 Republic of Korea Patent Publication No. 10-2366048 Republic of Korea Patent Publication No. 10-2468707

The present invention is intended to solve the above problems, and aims to provide a fixed approach warning device that can safely protect workers from collision accidents by increasing the recognition rate for moving objects and workers by combining cameras, sensors, and artificial intelligence. do.

In order to achieve the above object, a fixed approach warning device according to an embodiment of the present invention is installed on a wall spaced at a certain distance from an area of interest at risk of a collision accident, and recognizes workers approaching the area of interest. A camera module that takes images; a main body installed at a distance from the location where the camera module is installed, and detecting the risk of collision by detecting a worker by analyzing image data and sensor data transmitted from the camera module; An alarm is installed to be spaced apart from the camera module, receives a control signal from the main body when the operator enters the area of interest, and informs the operator of a dangerous situation through an alarm sound. It is installed on one side of the alarm, and the operator is connected to the area of interest. When entering the area of interest, a control signal is received from the main body and a warning light is provided at the same time as the alarm to inform the worker of a dangerous situation.

The camera module includes a sensor unit that recognizes a worker approaching the area of interest with a sensor, a camera unit that captures an image of a worker approaching the area of interest, and sensor data generated by the sensor unit and captured by the camera unit. It includes a camera communication unit that transmits image data to the main body through wireless communication.

The main body includes a main body receiver that receives sensor data and image data transmitted from the camera module, an artificial intelligence unit that detects the risk of collision by analyzing the sensor data and image data received from the main body receiver, and detects the risk of collision. It includes a main body transmitter that transmits a control signal to the alarm and warning light when entering the area of interest, an area of interest setting unit that can set the area of interest, and a display unit that displays the image of the worker detected by the artificial intelligence unit.

The artificial intelligence unit includes a data learning unit that learns data so that artificial intelligence can recognize the worker, and the camera unit that artificial intelligence analyzes based on sensor data received from the main body receiver and content learned by the data learning unit. It includes an object detection unit that detects a worker by comparing image data, and a collision risk detection unit that generates a control signal to control the alarm and warning lights when the worker detected by the object detection unit detects entering the area of interest.

The object detection unit is characterized in that it determines and detects the object as a worker only when at least one of the sensor data or the image data of the camera unit analyzed by the artificial intelligence recognizes the object as a worker.

In addition, in order to achieve the above object, an alarm method using a fixed approach alarm device according to an embodiment of the present invention includes a first step of setting the area of interest by the area of interest setting unit; A second step in which an object approaches the area of interest, the object is detected by the camera module, and an image of the object is captured; A third step in which the sensor data and image data for the object are analyzed by the artificial intelligence unit and the worker is detected; a fourth step in which the operator contacts a boundary of the region of interest to enter the region of interest; And a fifth step in which the collision risk detection unit generates a control signal, the alarm emits a warning sound at a 3-second interval, and the warning light emits a warning light at a 3-second interval.

In addition, a sixth step in which the operator enters the region of interest after the fourth step and is located inside the region of interest; And a seventh step in which the collision risk detection unit generates a control signal, the alarm emits a warning sound at a 1-second interval, and the warning light emits a warning light at a 1-second interval. Here, if the worker enters the interior from the border of the area of interest, it has the effect of conveying the danger more urgently.

In addition, in order to achieve the above object, an alarm method using a fixed approach alarm device according to another embodiment of the present invention includes a first step of setting the area of interest by the area of interest setting unit; A second step in which an object approaches the area of interest, the object is detected by the camera module, and an image of the object is captured; A third step in which the sensor data and image data of the object are analyzed by the artificial intelligence unit and the worker is detected; A fourth step in which the worker enters the area of interest and is located inside the area of interest; And a fifth step in which the collision risk detection unit generates a control signal, the alarm emits a warning sound at a 1-second interval, and the warning light emits a warning light at a 1-second interval.

Here, if the area of interest is placed within the shooting section of the camera module, and there is a blind area of the operator outside the area of interest, if there is a blind area of the operator on the border of the area of interest, and if there is a blind area of the operator inside the area of interest, When classified into a safe state, a semi-critical state, and a dangerous state, the warning sound generation cycle of the alarm in a semi-critical state and the warning light generation cycle of the warning light may proceed faster than the warning sound generation cycle of the alarm in the dangerous state and the warning light generation cycle of the warning light. You can.

As described above, the present invention has the advantage of preventing collision accidents in industrial sites by increasing the recognition rate for moving objects and workers by combining cameras and artificial intelligence.

Figure 1(a) shows the status of fatal accidents by major disaster types in 2022, Figure 1(b) shows the status of fatal accidents in the construction industry by major disaster types in 2022, and Figure 1(c) shows the status of fatal accidents in the construction industry by major disaster types in 2022. Indicates the status of fatal accidents in other industries by accident type.
Figure 2 is a configuration diagram of a fixed approach warning device according to an embodiment of the present invention.
Figure 3 is a configuration diagram of a camera module according to an embodiment of the present invention.
Figure 4 is a configuration diagram of the main body according to an embodiment of the present invention.
Figure 5 is a configuration diagram of an artificial intelligence unit according to an embodiment of the present invention.
Figure 6 shows that the fixed approach warning device according to an embodiment of the present invention fails to detect the worker in the area of interest.
Figure 7 shows that a fixed approach warning device according to an embodiment of the present invention detects a worker entering an area of interest.
Figure 8 shows a fixed approach warning device according to an embodiment of the present invention detecting the presence of a worker in the area of interest.
Figure 9 shows a fixed access alarm device according to an embodiment of the present invention installed in a factory.
Figure 10 shows a fixed approach alarm device installed around a packaging machine according to an embodiment of the present invention.
Figure 11 shows a fixed access warning device installed at a construction site according to an embodiment of the present invention.
Figure 12 shows a fixed access warning device according to an embodiment of the present invention installed in an actual industrial site.
Figure 13 is a flowchart of an alarm method using a fixed approach alarm device according to an embodiment of the present invention.
Figure 14(a) shows the first step in the warning method using a fixed access warning device according to an embodiment of the present invention, Figure 14(b) shows the third step, and Figure 14(c) shows the fifth step. represents.
Figure 15 is a flowchart of an alarm method using a fixed approach alarm device according to another embodiment of the present invention.

Since the present invention can make various changes and have various embodiments, specific embodiments will be illustrated in the drawings and described in detail in the detailed description. However, this is not intended to limit the present invention to specific embodiments, and should be understood to include all changes, equivalents, and substitutes included in the spirit and technical scope of the present invention. While describing each drawing, similar reference numerals are used for similar components.

Terms such as first, second, A, B, etc. may be used to describe various components, but the components should not be limited by the terms. The above terms are used only for the purpose of distinguishing one component from another.

For example, a first component may be named a second component, and similarly, the second component may also be named a first component without departing from the scope of the present invention. The term “and/or” includes any of a plurality of related stated items or a combination of a plurality of related stated items.

When a component is said to be "connected" or "connected" to another component, it is understood that it may be directly connected to or connected to the other component, but that other components may exist in between. It should be. On the other hand, when it is mentioned that a component is “directly connected” or “directly connected” to another component, it should be understood that there are no other components in between.

The terms used in this application are only used to describe specific embodiments and are not intended to limit the invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this application, terms such as “comprise” or “have” are intended to designate the presence of features, numbers, steps, operations, components, parts, or combinations thereof described in the specification, but are not intended to indicate the presence of one or more other features. It should be understood that this does not exclude in advance the possibility of the existence or addition of elements, numbers, steps, operations, components, parts, or combinations thereof.

Unless otherwise defined, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by a person of ordinary skill in the technical field to which the present invention pertains. Terms defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related technology, and unless explicitly defined in the present application, should not be interpreted in an ideal or excessively formal sense. No.

Hereinafter, a fixed access warning device according to an embodiment of the present invention will be described with reference to the drawings. Figure 2 is a configuration diagram of a fixed approach warning device according to an embodiment of the present invention. Referring to Figure 2, the fixed access warning device according to an embodiment of the present invention may be configured to include a camera module 100, a main body 200, an alarm 300, and a warning light 400.

First, the camera module 100 is installed on the upper part of the wall in the area of interest where there is a risk of a collision accident, and recognizes and photographs workers approaching the area of interest. The present invention is used to recognize and photograph multiple areas of interest. A plurality of camera modules 100 may be provided. Figure 3 is a configuration diagram of a camera module 100 according to an embodiment of the present invention. Referring to FIG. 3, a camera module according to an embodiment of the present invention may be configured to include a sensor unit 110, a camera unit 120, and a camera communication unit 130. The camera module 100 of the present invention is preferably installed on the top of a wall at a certain distance from a preset area of interest so that it can recognize workers approaching the area of interest.

The sensor unit 110 recognizes the operator approaching the area of interest with a sensor. The sensor unit 110 according to an embodiment of the present invention may use a LiDAR (Light Detection And Ranging, LiDAR) sensor. LiDAR sensor is a technology that can detect the distance, direction, speed, temperature, material distribution, and concentration characteristics of an object by shining a laser on the target. LiDAR sensors generally have high energy density and short cycle pulses. By utilizing the advantage of a laser that can generate signals, it can be used for more precise observation of physical properties in the atmosphere and distance measurement. The sensor unit 110 can recognize the shape of the object or worker by shining a laser on the area of interest and measuring the time it takes for the laser to be reflected from an object or worker approaching the area of interest. Additionally, the sensor data generated after the sensor unit 110 recognizes the shape is transmitted to the main body 200.

The camera unit 120 captures an image of a worker approaching the area of interest. The camera unit 120 of the present invention can use a camera with a 140° wide viewing angle to photograph a wide area. The camera unit 120 photographs objects or workers approaching the area of interest and transmits the captured images to the main body 200.

The camera communication unit 130 transmits sensor data generated by the sensor unit 110 and image data captured by the camera unit 120 to the main body 200 through wireless communication. The wireless communication of the present invention can be used by selecting one of short-range wireless communication, mid-range wireless communication, and long-distance wireless communication. Short-range wireless communication can be one of Bluetooth, Zigbee, Wi-Fi, IrDA, HomeRF, WLAN, WPAN, EnOcean, and Wi-SUN, and long-distance wireless communication can be LoRaWAN, SigFox, LTE-M, NB-IoT, and 5G. It may be one of the following, and in the present invention, sensor data and control data are transmitted and received according to at least one wireless communication technology, including wireless communication technologies not listed above.

Next, the main body 200 is installed away from the location where the camera module 100 is installed, and analyzes sensor data and image data transmitted from the camera module 100 to detect the worker and detect the risk of collision. Figure 4 is a configuration diagram of a main body according to an embodiment of the present invention. Referring to FIG. 4, the main body 200 according to an embodiment of the present invention includes a main body receiving unit 210, an artificial intelligence unit 220, a main body transmitting unit 230, a region of interest setting unit 240, and a display unit 250. ) may be configured to include.

The main body receiver 210 receives sensor data and image data transmitted from the camera communication unit 130. The main body receiver 210 and the camera communication unit 130 are connected through wireless communication.

The artificial intelligence unit 220 analyzes sensor data and image data received from the main body receiver 210 and detects the worker. Figure 5 is a configuration diagram of the artificial intelligence unit 220 according to an embodiment of the present invention. Referring to FIG. 5, the artificial intelligence unit 220 according to an embodiment of the present invention may be configured to include a data learning unit 221, an object detection unit 222, and a collision risk detection unit 223. In the present invention, artificial intelligence is used to detect workers from sensor data and image data received from the camera module 100. In the present invention, either deep learning or machine learning can be selected and used.

In object recognition using deep learning, a deep learning model such as a convolutional neural network (CNN) is used to automatically learn the unique characteristics of the object to identify it. For example, CNN can learn how to identify the differences between workers and moving objects by analyzing thousands of training images and learning features that distinguish workers from moving objects. Two methods perform object recognition using deep learning. There are several approaches.

First, to train a deep network from the ground up, collect a very large labeled data set, design a network architecture, learn features, and complete the model. This can produce excellent results, but this approach requires large amounts of training data and sets up layers and weights in the CNN.

And as a method of using pre-trained deep learning models, most deep learning applications use transfer learning, a process that involves fine-tuning the pre-trained model. This approach uses an existing network, such as AlexNet or GoogLeNet, to inject new data containing previously unknown classes. This method reduces time consumption and produces results quickly by pre-training the model with thousands or millions of images.

Machine learning technology is widely used in object recognition, and when machine learning technology is used in a different approach from deep learning, HOG feature extraction using the SVM (Support vector machine) machine learning model and features such as SURF and MSER are used. It is the first object detection framework to provide real-time object detection speed with the Bag of Words (BoW) model and the Viola-Jones algorithm, which can be used to recognize a variety of objects, including faces and upper bodies.

Additionally, to perform object recognition using standard machine learning methods, you gather images and select key features from each image. For example, feature extraction algorithms extract edge or corner features from the data that can be used to distinguish between classes. Then, these features are added to the machine learning model, dividing each feature into a unique category, and this information is used when analyzing and classifying new objects. Various machine learning algorithms and feature extraction methods are used to create an accurate object recognition model. Can be used in combination. When machine learning is used for object recognition, various features and classifiers can be optimally combined for learning, and accurate results can be obtained even with minimal data. In the present invention, it is more preferable to use machine learning because it is known which features of the image are best used to distinguish between classes of objects.

The data learning unit 221 learns data so that artificial intelligence can recognize the worker. In the present invention, the data learning unit 221 is provided with enough labeled data already learned.

The object detection unit 222 detects the worker by comparing the sensor data received from the main body receiver 210 with the image data of the camera unit 120 analyzed by artificial intelligence based on the content learned by the data learning unit 221. . The sensor data and image data of the camera module 100 may include various objects such as moving objects, workers, and objects being transported, and the object detection unit 222 uses the sensor data and the camera data based on the content learned in the data learning unit 221. The worker is detected by comparing the images captured by the unit 110. The object detection unit 222 can determine and detect a worker only when at least one of the sensor data or the image data of the camera unit 120 analyzed by artificial intelligence recognizes the worker. The object detection unit 222 analyzes the image data captured by the camera unit 120 with artificial intelligence to detect the worker. Since there may be an error in image recognition, the sensor unit 110 is used to improve more accurate detection sensitivity. The operator can be detected by using sensor data in parallel. The object detection unit 222 may determine that at least one of the sensor data recognized by the sensor unit 110 or the image data analyzed by artificial intelligence is a worker. That is, when the sensor data recognized by the sensor unit 110 recognizes an object approaching the area of interest as a worker, the object detection unit 222 determines that the object is a worker and detects it. Additionally, when image data captured by the camera unit 120 of an object approaching the area of interest is determined to be a worker by artificial intelligence, the object detection unit 222 determines the object to be a worker and detects it. Detecting workers by using the two methods in parallel as described above has the advantage of increasing detection sensitivity and overcoming problems caused by image recognition errors.

The collision risk detection unit 223 detects that the worker detected by the object detection unit 222 enters the area of interest and generates a control signal. Figure 6 shows that the fixed access warning device according to an embodiment of the present invention fails to detect the worker in the area of interest, and Figure 7 shows that the fixed access warning device according to an embodiment of the present invention fails to detect the worker entering the area of interest. 8 shows that a fixed approach warning device according to an embodiment of the present invention detects the presence of a worker in the area of interest. Referring to Figures 6 to 8, when artificial intelligence detects a worker, it recognizes this as the worker's blind area. The area where photography is possible due to the angle of view of the camera module 100 is referred to as the photography section, and the area where workers are prohibited from entering due to the risk of collision may be defined as the area of interest. If the area of interest is placed within the shooting section of the camera module 100, and there is a blind area of the operator outside the area of interest, if there is a blind area of the operator on the border of the area of interest, and if there is a blind area of the operator inside the area of interest. It can be classified into safe state, semi-critical state, and dangerous state.

The collision risk detection unit 223 does not detect the risk of collision when the worker's blind spot is outside the area of interest. However, the collision risk detection unit 223 detects the risk of collision when the worker's blind spot comes into contact with the interest area or when the worker's blind spot is located inside the interest area. When the collision risk detection unit 223 detects a collision risk, it generates a control signal to start the operation of the alarm 300 and the warning light 400. Additionally, the collision risk detection unit 223 generates a control signal to terminate the operation of the alarm 300 and the warning light 400 when the worker leaves the area of interest.

Meanwhile, the object detection unit 222 may be provided with a worker confirmation unit, and the worker confirmation unit determines whether the worker is a worker working at the location when the moving object is a worker based on data learned by artificial intelligence. is determined, and the determined signal is transmitted to the collision risk detection unit 223.

Here, the collision risk detection unit 223 may be provided with an area of interest range control unit, and the area of interest range control unit determines the area of interest as a first area of interest, includes the first area of interest, and is larger than the first area of interest. It can be divided into a second region of interest formed in a wide area. For example, the first region of interest and the second region of interest may have an area ratio of 1:1.1 to 1.5. If the worker information of the moving object transmitted from the worker confirmation unit is a worker working in the relevant area of interest, the area of interest is set as the first area of interest, and if the worker's information is not a worker working in the relevant area of interest, the area of interest is set. can be set as the second area of interest. Therefore, in the case of workers who do not work in the area, they are not familiar with the unfamiliar environment of the area of interest, so they can set the area of interest wider than workers working in the area, thereby raising awareness of work safety.

The main body transmitter 230 transmits a control signal to the alarm 300 and the warning light 400 when the worker enters the area of interest. The main body transmitter 230, the alarm 300, and the warning light 400 are connected through wireless communication. When the collision risk detection unit 223 detects a risk of collision, the main body transmitter 230 transmits a control signal for the alarm 300 and the warning light 400 to operate the alarm 300 and the warning light 400.

The region of interest setting unit 240 can set the region of interest within the shooting section. The area of interest setting unit 240 may designate an area with a high risk of collision between a worker and a moving object adjacent to the movement path of the moving object and set it as the area of interest. The administrator can check the set area of interest through the display unit 250, and the area of interest may be displayed in the form of a red solid line on the display unit 250. The area of interest may be automatically set by the area of interest setting unit 240, but the administrator must check the area of interest setting unit 240 displayed on the display unit 250 and operate the area provided at the bottom of the display unit 250. You can manually set the area of interest using the button. Meanwhile, the area of interest setting unit 240 may include a manual area of interest range setting unit, and the manager manually sets both the first area of interest and the second area of interest as the second area of interest in the area of interest range control unit, The area of interest can be set to a wider area than the first area of interest set in .

The display unit 250 displays the image of the worker detected by the artificial intelligence unit 220. The manager can check whether the artificial intelligence unit 220 properly detects the worker among the images captured by the camera unit 120 through the display unit 250. Additionally, the administrator can change the settings of the main body 200 through the display unit 250 and the operation buttons provided on the lower part of the display unit 250.

The main body 200 of the present invention may further include a worker information storage unit and a mobile device transmission unit. The worker information storage unit is inputted with the worker's telephone number information, and information about the worker's appearance is inputted based on the worker's photographic information. When the worker touches the area of interest or is determined to have entered the area of interest, the mobile device transmitter transmits a danger signal to the worker's mobile phone, and the worker receives a danger signal by vibrating or ringing the worker's mobile phone. It can be recognized.

Next, the alarm 300 is installed near the camera module 100, and when the worker enters the area of interest, it receives a control signal from the main body 200 and notifies the worker of the dangerous situation through an alarm sound. The alarm 300 notifies through sound that there is a risk of collision when a worker enters the area of interest, and can signal the danger with a siren of 100 to 150 dB. The alarm sound continues when the operator is located in the area of interest and may end when the operator leaves the area of interest. That is, when the worker leaves the area of interest, the impulse risk detection unit 223 generates a control signal that terminates the operation of the alarm 300, transmits it to the alarm 300 through the main body transmitter 230, and receives the control signal. One alarm 300 ends the alarm sound operation.

Next, the warning light 400 is installed on one side of the alarm 300, and when the worker enters the area of interest, it receives a control signal from the main body 200 and emits a warning light at the same time as the alarm 300 to alert the worker to a dangerous situation. inform. The warning light 400 operates by receiving a control signal from the main body transmitter 230 when the worker enters the area of interest. In the present invention, the warning light 400 can generate a red warning light by rotating a light located inside a red circular case by a motor provided therein. The light can rotate at a rotation speed of 10 to 150 rpm. In the present invention, the alarm 300 and the warning light 400 can be used simultaneously to provide a sufficient warning signal to workers entering the area of interest. The warning light continues when the operator is in the area of interest and can end when the operator leaves the area of interest. That is, when the worker leaves the area of interest, the impulse risk detection unit 223 generates a control signal that terminates the operation of the warning light 400, transmits it to the warning light 400 through the main body transmitter 230, and receives the control signal. One warning light 400 ends its operation.

Figure 9 shows a fixed access alarm device according to an embodiment of the present invention installed in a factory, Figure 10 shows a fixed access alarm device according to an embodiment of the present invention installed around a packaging machine, and Figure 11 shows the present invention. It shows that a fixed access warning device according to an embodiment of the present invention is installed at a construction site, and Figure 12 shows that a fixed access warning device according to an embodiment of the present invention is installed at an actual industrial site. As shown in FIGS. 9 to 12, the fixed access warning device of the present invention can be installed and used in places where moving objects enter and exit a factory, around packaging machines, and at construction sites where construction vehicles enter and exit. The fixed access warning device of the present invention can be installed and used without restrictions in places where workers exposed to the risk of collision can be protected.

The camera module 100 of the fixed access warning device of the present invention additionally includes a first mobile camera at the left edge of the front of the carrier, a second mobile camera at the right edge of the front of the carrier, and a second mobile camera at the right edge of the rear of the carrier. A third mobile camera and a fourth mobile camera are provided at the right edge of the rear of the support body, a first mobile area of interest captured by the first mobile camera, a second mobile area of interest captured by the second mobile camera, When a moving object is detected in the third moving area of interest captured by the third mobile camera or the fourth moving area of interest captured by the fourth mobile camera, the warning sound of the alarm 300 is generated, and the warning light of the warning light 400 is turned on. can be created to notify workers of danger. Accordingly, it is possible to detect danger based on images captured by the first to fourth mobile cameras in a forklift moving with the camera module 100 installed on the wall.

In addition, the fixed access warning device of the present invention further includes a plurality of warning LEDs installed on the workplace floor at regular intervals within the area of interest, and the selective warning LED control unit provided in the artificial intelligence unit includes an object detection unit. When a worker enters the area of interest, the warning LED installed in the area where the worker is located is turned on and off at regular intervals, and the warning LED in the place where the worker is not located is not turned on and off to notify the worker of the danger. Here, a plurality of warning LEDs can be installed movably on the workplace floor in the form of pads. The location of the worker can cause a plurality of warning LEDs within a diameter of 1 m to operate centered on the worker.

Additionally, the main body further includes a camera position detection unit that detects the position of the camera module. If a moving object or moving object collides with the camera and the angle of the camera deviates from the area of interest, the camera position detection unit may detect this and transmit a signal of the change in position to the display unit. For example, when a physical force is applied to the camera to change its position, the camera position detection unit detects this using a horizontal sensor provided inside and transmits a signal of the position change to the display unit to inform the camera of this change.

Hereinafter, an alarm method using a fixed approach alarm device according to an embodiment of the present invention will be described with reference to the drawings. Figure 13 is a flowchart of an alarm method using a fixed approach alarm device according to an embodiment of the present invention. Figure 14(a) shows the first step in the warning method using a fixed access warning device according to an embodiment of the present invention, Figure 14(b) shows the third step, and Figure 14(c) shows the fifth step. represents. Referring to Figures 13 and 14, the warning method using the fixed access warning device according to an embodiment of the present invention may be composed of the following five steps.

The first step (S10) is a step in which a region of interest is set by the region of interest setting unit 240. The area of interest setting unit 240 may set an area where there is a possibility of collision between a carrier or an object and a worker in the capturing section of the camera module 100 as the area of interest. The area of interest set in the area of interest setting unit 240 is displayed as a red solid line on the display unit 250, and the user can manually or automatically set the area of interest setting unit 240 through the display unit 250.

The second step (S20) is a step in which an object approaches the area of interest, the object is detected by the camera module 100, and an image of the object is captured. The sensor unit 110 can recognize an object by shining a laser on an object approaching the area of interest, and the camera unit 120 photographs the object approaching the area of interest. Sensor data recognized by the sensor unit 110 and image data captured by the camera unit 120 are transmitted to the main body 200 through the camera communication unit 130.

The third step (S30) is a step in which the artificial intelligence unit 220 analyzes the image of the object and detects the worker. The main body receiver 210, which receives the sensor data and image data, transmits the data to the artificial intelligence unit 220, and the object detection unit 222 analyzes the object based on what was learned from the data learning unit 221 and detects the object. Detect whether is a worker. The object detection unit 222 determines and detects an object as a worker only when at least one of the sensor data or the image data from the camera unit analyzed by artificial intelligence recognizes the object as a worker.

The fourth step (S40) is a step in which the worker contacts the boundary of the area of interest to enter the area of interest. When the worker touches the boundary of the area of interest, the collision risk detection unit 223 detects this. That is, the collision risk detection unit 223 can check whether the worker's blind spot is in contact with the area of interest.

In the fifth step (S50), the collision risk detection unit 223 generates a control signal, the alarm 300 emits a warning sound every 3 seconds, and the warning light 400 emits a warning light every 3 seconds. The control signal generated by the collision risk detection unit 223 is transmitted to the alarm 300 and the warning light 400 by the main body transmitter 230. In the fourth step (S40), when the worker touches the boundary of the region of interest, that is, if the worker has not entered the interior of the region of interest, a warning signal is given to prevent the worker from entering the region of interest. In the present invention, the alarm 300 emits a warning sound every 3 seconds, and the warning light 400 emits a warning light every 3 seconds to warn the worker of danger and prevent the worker from entering the area of interest.

Hereinafter, an alarm method using a fixed approach alarm device according to another embodiment of the present invention will be described with reference to the drawings. Figure 15 is a flowchart of an alarm method using a fixed approach alarm device according to another embodiment of the present invention. Referring to FIG. 15, the alarm method using the fixed approach alarm device according to another embodiment of the present invention may be composed of the following five steps.

The first step (R10) is a step in which a region of interest is set by the region of interest setting unit 240. The area of interest setting unit 240 may set an area where there is a possibility of collision between a carrier or an object and a worker in the capturing section of the camera module 100 as the area of interest. The area of interest set in the area of interest setting unit 240 is displayed as a red solid line on the display unit 250, and the user can manually or automatically set the area of interest setting unit 240 through the display unit 250.

The second step (R20) is a step in which an object approaches the area of interest, the object is detected by the camera module 100, and an image of the object is captured. The sensor unit 110 can recognize an object by shining a laser on an object approaching the area of interest, and the camera unit 120 photographs the object approaching the area of interest. Sensor data recognized by the sensor unit 110 and image data captured by the camera unit 120 are transmitted to the main body 200 through the camera communication unit 130.

The third step (R30) is a step in which the artificial intelligence unit 220 analyzes the image of the object and detects the worker. The main body receiver 210, which receives the sensor data and image data, transmits the data to the artificial intelligence unit 220, and the object detection unit 222 analyzes the object based on what was learned from the data learning unit 221 and detects the object. Detect whether is a worker. The object detection unit 222 determines and detects a worker as a worker only when at least one of the sensor data or the image data of the camera unit 120 analyzed by artificial intelligence recognizes the worker.

The fourth step (R40) is the step where the worker enters the area of interest and is located inside the area of interest. When the worker enters the area of interest, the collision risk detection unit 223 detects it. That is, the collision risk detection unit 223 can check whether the worker's blind spot has entered the area of interest.

In the fifth step (R50), the collision risk detection unit 223 generates a control signal, the alarm 300 emits a warning sound at a 1-second interval, and the warning light 400 emits a warning light at a 1-second interval. The control signal generated by the collision risk detection unit 223 is transmitted to the alarm 300 and the warning light 400 by the main body transmitter 230. In the fourth step (R40), if the worker is located inside the area of interest, that is, if the worker has already entered the inside of the area of interest, a warning signal is given to inform the worker of the imminent risk of collision with a moving object. In the present invention, the alarm 300 emits a warning sound at a 1-second interval, and the warning light 400 emits a warning light at a 1-second interval to warn the worker of the risk of collision, allowing the worker to urgently escape the area of interest.

The generation cycle of the warning sound of the alarm 300 and the flashing cycle of the warning light of the warning light 400 in a state in which the worker's blind area of the present invention is located inside the area of interest are determined when the worker's blind area is in contact with the area of interest. The warning sound generation period of the alarm 300 may be shorter than the warning light flashing period of the warning light 400. Therefore, it has the advantage of being able to distinguish and transmit more urgent danger signals when a worker enters the area of interest than when the worker is across the area of interest.

The present invention is not limited to the above-mentioned embodiments, but can be manufactured in various different forms, and those skilled in the art may manufacture the present invention in other specific forms without changing the technical idea or essential features of the present invention. You will understand that it can be done. Therefore, the embodiments described above should be understood in all respects as illustrative and not restrictive.

100: Camera module
110: sensor unit
120: camera unit
130: Camera communication unit
200: main body
210: main body receiver
220: Artificial Intelligence Department
221: Data learning department
222: object detection unit
223: Collision risk detection unit
230: main body transmitter
240: Area of interest setting unit
250: display unit
300: alarm
400: warning light

Claims (6)

A camera module installed on a wall at a certain distance from an area of interest at risk of a collision, to recognize workers approaching the area of interest and capture images; a main body installed at a distance from the location where the camera module is installed, and detecting the risk of collision by detecting a worker by analyzing image data and sensor data transmitted from the camera module; An alarm is installed to be spaced apart from the camera module, receives a control signal from the main body when the operator enters the area of interest, and informs the operator of a dangerous situation through an alarm sound. It is installed on one side of the alarm, and the operator is connected to the area of interest. When entering the area of interest, a control signal is received from the main body and a warning light is provided simultaneously with the alarm to inform the worker of a dangerous situation.
The camera module includes a sensor unit that recognizes a worker approaching the area of interest with a sensor, a camera unit that captures an image of a worker approaching the area of interest, and sensor data generated by the sensor unit and captured by the camera unit. It includes a camera communication unit that transmits image data to the main body through wireless communication,
The main body includes a main body receiver that receives sensor data and image data transmitted from the camera module, an artificial intelligence unit that detects the risk of collision by analyzing the sensor data and image data received from the main body receiver, and detects the risk of collision. When entering the area of interest, it includes a main body transmitter that transmits a control signal to the alarm and warning light, an area of interest setting unit that can set the area of interest, and a display unit that displays the image of the worker detected by the artificial intelligence unit,
The artificial intelligence unit includes a data learning unit that learns data so that artificial intelligence can recognize the worker, and the camera unit that artificial intelligence analyzes based on sensor data received from the main body receiver and content learned by the data learning unit. An object detection unit that compares image data to detect a worker, and a collision risk detection unit that generates a control signal to control the alarm and warning lights when it detects that the worker detected by the object detection unit enters the area of interest,
The object detection unit determines and detects a worker only when at least one of the sensor data or the image data of the camera unit analyzed by the artificial intelligence recognizes the object as a worker,
The object detection unit includes a worker confirmation unit, and the worker confirmation unit determines whether the worker corresponds to a worker working at the location when the moving object is a worker based on data learned by artificial intelligence, and makes the determination. transmits the signal to the collision risk detection unit,
The collision risk detection unit may include an area of interest range control unit, and the area of interest range control unit may define the area of interest as a first area of interest, and a second area that includes the first area of interest and is formed to have a wider range than the first area of interest. It is divided into areas of interest, and if the worker information of the moving object transmitted from the worker confirmation unit is a worker working in the area of interest, the area of interest is set as the first area of interest, and the worker's information is set as the first area of interest. If you are not a worker, set the area of interest as the second area of interest,
It further includes a plurality of warning LEDs installed on the workplace floor at regular intervals within the area of interest, and the selective warning LED control unit provided in the artificial intelligence unit allows the worker to enter the area of interest through the object detection unit. In this case, the warning LED installed in the place where the worker is located is turned on and off at regular intervals, the warning LED in the place where the worker is not located is not turned on and off, and the plurality of warning LEDs are in the form of pads that can be moved on the workplace floor. It is installed,
The main body further includes a camera position detection unit that detects the position of the camera module, and when a moving object or moving object collides with the camera unit and the angle of the camera unit deviates from the region of interest, the camera position detection unit detects this and In an alarm method using a fixed approach alarm device that transmits a position change signal to the display unit,
A first step of setting the region of interest by the region of interest setting unit; A second step in which an object approaches the area of interest, the object is detected by the camera module, and an image of the object is captured; A third step in which the sensor data and image data for the object are analyzed by the artificial intelligence unit and the worker is detected; a fourth step in which the worker enters the region of interest and touches a border of the region of interest; And a fifth step in which the collision risk detection unit generates a control signal, the alarm emits a warning sound at a 3-second interval, and the warning light emits a warning light at a 3-second interval.
A sixth step in which the operator enters the region of interest after the fourth step and is located inside the region of interest; And a seventh step of generating a control signal from the collision risk detection unit, wherein the alarm emits a warning sound at a 1-second interval and the warning light emits a warning light at a 1-second interval;
When the area of interest is placed within the shooting section of the camera module and the operator's blind area is outside the area of interest, it is a safe state; when there is a blind area of the worker at the border of the area of interest, it is a semi-critical state; and inside the area of interest. If there is a blind spot for the operator in a dangerous state, the cycle of generating the warning sound of the alarm and the generation cycle of the warning light of the warning light is greater than the cycle of generation of the warning sound of the alarm and the generation cycle of the warning light of the warning light in a dangerous state. An alarm method using a fixed approach alarm device, characterized in that it progresses quickly. delete delete delete delete delete