KR102023270B1 - Construction equipment real-time safety monitoring apparatus and method - Google Patents

Abstract

The present invention relates to a real-time safety monitoring device of construction equipment. Device for real-time safety monitoring of construction equipment according to the present invention, the data storage unit for storing the image data about the image image obtained from the camera installed on one side of the construction equipment; An image editing unit correcting the distorted image by the installation angle of the camera or the distortion of coordinates, and extracting a region of interest based on an image image having a predetermined height and width in a coordinate plane on which an operator or an object is placed; An image reading unit determining whether safety equipment to be worn by an operator exists from the image image extracted by the image editing unit based on a control signal of the controller; And an image display unit which displays the determination result of the image reading unit on the display device of the construction equipment based on a control signal of the control unit.

Description

Construction equipment real-time safety monitoring apparatus and method

The present invention relates to an apparatus and method for real-time safety monitoring of construction equipment, and more particularly, because of a blind area that is not secured in real time to the driver's field of view when driving construction equipment, an accident occurs to an operator or a driver. A device and method for real-time safety monitoring of construction equipment that informs a driver that there are workers or objects present in the buried zone, etc., so that they can be suppressed.

According to 2012 deaths in construction industry statistics, more than 25% of all accidents related to the operation of construction equipment in construction sites account for more than 25%. In addition, the number of deaths related to operating construction equipment in construction sites was 91 in 2012, an increase of about 38% over 66 in 2008. According to the statistics of the Ministry of Employment and Labor, the economic loss from industrial accidents in 2012 accounted for about 33% of the total cost of industrial accidents.

Based on the ratio of the number of fatalities injured (357) related to construction equipment operation out of the number of fatalities injured in construction sites in 2012 (91), the economic cost of industrial accidents related to the operation of construction equipment It is estimated to amount to about 1.6 trillion won (Korea Occupational Safety and Health Agency, 2012).

As this trend is expected to increase every year, there has been a need for the development of construction equipment safety technology to prevent safety accidents related to the operation of construction equipment in the construction site. According to the Construction Safety Accident Statistics of the Occupational Safety and Health Administration (OSHA), 293 serious casualties were caused by industrial trucks.

For this reason, as shown in Fig. 1, 57.7% of the work environment-related factors such as blindness due to blind area or collision with loads and other workers accounted for 57.7% (Korea Occupational Safety and Health). Industrial Complex, 2008). In order to prevent safety accidents related to the operation of construction equipment in the construction site caused by factors related to the work environment, it is necessary to develop a technology capable of real-time detection of risk factors throughout the construction equipment working environment.

Accordingly, a number of technologies for real-time detection of hazards in the working environment of construction equipment have been proposed. The proposed method is based on distance sensors such as ultrasonic sensors, sensors, time of arrival (ToA), radio frequency identification (RFID) technology, pyroelectric infrared ray (PIR) sensors, and two-dimensional laser scanners. It is divided into a method for recognizing a worker by using an image sensor and an image sensor.

In the "Heavy Duty Worker Approach Detection Device (Application No .: 10-1997-0066678)" and "Proximity Alarming System (Application No .: 10-2001-0014181)", ultrasonic sensors are used to recognize workers in the work space of construction equipment. In the construction equipment safety alarm system (application number: 20-2005-0000816), a detection sensor was used to recognize workers in the construction equipment workspace. “Tunnel construction site equipment-manpower safety management system (application number) : 10-2011-0031713) ”and“ Safety Accident Prevention System at Construction Sites (10-2011-0065579) ”attempted to recognize workers in the construction equipment workspace by using a ToA (Time of Arrival) transceiver.

The Heavy Construction Approach Monitoring System (Application No. 10-2013-0033357) used PIR (Pyroelectric Infrared Ray) sensors and ultrasonic sensors to identify workers in the construction equipment workplace. Development of detection technology (So Ji-yoon et al., 2010) ”attempted to recognize workers in the construction equipment workspace using a two-dimensional laser scanner.

As such, the method for recognizing a worker using a distance sensor is a method using a ultrasonic sensor, a sensor, and a pyroelectric infrared ray (PIR) sensor. About 12m) There is a disadvantage that the operator can not be recognized. In addition, the method using RFID (Radio Frequency Identification) technology has a disadvantage in that the detection distance is sufficient, but all construction equipment and workers in the construction site must be tagged. In the case of a method using a two-dimensional laser scanner, the detection distance is sufficient, but the cost of mounting the sensor may be several million to tens of millions of won.

On the other hand, the video image obtained by using an image sensor from the construction site has the advantage that it can be used to recognize the workers in the working radius of the construction equipment (about 12m) at a relatively economical cost. Hat Detection in Video Sequences based on Face Features, Motion and Color Information (Du et al., 2011) ”, extracts moving pixels from video images acquired using image sensors from the construction site, Attempted to recognize the worker.

However, according to the operator identification method using such an image sensor, since the operator or the object has been discriminated by recognizing a specific type of object or a human image, the operator who takes various types of operations or the object of various shapes can be easily read. As the failure occurs frequently, there was a problem causing a safety accident.

The present invention has been made to solve the above problems, an object of the present invention is to clarify in real time to the driver whether a worker or an object is present around the construction equipment, especially in the blind area when the construction equipment is operating. It is an object of the present invention to provide a real-time safety monitoring device and method for construction equipment that can be distinguished and informed.

Real-time safety monitoring device for construction equipment according to the present invention for achieving the above object, Data storage unit for storing the image data about the image image obtained from the camera installed on one side of the construction equipment; An image editing unit correcting the distorted image by the installation angle of the camera or the distortion of coordinates, and extracting a region of interest based on an image image having a predetermined height and width in a coordinate plane on which an operator or an object is placed; An image reading unit determining whether safety equipment to be worn by an operator exists from the image image extracted by the image editing unit based on a control signal of the controller; And an image display unit which displays the determination result of the image reading unit on the display device of the construction equipment based on a control signal of the control unit.

Preferably, the image reading unit determines whether the image image corresponds to an operator using the color of the safety equipment.

The image reading unit may include a first reading unit determining whether the image image primarily includes safety equipment, and secondly, when the image image does not include safety equipment as a result of the determination of the first reading unit. It is preferable to include a second reading unit for determining whether a person or an object.

Preferably, the second reader compares a data value corresponding to the video image with a preset setting value with respect to the video data of the person, and determines whether the video image is about a person.

The setting values are classified according to standing, walking, bowing and sitting of the person, and the second reading unit reads the video image according to whether the video image corresponds to the classified view. desirable.

The second reading unit determines whether any one of the image images extracted by the image editing unit is located in front of the other despite being smaller in height than the other, and based on the result, one of the image images K may be further configured to determine whether the person corresponds to a sitting person.

The image editing unit acquires an image corrected by correcting coordinates distorted by the image acquisition angle of the camera to actual plane coordinates at which the operator is located, removes a shadow from the acquired image, and then stores the actual plane coordinates. It is desirable to classify objects having a predetermined height and width into different colors, and to extract the matched objects as a region of interest.

Real-time safety monitoring method of the construction equipment to achieve the above object, Data acquisition step of storing and securing the image data about the image image acquired from the camera installed on one side of the construction equipment; An image extraction step of correcting the image data secured in a distorted state by an installation angle of the camera or a distortion of coordinates, and extracting a region of interest based on an image image having a predetermined height and width in a coordinate plane; An image reading step of determining whether there is a safety tool to be worn by a worker from the extracted image image and whether the safety tool is a second person or an object when the safety image is not included in the image image; And displaying a predetermined mark according to the read result on the display device of the construction equipment based on a control signal of the control unit.

The real-time safety monitoring device for construction equipment according to the present invention having the configuration as described above is based on the image image having a certain height and width in the plane where the worker or the object is located, if the predetermined reference value exceeds all By classifying, but eliminating non-reading image noise such as shadows prior to reading, you can accurately and quickly read only those objects that are likely to pose a risk or expose you to real-time results. As it can be given, it has the effect of significantly reducing the occurrence of safety accidents.

According to the exemplary embodiment in which the image reading unit includes the first reading unit and the second reading unit, the first reading unit may notify the driver in real time whether the operator is located in the blind zone according to whether or not the safety equipment is used. Of course, even if the acquired video image does not correspond to the safety equipment, the second reading unit is configured to read out the video image separately as a person or an object, so that the driver of the construction equipment sees the marking displayed on the display device, The advantage of enabling the operation to apply an appropriate control signal to the drive unit is described, which can significantly reduce accidents caused by inadvertent damage or other objects to people who do not wear safety equipment.

1 is a view for explaining the problems of the prior art.
Figure 2 is a block diagram of the configuration of a real-time safety monitoring device of construction equipment according to an embodiment of the present invention.
3 is a flowchart illustrating a process flow of an embodiment of the present invention.
4 is a flowchart for explaining a monitoring method according to an embodiment of the present invention.
5 is a view for explaining an implementation of some components constituting an embodiment of the present invention.
6 is a view for explaining an example of the use of one embodiment of the present invention.
7 is a view for explaining an embodiment of the first reader adopted in one embodiment of the present invention.
8 is a view for explaining an embodiment of a second reader adopted in one embodiment of the present invention.

Hereinafter, a real-time safety monitoring device for construction equipment according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

Figure 2 is a block diagram of the configuration of a real-time safety monitoring device for construction equipment according to an embodiment of the present invention, Figure 3 is a flow chart for explaining the process flow of an embodiment of the present invention, Figure 4 is an embodiment of the present invention 5 is a flowchart for explaining a monitoring method according to an example, and FIG. 5 is a view for explaining an implementation of some components of an embodiment of the present invention, and FIG. 6 is for explaining an example of use of an embodiment of the present invention. 7 is a view for explaining an embodiment of the first reader adopted in one embodiment of the present invention, Figure 8 is a view for explaining an embodiment of the second reader adopted in an embodiment of the present invention. .

As shown in FIG. 2, a real-time safety monitoring apparatus for construction equipment according to an embodiment of the present invention includes a data storage unit 2, an image editing unit 3, an image reading unit 5, and an image display unit 6. )

The data storage unit 2 is a portion in which image data regarding an image image acquired from the camera 100 installed on one side of construction equipment is stored as shown in FIG. 6. Here, the image acquisition unit 1 for acquiring the image may be implemented by various image sensors, without being limited to the camera 100.

As illustrated in FIG. 5, the image editing unit 3 displays a video image stored in the data storage unit 2 in a state of being distorted due to an installation angle of the camera 100 or a distortion of coordinates. Corrected by the unwarping method, it is possible to measure the height or width on the plane where the real worker or the object is located, delete the shadows placed on this plane, and the image image with the constant height and width on the plane. It extracts a region of interest to be read based on.

Here, the advantages derived by the image editing unit employed in the embodiment of the present invention will be described in comparison with the prior art.

That is, according to the related art, although there is a similarity to one embodiment of the present invention in identifying a person or an object located in a blind zone by an image acquisition means such as the camera 100, the operation of a person or As it is configured to recognize a specific shape, there is a drawback of not recognizing a worker or an object that is not a specific shape, such as a work of moving a hand mainly.

However, according to the present embodiment, to compensate for this disadvantage, the image, such as a shadow corresponding to noise (no height as it is located on the same plane as the plane on which the worker is located) is deleted from the first acquired image image, It is configured to classify all objects as read only when a certain level is exceeded based on an image image having a certain height and width in the plane where the worker is located, so that only those objects that are likely to cause a risk or may be exposed to a risk are precisely classified. Readability can significantly reduce the occurrence of safety accidents.

 The image reading unit 5 is a part for determining whether safety equipment to be worn by an operator exists from the image image extracted by the image editing unit 3 based on the control signal of the control unit 4. The presence or absence of the safety equipment can be distinguished by various methods, such as the type and shape of the safety equipment, when considering the type or shape of the safety equipment, considering that a lot of time and power is consumed in the data processing, as shown in FIG. As noted, it is desirable to distinguish by the color of the acquired image.

The image display unit 6 serves to display the determination result of the image reading unit 5 on the display device of the construction equipment 200 based on the control signal of the control unit 4.

Real-time safety monitoring device for construction equipment according to an embodiment of the present invention having such a configuration, the target to read all when a certain reference value is exceeded based on a video image having a certain height and width in the plane where the worker or the object is located Classified as, but by removing the non-reading image noise such as shadows prior to reading, it is possible to accurately and quickly read only the objects that are likely to cause or expose the risk, and then display the results in real time to the driver As it can be made, it can derive the advantage that can significantly reduce the occurrence of safety accidents.

As illustrated in FIG. 7, the image reading unit 5 employed in the present embodiment is configured to determine whether the image image corresponds to an operator using the color of the safety equipment.

According to the embodiment having such a configuration, it is determined whether or not the operator is located in the blind zone by the color of the safety equipment regardless of the shape or type of the safety equipment, so that the color information is obtained from the image image acquired by the camera 100. It is expected to be easy to identify and to relatively easily implement a program for reading safety equipment, and to have the advantage of not causing errors in failing to identify safety equipment.

3 and 4, the image reading unit 5 employed in the present embodiment may include a first reading unit 51 that determines whether the image image primarily includes safety equipment. In addition to the reading by the first reading unit 51, a second reading unit 52 which secondly determines whether a person or an object is included.

The present embodiment having such a configuration enables the first reading unit 51 to inform the driver in real time whether the operator is located in the blind zone according to whether or not the safety equipment is used. Even if it does not correspond to the safety equipment, the second reading unit 52 is configured to read the image image to distinguish whether it is a person or an object, so that the driver of the construction equipment 200 sees the mark displayed on the display device, which is an actuator. It is possible to perform an operation for applying an appropriate control signal to the drive unit 7, so that an effect that can significantly reduce the occurrence of accidents caused by people or other objects not wearing safety equipment can be expected.

The second reader 52 is configured to compare a data value corresponding to the video image with a preset setting value with respect to the video data of the person, and determine whether the video image is about a person. desirable.

That is, as illustrated in FIG. 8, the set values are classified according to, for example, standing, walking, bowing and sitting of the person, and the second reading unit 52 is the image image. Is preferably configured to read the video image according to whether or not corresponds to the sorted appearance.

According to the present embodiment having such a configuration, whether or not the first reading unit 51 that reads the safety equipment and the image image acquired separately from the first reading unit 51 is a person's work behavior type. The second reading unit 52, which is classified and read in detail, is organically coupled to each other, thereby distinguishing between the worker wearing the safety tool, the non-wearing worker, and the non-human object, and reading the read result with different marks. Displayed distinctly on the image display section 6 (see the lowest block in FIG. 5), the driver can actively take action according to each reading result.

And, the second reader 52 is located in the front (near to the construction equipment) any one of the video images extracted by the video editing unit 3, despite the height is smaller than the other It can be configured to determine whether.

That is, according to the principle that the image is acquired from the camera 100, it is natural that the height of the image of the person located in front is higher when several workers stand in the same way based on the distortion correction picture of FIG. Will do. In spite of this, if the height of the image of the person located in front is low, it is necessary to further read whether the video image having the low height corresponds to the sitting state of the person.

According to this embodiment, all images acquired by the first reading unit 51 and the second reading unit 52 are not classified by comparison with the set values set for reading, and in the case of images belonging to a specific range, By comparing only the set values and being able to read them, it is expected to further increase the real-time safety monitoring efficiency of the construction equipment 200 according to the improvement of data processing efficiency.

Then, the image editing unit 3 employed in the present embodiment corrects the coordinates distorted by the image acquisition angle of the camera 100 to the actual plane coordinates where the worker is located, as shown in the middle block of FIG. 5. Acquire the corrected image, remove the shadow from the acquired image, classify the objects having a certain height and width in the actual plane coordinates by different colors, and classify the matched objects into the region of interest. Extract.

In this embodiment having such a configuration, the image editing unit 3 is configured to more clearly classify the object to be read and then read out, thereby not reading the unnecessary image corresponding to the object or the noise, which is the object to be read. Only images can be read intensively, further improving the precision and speed of real-time reading of the operator or object.

On the other hand, in the monitoring method using the real-time safety monitoring device of the construction equipment, as shown in Figure 3, to acquire the image data from the camera 100 mounted on the rear of the construction equipment 200, the image data After extracting the region of interest that is the object of image editing and reading through correction of distortion correction and shadow deletion, and determining whether there is a safety tool in the extracted image, the image image is secondary It is configured to read whether it is a person or an object and to display a mark corresponding to the reading result on the display device of the driver and to apply a control signal for driving the actuator of the construction equipment.

Although various embodiments of the present invention have been described above, the present embodiment and the accompanying drawings are only clearly showing a part of the technical spirit included in the present invention, and are included in the specification and drawings of the present invention. Modifications and specific embodiments that can be easily inferred by those skilled in the art within the scope of the technical idea will be apparent to be included in the scope of the present invention.

1: Image Acquisition Unit 2: Data Storage Unit
3: image editor 4: controller
5: Image Reader 51: First Reader
52: Second reading part 6: Image display part
7: drive unit 100: camera
200: construction equipment

Claims (8)

A data storage unit for storing image data regarding an image image acquired from a camera installed at one side of construction equipment;
An image editing unit correcting the distorted image by the installation angle of the camera or the distortion of coordinates, and extracting a region of interest based on an image image having a predetermined height and width in a coordinate plane on which an operator or an object is placed;
An image reading unit determining whether safety equipment to be worn by an operator exists from the image image extracted by the image editing unit based on a control signal of the controller; And
And an image display unit which displays the determination result of the image reading unit on a display device of the construction equipment based on a control signal of the control unit.
The image editing unit,
After obtaining the corrected image by correcting the coordinates distorted by the image acquisition angle of the camera to the actual plane coordinates at which the worker is located, removing the shadow from the acquired image, and then fixing the height and width to the actual plane coordinates. Real-time safety monitoring device for construction equipment, characterized in that to match and classify the objects with different colors, and to extract the matched object to the region of interest. The method of claim 1,
The image reading unit, the real-time safety monitoring device for construction equipment, characterized in that by using the color of the safety equipment determines whether the image image corresponds to the worker. The method of claim 1,
The image reading unit may include a first reading unit determining whether the image image primarily includes safety equipment, and secondly, when the image image does not include safety equipment as a result of the determination of the first reading unit. Real-time safety monitoring device for construction equipment, characterized in that it comprises a second reader for determining whether a person or an object. The method of claim 3,
The second reader may compare the data value corresponding to the video image with a preset setting value with respect to the video data of the person, and determine whether the video image is about a person. Real time safety monitoring device. The method of claim 4, wherein
The setting value is classified according to the person standing, walking, upper body and sitting state,
The second reader, the real-time safety monitoring device for construction equipment, characterized in that for reading the video image according to whether the video image corresponds to the sorted state. The method of claim 5,
The second reading unit determines whether any one of the image images extracted by the image editing unit is located in front of the other despite being smaller in height than the other, and based on the result, one of the image images The apparatus for real-time safety monitoring of construction equipment, characterized in that it further determines whether or not corresponds to the person sitting. delete A data securing step of storing and securing image data about an image image acquired from a camera installed at one side of construction equipment;
An image extraction step of correcting the image data secured in a distorted state by an installation angle of the camera or a distortion of coordinates, and extracting a region of interest based on an image image having a predetermined height and width in a coordinate plane;
An image reading step of determining whether there is a safety tool to be worn by a worker from the extracted image image and whether the safety tool is a second person or an object when the safety image is not included in the image image; And
And displaying a predetermined mark according to the read result on the display device of the construction equipment based on a control signal of the controller.
The video extraction step,
After obtaining the corrected image by correcting the coordinates distorted by the image acquisition angle of the camera to the actual plane coordinates at which the worker is located, removing the shadow from the acquired image, and then fixing the height and width to the actual plane coordinates. And classifying objects with different colors into different colors, and extracting the matched objects into a region of interest.

Images