KR102115322B1 - System for smart video analysis - Google Patents

Abstract

A smart image analysis system according to an embodiment of the present invention includes an image analysis unit for setting a region of interest in each of the acquired images, and detecting an abnormal situation in the set region of interest; And an integrated management server that receives and stores the acquired image and detected abnormality information, and provides the stored image and abnormality information to a user.

Description

Smart video analysis system {SYSTEM FOR SMART VIDEO ANALYSIS}

The present invention relates to a smart image analysis system, and relates to a smart image analysis system that detects an alarm situation quickly and accurately by analyzing an acquired image and determining whether an abnormal condition has occurred.

With the current CCTV monitoring system, there is a problem that an operator (security guard) performs patrolling and monitors the monitor at the same time, so that the violation may not be noticed immediately or may be missed. This will lead to a late response in an emergency situation, and in such a situation, it is inevitable to increase the operating personnel for the security system to expect the efficiency of the monitoring function, and this will cause excessive cost expenditure to continue, resulting in cost loss for the operator operating the security system. Becomes inevitable.

In order to solve this problem, by applying various types of surveillance cameras based on various signal information acquired through various detection means, it is possible to check the abnormal conditions in each region from a remote location, thereby maximizing the maximum effect with minimum manpower. An integrated security remote monitoring system and method for providing an integrated security system are disclosed.

However, the prior art also only helps the operator (security guard) to check various types of surveillance cameras, and overcomes the limitation of monitoring by the operator (security guard) by analyzing the acquired CCTV image itself and determining whether an abnormality has occurred. There is a problem that cannot be done.

KR20160080159A

The present invention was invented to solve the above problems, and relates to a smart image analysis system that overcomes the limitations of visual surveillance by a monitoring agent by detecting an abnormal operation of a fire, intrusion, or equipment by analyzing the acquired image.

A smart image analysis system according to an embodiment of the present invention includes an image analysis unit for setting a region of interest in each of the acquired images, and detecting an abnormal situation in the set region of interest; And an integrated management server that receives and stores the acquired image and detected abnormality information, and provides the stored image and abnormality information to a user.

The image analysis unit may calculate a color feature value of the first image among the acquired images, and detect fire or smoke based on the calculated color feature value.

The image analysis unit calculates the pixel information of the region of interest set in the second image among the obtained images, and compares the calculated pixel information with the reference pixel preset in the region of interest, thereby preventing intrusion. Can be detected.

The image analysis unit may determine that an intruder has occurred when the calculated pixel information and the preset reference pixel differ by a predetermined level or more, and may track the intruder in the acquired images.

The image analysis unit determines the pattern of the region of interest set in the third image among the obtained images, and compares the determined pattern with the abnormal pattern preset in the region of interest, and the interest of the third image It is possible to detect an abnormal condition of the machinery included in the area.

The preset abnormal pattern may include an abnormal termination pattern in which the operation of the mechanical equipment is abnormally terminated, and an abnormal operation pattern in which the mechanical equipment is abnormally operated.

The integrated management server may designate regions of interest for each of the acquired images, and transmit the designated regions of interest to the image analysis unit.

The image analysis unit may receive the designated regions of interest and set an region of interest to images corresponding to each of the received regions of interest.

The integrated management server may output an alarm when the abnormality is detected, and store and store the received abnormality detected image and information about the received abnormality in pairs.

According to the smart image analysis system according to an embodiment of the present invention, there is an effect of overcoming the limitations of visual surveillance by detecting the real-time condition of an industrial facility by analyzing the acquired image.

In addition, by detecting an intrusion situation, a fire occurrence situation, etc., it is possible to protect facilities and prevent industrial accidents.

In addition, there is an effect that can quickly and accurately detect the alarm situation by checking the abnormal situation in real time.

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are included as part of the detailed description to aid understanding of the present invention, provide embodiments of the present invention, and describe the technical features of the present invention together with the detailed description.
1 is a block diagram illustrating a smart image analysis system according to an embodiment of the present invention.
2 is a view showing an embodiment of a smart image analysis system according to an embodiment of the present invention.
3 is a diagram illustrating an image analysis graph in a smart image analysis system according to an embodiment of the present invention.

In the present specification, terms such as first and / or second are used only to distinguish one component from another component. That is, it is not intended to limit the components by the terms.

Elements, features, and steps referred to in the phrase 'include' in this specification mean that the elements, features, and steps exist, and are intended to exclude one or more other elements, features, steps, and the like. This is not.

In this specification, unless otherwise specified in the singular form, the plural form is included. That is, the components and the like referred to herein may mean the presence or addition of one or more other components.

Unless defined otherwise, all terms used in this specification, including technical or scientific terms, have the same meaning as commonly understood by a person having ordinary skill in the art to which the present invention pertains (normal artisan). to be.

In other words, terms such as those defined in a commonly used dictionary should be interpreted as meanings consistent with meanings in the context of related technologies, and are interpreted as ideal or excessively formal meanings unless explicitly defined herein. Does not work.

Hereinafter, a smart image analysis system according to an embodiment of the present invention will be described in detail with reference to the accompanying drawings.

1 is a block diagram illustrating a smart image analysis system according to an embodiment of the present invention.

Referring to FIG. 1, a smart image analysis system 100 according to an embodiment of the present invention may include an image analysis device 101 and an integrated management server 103.

The image analysis device 101 may set a region of interest in each of the acquired images, and detect an abnormal situation in the set region of interest.

For example, the acquired images are images captured through a camera such as a CCTV or IPZ camera, and the image analysis device 101 can acquire images captured by cameras pre-installed in an industrial facility.

The region of interest is designated by the integrated management server 103, and the integrated management server 103 may designate regions of interest for each of the acquired images, and transmit the designated regions of interest to the image analysis unit 101. The image analysis unit 101 may receive the designated regions of interest and set the region of interest to images corresponding to each of the received regions of interest.

For example, the user can connect to the integrated management server 103 and designate the location and size of the region of interest in the acquired image. Alternatively, the integrated management server 103 may extract important areas such as industrial facilities and entrances from the acquired image, and designate the extracted important areas as areas of interest.

The images acquired by the image analysis device 101 include a first image obtained from a camera installed to detect fire or smoke, a second image obtained from a camera installed to detect intrusion, and a camera installed to detect an abnormal condition of the mechanical equipment. It may include a third image obtained from.

For example, a camera photographing the first image may be installed at a location capable of photographing the entire interior of the industrial facility, and a camera photographing the second image may be installed near the entrance of the industrial facility to detect intruders. In addition, the camera for photographing the third image may be installed at a location where an image of an industrial facility such as a coal feeder, a conveyor belt, a small hydro turbine, or a center pipe can be photographed.

The image analysis device 101 may calculate a color feature value of the first image among the acquired images, and detect fire or smoke based on the calculated color feature value.

For example, a fire reference value capable of determining whether a fire has occurred may be preset in the image analysis device 101. The image analysis device 101 may calculate a color feature value based on the color, saturation, and brightness of the first image, and compare the calculated color feature value with a fire reference value to determine whether a fire has occurred.

In addition, a smoke reference value capable of determining whether smoke is generated as well as a fire may be preset in the image analysis device 101. The image analysis device 101 may determine whether smoke is generated by comparing the color feature values calculated based on the color, saturation, and brightness of the first image with the smoke reference value.

The image analysis device 101 calculates pixel information of a region of interest set in the second image among the acquired images, and detects an intrusion based on a result of comparing the calculated pixel information with a reference pixel preset in the region of interest. Can be.

When the calculated pixel information and the predetermined reference pixel differ by a predetermined level or more, the image analysis device 101 may determine that an intruder has occurred, detect an intrusion, and track the intruder in the acquired images.

For example, if the second image is an image in which the doorway is photographed from the inside of an industrial facility, the preset reference pixel of the second image may be a pixel of the doorway. The entrance is set as the region of interest, and the image analysis device 101 can calculate pixel information of the region of interest from the acquired second image and compare it with a preset reference pixel. When the intruder enters through the entrance, the calculated pixel information of the region of interest may be different from the preset reference pixel. Therefore, the image analysis device 101 may detect an intrusion by comparing the pixel information calculated from the acquired second image with a preset reference pixel.

Additionally, when detecting an intrusion, the image analysis device 101 may track the intruder using all the acquired images.

For example, the image analysis device 101 may track the intruder in the second image, select the image captured by the camera installed in the direction in which the intruder moved, and continue to track the intruder in the selected image. Therefore, the image analysis device 101 can detect the intrusion and track the intruder, so that even if the monitoring agent is not checking the image, the intrusion situation can be dealt with.

The image analysis apparatus 101 may set a region of interest in a third image among the acquired images, and determine a pattern of the region of interest. The image analysis device 101 may compare the determined pattern of the region of interest with a predetermined abnormal pattern in the region of interest, and detect an abnormal condition of the mechanical equipment included in the region of interest.

For example, the third image may include an image of machinery such as a coal feeder, a conveyor belt, and a center pipe. The image analysis device 101 may set a region of interest on a cross section of a coal conveying part of a coal feeder, a conveying part of a conveyor belt, and a center pipe.

The preset abnormal pattern may include an abnormal termination pattern in which the operation of the mechanical equipment is abnormally terminated, and an abnormal operation pattern in which the mechanical equipment is abnormally operated. When the pattern of the region of interest corresponds to a predetermined abnormal pattern, the image analysis device 101 may determine that an abnormal condition has occurred.

The integrated management server 103 may receive and store information on the acquired image and the detected abnormal situation from the image analysis device 101, and provide the stored image and the information on the abnormal situation to the user.

The integrated management server 103 may output an alarm to the user when the video analysis device 101 detects an abnormal situation, and may store and store the detected abnormal image and the information on the received abnormal situation in pairs. .

For example, a user may access the integrated management server 103 through an application installed on the user's terminal or a web page, and obtain information about an image and an abnormal situation stored in the integrated management server 103.

That is, the smart image analysis system according to an embodiment of the present invention has the advantage of overcoming the limitations of monitoring by the monitoring personnel by performing fire detection, smoke detection, intrusion detection, and abnormal condition detection of mechanical equipment based on image analysis. There is this.

In addition, there is an advantage that it is possible to determine whether an abnormal situation occurs in real time, and output an alarm to the user as soon as the abnormal situation occurs, thereby minimizing damage caused by the abnormal situation.

2 is a view showing an embodiment of a smart image analysis system according to an embodiment of the present invention.

Specifically, FIG. 2A is a first image photographed for detecting fire and smoke, FIG. 2B is a second image photographed for detecting intrusion, and FIG. 2C is a third image photographed for detecting an abnormal condition of a machine. .

Referring to FIG. 2A, the first image is an image for detecting fire and smoke, and since the fire and smoke are not generated in a specific region, the region of interest may be set as the entire first image.

The image analysis device 101 may detect fire and smoke by calculating color feature values of the entire first image and comparing the calculated color feature values with a fire reference value and a smoke reference value.

Referring to FIG. 2B, the second image is an image for detecting intrusion, and in the second image, an area of interest may be set near the entrance of the facility.

The image analysis device 101 sets the region of interest in the vicinity of the doorway in the second image, calculates pixel information of the set region of interest, and compares the calculated pixel information with a preset reference pixel set in the region of interest. Can be detected.

Referring to FIG. 2C, the third image is an image for detecting an abnormal condition of the machine, and a region of interest may be set in the machine in the third image.

The image analysis device 101 may determine a pattern of the region of interest set in the third image, and compare the determined pattern with an abnormal pattern preset in the region of interest to detect an abnormal condition of the machine.

3 is a diagram illustrating an image analysis graph in a smart image analysis system according to an embodiment of the present invention.

Referring to FIG. 3, the image analysis apparatus 101 may set a region of interest 201 in an image, and adjust color, saturation, and brightness of the image of the region of interest 201 set. The adjusted image 203 shows the analyzed result by adjusting the color, saturation, and brightness.

The setting area 205 consists of items that can be set by the user, and can designate a region of interest (ROI), and set a background threshold, a detection threshold, and the like. That is, the user can set algorithms and parameters of image analysis in real time.

The image analysis graph 207 shows a result graph analyzed for each frame of the acquired image, and graphs processed in real time may be output. In addition, the image analysis graph 207 may display the maximum and minimum values of the threshold, so that the user can confirm the threshold by checking the image analysis graph 207.

Although the description herein appears in several exemplary aspects, various modifications or changes may be made from the scope defined by the following claims, and the technical protection scope of the present invention is described in the following claims. It will have to be decided by.

100: smart image analysis system
101: image analysis device
103: integrated management server
201: area of interest
203: adjusted image
205: setting area
207: image analysis graph

Claims (8)

An image analysis unit for setting a region of interest in each of the acquired images, and detecting an abnormal situation in the set region of interest; And
And an integrated management server that receives and stores the acquired image and detected abnormality information, and provides the stored image and abnormality information to a user.
The acquired image includes a first image, a second image, and a third image,
The image analysis unit,
The pattern of the region of interest set in the third image among the obtained images is determined, and the determined pattern is compared with the abnormal pattern preset in the region of interest, so that the determined pattern corresponds to the abnormal pattern preset in the region of interest If it is, it is determined that an abnormal condition of the mechanical equipment included in the region of interest of the third image has occurred,
The preset abnormal pattern includes an abnormal termination pattern in which the operation of the mechanical equipment is abnormally terminated, and an abnormal operation pattern in which the mechanical equipment is abnormally operated,
Smart image analysis system. According to claim 1,
The image analysis unit,
Calculating a color feature value of the first image among the acquired images, and detecting fire or smoke based on the calculated color feature value,
Smart image analysis system. According to claim 1,
The image analysis unit,
Calculating pixel information of a region of interest set in a second image among the obtained images, and detecting intrusion based on a result of comparing the calculated pixel information with a reference pixel preset in the region of interest,
Smart image analysis system. According to claim 3,
The image analysis unit,
If the calculated pixel information and the preset reference pixel differ by more than a certain level, it is determined that an intruder has occurred,
Tracking the intruder in the acquired images,
Smart image analysis system. delete delete According to claim 1,
The integrated management server,
Designating regions of interest for each of the acquired images, transmitting the designated regions of interest to the image analysis unit,
The image analysis unit,
Receiving the designated regions of interest, and setting regions of interest in images corresponding to each of the received regions of interest,
Smart image analysis system. According to claim 1,
The integrated management server,
When the above situation is detected, an alarm is output,
The image of the received abnormal situation is detected and information about the received abnormal situation is paired and stored.
Smart image analysis system.

Images