KR20210156980A - System and method for image processing based on smoke area localization, and a recording medium having computer readable program for executing the method - Google Patents

Abstract

Disclosed are a smoke area zoning-based image processing system, a method, and a recording medium having a computer readable program for executing the method thereof for acquiring high-quality image information of a fire disaster site while using a general visible light color camera. The image processing system comprises an image acquisition unit, a smoke detection unit, a region segmentation unit, a smoke concentration calculation unit, and a smoke removal unit. The image acquisition unit acquires an image of a fire scene, and the smoke detection unit detects smoke from the image. The area zoning unit partitions a smoke area in which smoke is detected. The smoke concentration calculation unit calculates the concentration of smoke corresponding to the smoke area. The smoke removal unit calculates an image from which smoke in the smoke area is removed according to the smoke concentration.

Description

A smoke area zoning-based image processing system, method, and a recording medium recording a computer readable program for executing the method. {SYSTEM AND METHOD FOR IMAGE PROCESSING BASED ON SMOKE AREA LOCALIZATION, AND A RECORDING MEDIUM HAVING COMPUTER READABLE PROGRAM FOR EXECUTING THE METHOD}

The present invention relates to image processing-related technology, and more particularly, to a system and method for effectively acquiring a high-quality image in an area where visibility is unclear due to smoke.

Smoke, fog, and smog generated at the site of a fire disaster interfere with the observation of video information essential for firefighting, rescue, and first aid (fire point, person in need, dangerous object, and landmark). Accordingly, in recent years, as the spread of thermal imaging cameras has been activated, the use of thermal imaging cameras for searching for fire spots or people who are hidden by smoke at a fire disaster site is increasing.

However, thermal imaging cameras are still expensive and high-resolution cameras are difficult to actively use because there is a risk of cost loss in case of damage. In addition, the thermal image has a lower resolution than the visible light band image sensor, so it is not easy to search for a fire point at a long distance, and there is a limit when trying to identify the color of the person's clothing in the smoke.

KR 102087000 B1

The present invention has been devised to solve the above-described problems in the prior art, and provides a system and method for obtaining excellent quality image information of a fire disaster site while using a general visible light color camera without additional expensive equipment. aim to

In order to achieve the above object, an image processing system based on smoke region zoning according to the present invention includes an image acquisition unit, a smoke detection unit, a region zoning unit, a smoke concentration calculation unit, and a smoke removal unit. The image acquisition unit acquires an image of the fire scene, the smoke detection unit detects smoke from the acquired image, the area zoning unit partitions a smoke area in which smoke is detected, and the smoke concentration calculation unit calculates a concentration of smoke corresponding to the smoke area and the smoke removal unit calculates an image from which smoke in the smoke area is removed according to the smoke concentration.

According to this configuration, since the effect of smoke on the image is automatically removed just by shooting a fire image, it is possible to use a general visible light color camera without additional expensive equipment such as a thermal imaging camera, while providing excellent quality image information of the fire disaster site. can be obtained

In addition, since the smoke perspective image is automatically provided without user's separate manipulation, when applied to a helmet or the like, it is possible to provide great help to firefighters having a low degree of freedom.

In this case, the image processing system may further include a fire detection unit that detects the occurrence of a fire, and the smoke removal unit may calculate an image from which smoke is removed by using the fire information detected by the fire detection unit. According to such a configuration, it is possible to perform a smoke removal operation most suitable for a fire type without a separate operation by a user by using various types of fire information detected using various types of sensors.

In addition, the apparatus may further include a photographing controller configured to calculate a photographing parameter of the image acquisition unit by using the smoke density information. According to this configuration, it is possible to secure a more effective smoke perspective image by automatically changing the setting of the camera as well as image processing.

The apparatus may further include an object-of-interest detector configured to detect an object of interest in the smoke region, and the imaging controller may calculate a photographing variable by further using information on the object of interest. According to such a configuration, it is possible to automatically set a shooting condition for best observation of an object of interest according to information on the object of interest.

In addition, it further includes an image transmitter for transmitting the image from which the smoke has been removed to the outside, and a compression information acquirer for acquiring image compression information from the image transmitter, wherein the smoke remover may further use the compression information to remove smoke. According to this configuration, by reflecting the compressed information of the image transmitted to the outside in the image processing, it is possible to secure a more effective smoke perspective image.

In addition, the area partitioning unit may partition a plurality of smoke areas according to the concentration of smoke, and the smoke removal unit may calculate an image from which smoke has been removed for regions having similar smoke concentrations to each other, and synthesize the plurality of calculated images. . According to such a configuration, since it is possible to provide an image in which smoke has been removed for each of a plurality of smoke areas existing in one frame, visibility of the entire image can be more effectively improved.

In addition, the invention implementing the system in the form of a method and a recording medium recording a computer readable program for executing the method are disclosed together.

According to the present invention, since the effect of smoke on the image is automatically removed simply by shooting a fire image, image information of a fire disaster site of excellent quality is obtained using a general visible light color camera without separate expensive equipment such as a thermal imaging camera. be able to do

In addition, since the smoke perspective image is automatically provided without user's separate manipulation, when applied to a helmet or the like, it is possible to provide great help to firefighters having a low degree of freedom.

In addition, by using various types of fire information detected using various types of sensors, it is possible to perform a smoke removal operation most suitable for the type of fire without a separate operation by the user.

In addition, it is possible to secure a more effective smoke perspective image by automatically changing the camera settings as well as image processing.

In addition, it is possible to automatically set the shooting conditions for best observing the object of interest according to the information of the object of interest.

In addition, by reflecting the compressed information of the image transmitted to the outside in the image processing, it is possible to secure a more effective smoke perspective image.

In addition, since it is possible to provide an image in which smoke has been removed for each of a plurality of smoke areas existing in one frame, visibility of the entire image may be more effectively improved.

1 is a schematic block diagram of an image processing system based on smoke area zoning according to an embodiment of the present invention;
FIG. 2 is a block diagram illustrating an actual implementation example of the image processing system of FIG. 1;
FIG. 3 is a flowchart of an image processing method performed by the image processing system of FIG. 1 according to a first embodiment;
4 is a diagram illustrating a process of calculating a plurality of smoke removal images and synthesizing them by the process of FIG. 3 .
5 is a flowchart of a second embodiment of an image processing method performed by the image processing system of FIG. 1;
6 is a flowchart of a third embodiment of an image processing method performed by the image processing system of FIG. 1;
7 is a flowchart of a fourth embodiment of an image processing method performed by the image processing system of FIG. 1;

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings.

1 is a schematic block diagram of an image processing system based on smoke area zoning according to an embodiment of the present invention, and FIG. 2 is a block diagram illustrating an actual implementation of the image processing system of FIG. 1 .

In FIG. 1 , the image processing system 100 includes an image acquisition unit 110 , a smoke detection unit 120 , a region zoning unit 130 , a smoke concentration calculation unit 140 , a smoke removal unit 150 , and a fire detection unit 160 . ), an imaging control unit 170 , an object of interest detection unit 180 , an image transmission unit 190 , and a compressed information acquisition unit 195 .

The image acquisition unit 110 acquires an image of a fire scene, and the smoke detection unit 120 detects smoke from the image. At this time, the smoke may include any other material that floats in the air, such as dust, fog, smog, etc. generated at the site of a fire disaster and blocks the view.

The area dividing unit 130 partitions a smoke area in which smoke is detected, and the smoke concentration calculating unit 140 calculates a concentration of smoke corresponding to the smoke area. More specifically, whether a pixel located at an arbitrary position (x,y) is included in the smoke area is determined based on the minimum value DCV(x,y) among RGB values of pixels located in a specific radius around the pixel do. In this case, DCV means a dark channel value.

In a normal image without smoke, the DCV(x,y) value is close to 0. On the other hand, the higher the smoke concentration, the larger the DCV value. If an area is divided based on the DCV value (eg, a segmentation technique such as clustering is applied), the smoke area can be zoned and the average smoke concentration for each area can be calculated.

The smoke removal unit 150 calculates an image from which smoke in the smoke area is removed according to the smoke concentration. According to such a configuration, when it is difficult to use a thermal imaging camera or when high-resolution observation is required, when color identification of a search target is required, a visible light color camera that is widely available and can be purchased at a relatively low price can be used in smoke. object identification can be improved. In particular, it is possible to improve the performance of improving the image in the smoke by zoning the smoke area in the image and calculating the smoke concentration and the like.

The area partitioning unit 130 may partition a plurality of smoke areas according to the concentration of smoke. In this case, the smoke removal unit 150 may calculate an image from which smoke has been removed for each area having a similar smoke concentration, and synthesize the plurality of calculated images.

That is, the smoke area displayed in one frame may be divided into a plurality of areas according to the concentration. In this case, overall visibility may be improved by applying image processing optimized according to the smoke concentration.

3 is a flowchart of a first embodiment of an image processing method performed by the image processing system of FIG. 1 , and FIG. 4 is a diagram illustrating a process of calculating and synthesizing a plurality of smoke removal images by the process of FIG. 3 .

The image transmission unit 190 transmits the smoke-free image to the outside, and the compression information acquisition unit 195 acquires image compression information from the image transmission unit 190 . In this case, the smoke removal unit 150 may remove the smoke by further using the compressed information. According to this configuration, by reflecting the compressed information of the image transmitted to the outside in the image removal, it is possible to secure a more effective smoke perspective image.

Areas obscured by smoke usually have very low contrast. Therefore, when the transmission channel speed is low, it is necessary to increase the compression rate for image transmission. In this case, there is a risk that image information in a portion obscured by smoke may be lost in the compression process.

Therefore, in order to make it easier to identify objects obscured by smoke in conjunction with the compression ratio, the lower the target bit per pixel (BPP) (the same meaning as the higher compression ratio), the greater the smoke removal treatment intensity.

FIG. 5 is a flowchart of a second embodiment of an image processing method performed by the image processing system of FIG. 1 . 5 shows a process of performing the smoke removal processing intensity in association with the image compression rate.

The fire detection unit 160 detects the occurrence of a fire, and the smoke removal unit 150 may calculate an image from which smoke is removed by using the fire information detected by the fire detection unit 160 . The fire detection unit 160 may be implemented using an NIR sensor or a thermal image sensor, and a non-image sensor such as an illuminance sensor may be used in combination in addition to the image sensor.

According to such a configuration, it is possible to perform a smoke removal operation most suitable for a fire type without a separate operation by a user by using various types of fire information detected using various types of sensors.

The photographing control unit 170 calculates a photographing variable of the image obtaining unit 110 by using the smoke density information. This is to determine the optimal camera shooting parameters to identify objects covered by smoke. When it is determined that there is an object of interest in the smoke area, it is to set the optimal shooting conditions to best observe the object of interest according to the information of the object of interest. Examples of the shooting variable include exposure time, focus, gain, and dynamic range of the camera.

According to such a configuration, it is possible to secure a more effective smoke perspective image by automatically changing not only image processing but also the setting of the camera itself. 6 is a flowchart of a third embodiment of an image processing method performed by the image processing system of FIG. 1 . In FIG. 6 , a process of determining a camera shooting parameter according to a smoke concentration is illustrated.

The object-of-interest detector 180 detects the object of interest in the smoke area. In this case, the photographing control unit 170 may calculate a photographing variable by further using information on the object of interest. For example, the object of interest may be preset as an object of a specific color, such as red. According to such a configuration, it is possible to automatically set a photographing condition for best observing the object of interest according to information on the object of interest.

7 is a flowchart of a fourth embodiment of an image processing method performed by the image processing system of FIG. 1 . When it is determined that there is an object of interest in the smoke area, a process of setting an optimal imaging condition for best observation of the object of interest according to information on the object of interest is illustrated.

According to the present invention, when it is difficult to use a thermal imaging camera or high-resolution observation is required, or when it is necessary to identify a color of a search target, it is possible to improve object identification in smoke by using a universally used visible light color camera. be able to In addition, the smoke region in the image is zoned, the smoke concentration, etc. are calculated, and the performance of improving the image in the smoke can be improved by using this.

Although the present invention has been described with reference to some preferred embodiments, the scope of the present invention should not be limited thereto, but should also extend to modifications or improvements of the above embodiments supported by the claims.

100: smoke area zoning based image processing system
110: image acquisition unit
120: smoke detection unit
130: area zoning unit
140: smoke concentration calculator
150: smoke removal unit
160: fire detection unit
170: shooting control unit
180: object of interest detection unit
190: video transmitter
195: compressed information acquisition unit

Claims (8)

an image acquisition unit for acquiring an image;
a smoke detector for detecting smoke from the image;
an area zoning unit partitioning a smoke area in which the smoke is detected;
a smoke concentration calculator configured to calculate a concentration of smoke corresponding to the smoke region; and
and a smoke removal unit configured to calculate an image from which the smoke of the smoke area is removed according to the smoke concentration.
The method according to claim 1,
Further comprising a fire detection unit for detecting the occurrence of a fire,
The smoke removal unit calculates an image from which the smoke is removed by using the fire information detected by the fire detection unit.
The method according to claim 1,
The image processing system according to claim 1, further comprising a photographing controller configured to calculate a photographing parameter of the image acquisition unit by using the information on the concentration of the smoke.
4. The method according to claim 3,
Further comprising an object-of-interest detection unit for detecting an object of interest in the smoke area,
The image processing system, characterized in that the photographing control unit calculates the photographing variable by further using the information on the object of interest.
The method according to claim 1,
an image transmitter for transmitting the smoke-free image to the outside; and
Further comprising a compression information acquisition unit for acquiring the image compression information from the image transmission unit,
The image processing system according to claim 1, wherein the smoke removal unit removes the smoke by further using the compression information.
The method according to claim 1,
The area partitioning unit partitions a plurality of smoke areas according to the concentration of the smoke,
The smoke removal unit calculates an image from which smoke has been removed for each area having a similar smoke concentration, and synthesizes the plurality of calculated images.
An image processing method performed by an image processing system, comprising:
an image acquisition step of acquiring an image;
a smoke detection step of detecting smoke from the image;
an area zoning step of dividing a smoke area in which the smoke is detected;
a smoke concentration calculating step of calculating a concentration of smoke corresponding to the smoke area; and
and a smoke removal step of calculating an image from which the smoke of the smoke area is removed according to the smoke concentration.
A recording medium in which a computer readable program for performing the method of claim 7 is recorded.

Images