CN108520615B - An image-based fire identification system and method - Google Patents

Abstract

The invention provides an image-based fire identification system and method. A CCD camera is used to collect surrounding images in real time. A brightness identification unit performs brightness identification of the images to determine whether there is a highlight object. When a highlight object is detected, a speed measurement is used. The unit calculates the moving speed of the highlighted object, and comprehensively uses the weather data, the geographic coordinates of the location of the CCD camera, the rotation angle of the CCD camera, and the local time of the location of the CCD camera to determine whether the highlighted object is a flame.

Description

An image-based fire identification system and method

technical field

The invention belongs to the technical field of image processing, and in particular relates to an image-based fire identification system and method.

Background technique

Because the forest area is sparsely populated, it is difficult to be discovered in the early stage of the fire. Several major forest fires in the history of our country were only discovered after the fire spread to a certain scale. Due to the rugged mountain roads, lush vegetation, and inconvenient transportation in forest areas, it is difficult to transport fire-fighting equipment to the fire scene, which brings extreme difficulty to forest fire fighting.

At present, the main working principle of forest fire monitoring based on video technology is to analyze the color and brightness of the image after acquiring the video image, so as to determine whether there is a fire. However, due to the great uncertainty of the color and shape of the flame, the monitoring device cannot distinguish light sources such as the sun, the moon, lightning, and car lights from the flame in the early stage of the fire, and it is easy to falsely report the danger.

SUMMARY OF THE INVENTION

The present invention provides an image-based fire identification system and method, aiming at solving the above-mentioned technical problems.

An image-based fire identification system, comprising:

CCD camera, CCD camera rotates 360 degrees around the installation pole to collect forest images in real time.

The CCD camera rotation control unit is used to control the rotation speed of the CCD camera.

The image acquisition unit acquires the current image of the video surveillance area of the CCD camera.

The brightness identification unit identifies whether there is a bright object in the current image. When the highlighted object is identified, the CCD camera rotation control unit controls the CCD camera to stop rotating, and the CCD camera is aimed at the position of the highlighted object and continuously captures images.

The speed measurement unit is used to track the identified highlighted object and calculate the moving speed of the highlighted object. If the speed is greater than the threshold, it is determined that the highlighted object is not a flame, and the CCD camera resumes rotation.

The weather data reading unit is used to retrieve the real-time weather data of the location of the CCD camera in the weather system.

The rotation angle acquisition unit of the CCD camera is used to acquire the rotation angle of the video surveillance unit when a bright object is captured.

The central processing unit, if the weather data read by the weather data reading unit is cloudy or rainy, exclude the possibility that the highlighted object is the sun or the moon, and the central processing unit sends an alarm command to the alarm unit; if the real-time weather is On sunny or cloudy days, the CCD camera rotation angle acquisition unit acquires the CCD camera rotation angle when a bright object is captured, and the central processing unit calculates the sun and moon phases based on the geographic coordinates of the CCD camera, the CCD camera rotation angle, and the local time where the CCD camera is located. For the position of the CCD camera, match the position of the sun and the moon to the position of the CCD camera and the position of the highlighted object. If they do not coincide, the central processing unit sends an instruction to the alarm unit. Marked as celestial light source, the CCD camera resumes rotation.

The alarm unit is used to obtain the alarm instruction issued by the central processing unit and issue a fire alarm signal.

Further, the image-based fire identification system further includes an image preprocessing unit, which is used for preprocessing to exclude illumination from the image acquired by the image acquisition unit. For the image acquired by the image acquisition unit, first convert the color image into a grayscale image, and then use the gamma transformation method to remove excess light, wherein the threshold of the gamma transformation is dynamically determined by calculating the maximum value of the pixel grayscale in the image;

The basic form of gamma transform is:

s=cr ^γ

Convert the color image to a grayscale image, and then traverse the entire image to find the highest grayscale value g in all pixels, that is

g=max{g ₁ , g2, ..., g _i }

Then use the deformation formula of the gamma transform:

s=c(rg) ^γ

and

s=c(r+L-1-g) ^γ

where r is the grayscale value of the input image, s is the grayscale value of the output image, c and γ are normal numbers, L is the grayscale level of the grayscale image, and _gi represents the grayscale value of the ith pixel in the image .

An image-based fire identification method, comprising the following steps;

S1. Under the control of the CCD camera rotation control unit, the CCD camera rotates 360 degrees around the installation rod to collect forest images in real time.

S2. The image acquisition unit acquires the current image of the video surveillance area of the CCD camera.

S3. The brightness identification unit identifies whether there is a highlight object in the current image. When a highlight object is identified, the CCD camera rotation control unit controls the CCD camera to stop rotating, and the CCD camera is aimed at the position of the highlight object and continuously captures images.

S4. The speed measuring unit tracks the identified highlighted object, and calculates the moving speed of the highlighted object. If the speed is greater than the threshold, it is determined that the highlighted object is not a flame, and the CCD camera resumes rotation.

S5. If the moving speed of the highlighted object is less than the threshold value, retrieve the real-time weather data of the location of the CCD camera in the weather system.

S6, if the weather data read by the weather data reading unit is cloudy or rainy, then exclude the possibility that the highlighted object is the sun or the moon, the central processing unit sends an alarm instruction to the alarm unit, and the alarm unit obtains the central processing unit The unit sends an alarm command and sends out a fire alarm signal; if the real-time weather is sunny or cloudy, the CCD camera rotation angle acquisition unit obtains the rotation angle of the CCD camera when the highlighted object is captured, and the central processing unit rotates the CCD camera according to the geographic coordinates of the CCD camera. The angle and the local time where the CCD camera is located, calculate the position of the sun and the moon relative to the position of the CCD camera, and match the position of the sun and the moon relative to the position of the CCD camera and the position of the highlighted object. If they do not coincide, the central processing unit will alert the police. The unit sends an instruction, and the alarm unit obtains the alarm instruction sent by the central processing unit and sends out a fire alarm signal. If it coincides, the central processing unit marks the highlight as a celestial light source, and the CCD camera resumes rotation.

Further, between steps S2 and S3, an image preprocessing step is also included, which is used to perform preprocessing of eliminating illumination on the image acquired by the image acquisition unit. For the image acquired by the image acquisition unit, first convert the color image into a grayscale image, and then use the gamma transformation method to eliminate excess light, wherein the threshold value of the gamma transformation is dynamically determined by calculating the maximum value of pixel grayscale in the image.

The basic form of gamma transform is:

s=cr ^γ

Convert the color image to a grayscale image, and then traverse the entire image to find the highest grayscale value g in all pixels, that is

g=max{g ₁ , g ₂ , ..., gi}

Then use the deformation formula of the gamma transform:

s=c(rg) ^γ

and

s=c(r+L-1-g) ^γ

where r is the grayscale value of the input image, s is the grayscale value of the output image, c and γ are normal numbers, L is the grayscale level of the grayscale image, and _gi represents the grayscale value of the ith pixel in the image .

Description of drawings

Figure 1 is a working flow chart of an image-based fire identification system.

Detailed ways

The embodiments are described in detail below with reference to the accompanying drawings.

Fig. 1 shows the working flow chart of an image-based fire identification system of the present invention. An image-based fire identification system of the present invention includes: a CCD camera, which rotates 360 degrees around a mounting rod to perform a forest image Real-time collection.

The CCD camera rotation control unit is used to control the rotation speed of the CCD camera.

The image acquisition unit acquires the current image of the video surveillance area of the CCD camera.

The brightness identification unit identifies whether there is a bright object in the current image. When the highlighted object is identified, the CCD camera rotation control unit controls the CCD camera to stop rotating, and the CCD camera is aimed at the position of the highlighted object and continuously captures images.

The speed measurement unit is used to track the identified highlighted objects and calculate the moving speed of the highlighted objects. If the speed is greater than the threshold, it is determined that the highlighted object is not a flame; the highlighted object is a fast-moving light source such as lightning or car lights, and the CCD camera resumes rotation.

The weather data reading unit is used to retrieve the real-time weather data of the location of the CCD camera in the weather system through the network. If the moving speed of the highlighted object is less than the threshold, the real-time weather data of the location of the CCD camera in the weather system is retrieved.

The rotation angle acquisition unit of the CCD camera is used to acquire the rotation angle of the video surveillance unit when a bright object is captured.

The central processing unit, if the weather data read by the weather data reading unit is cloudy or rainy, exclude the possibility that the highlighted object is the sun or the moon, and the central processing unit sends an alarm command to the alarm unit; if the real-time weather is On sunny or cloudy days, the CCD camera rotation angle acquisition unit acquires the CCD camera rotation angle when a bright object is captured, and the central processing unit calculates the sun and moon phases based on the geographic coordinates of the CCD camera, the CCD camera rotation angle, and the local time where the CCD camera is located. For the position of the CCD camera, match the position of the sun and the moon to the position of the CCD camera and the position of the highlighted object. If they do not coincide, the central processing unit sends an instruction to the alarm unit. Marked as celestial light source, the CCD camera resumes rotation.

The alarm unit is used to obtain the alarm instruction issued by the central processing unit and issue a fire alarm signal.

Further, an image preprocessing unit is also included, configured to perform preprocessing of excluding illumination on the image acquired by the image acquisition unit. For the image acquired by the image acquisition unit, first convert the color image into a grayscale image, and then use the gamma transformation method to eliminate excess light, wherein the threshold value of the gamma transformation is dynamically determined by calculating the maximum value of pixel grayscale in the image.

In order to reduce the influence of light. A method of gamma transformation (also called power transformation) is used to process the acquired current image.

The basic form of gamma transform is:

s=cr ^γ

Convert the color image to a grayscale image, and then traverse the entire image to find the highest grayscale value g in all pixels, that is

g=max{g ₁ , g ₂ , ..., g _i }

Then use the deformation formula of the gamma transform:

s=c(rg) ^γ

and

s=c(r+L-1-g) ^γ

where r is the grayscale value of the input image, s is the grayscale value of the output image, c and γ are normal numbers, L is the grayscale level of the grayscale image, and _gi represents the grayscale value of the ith pixel in the image .

The present invention also provides an image-based fire identification method using the above-mentioned image-based fire identification system, which includes the following steps.

S1. Under the control of the CCD camera rotation control unit, the CCD camera rotates 360 degrees around the installation rod to collect forest images in real time.

S2. The image acquisition unit acquires the current image of the video surveillance area of the CCD camera.

S3. The brightness identification unit identifies whether there is a highlighted object in the current image. When a highlighted object is identified, the CCD camera rotation control unit controls the CCD camera to stop rotating, and the CCD camera is aimed at the position of the highlighted object and continuously captures images.

S4, the speed measuring unit tracks the identified highlighted object, and calculates the moving speed of the highlighted object. If the speed is greater than the threshold, it is determined that the highlighted object is not a flame; the highlighted object is a fast-moving light source such as lightning or car lights, and the CCD camera resumes rotation.

S5. The weather data reading unit retrieves the real-time weather data of the location of the CCD camera in the weather system through the network. If the moving speed of the highlighted object is less than the threshold, the real-time weather data of the location of the CCD camera in the weather system is retrieved.

S6. If the weather data read by the weather data reading unit is cloudy or rainy, exclude the possibility that the highlighted object is the sun or the moon, and the central processing unit sends an alarm instruction to the alarm unit; if the real-time weather is sunny or Cloudy, the CCD camera rotation angle acquisition unit acquires the CCD camera rotation angle when a bright object is captured, and the central processing unit calculates the sun and moon phases for the CCD based on the geographic coordinates of the CCD camera, the CCD camera rotation angle, and the local time where the CCD camera is located. The orientation of the camera, and the sun and moon phases are matched with the orientation of the CCD camera and the orientation of the highlighted object. If they do not coincide, the central processing unit sends an instruction to the alarm unit. If they overlap, the central processing unit marks the highlighted object as The celestial light source, the CCD camera resumes rotation.

S7, the alarm unit acquires the alarm instruction sent by the central processing unit and sends out a fire alarm signal.

Further, between steps S2 and S3, an image preprocessing step is also included, which is used to perform preprocessing of eliminating illumination on the image acquired by the image acquisition unit. For the image acquired by the image acquisition unit, first convert the color image into a grayscale image, and then use the gamma transformation method to eliminate excess light, wherein the threshold value of the gamma transformation is dynamically determined by calculating the maximum value of pixel grayscale in the image.

In order to reduce the influence of illumination, the method of gamma transformation (also called power transformation) is used to process the acquired current image. The basic form of gamma transformation is:

s=cr ^γ

Convert the color image to a grayscale image, and then traverse the entire image to find the highest grayscale value g in all pixels, that is

g=max{g ₁ , g ₂ , ..., g _i }

Then use the deformation formula of the gamma transform:

s=c(rg) ^γ

and

s=c(r+L-1-g) ^γ

where r is the grayscale value of the input image, s is the grayscale value of the output image, c and γ are normal numbers, L is the grayscale level of the grayscale image, and _gi represents the grayscale value of the ith pixel in the image .

The above embodiments are only preferred specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. , all should be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (4)

1. an image-based fire identification system, is characterized in that, comprises: CCD camera, the CCD camera rotates 360 degrees around the installation pole, and collects the surrounding forest images in real time; The CCD camera rotation control unit is used to control the rotation speed of the CCD camera; The image acquisition unit acquires the current image of the video surveillance area of the CCD camera; The brightness recognition unit identifies whether there is a highlight object in the current image acquired by the image acquisition unit. When a highlight object is identified, the brightness recognition unit sends a signal to the CCD camera rotation control unit. The CCD camera rotation control unit controls the CCD camera to stop rotating, and the CCD camera stops rotating. The camera is aimed at the location of the highlighted object and continuously captures images; The speed measuring unit is used to track the identified highlighted object and calculate the moving speed of the highlighted object. If the speed is greater than the threshold, it is determined that the highlighted object is not a flame, and the speed measuring unit sends a signal to the CCD camera rotation control unit, and the CCD camera rotates The control unit controls the CCD camera to resume rotation; The weather data reading unit is used to retrieve the real-time weather data of the location of the CCD camera in the weather system through the network; The CCD camera rotation angle acquisition unit is used to acquire the rotation angle data of the video surveillance unit when the highlighted object is captured; The central processing unit, if the weather data read by the weather data reading unit is cloudy or rainy, exclude the possibility that the highlighted object is the sun or the moon, and the central processing unit sends an alarm command to the alarm unit; if the real-time weather is On sunny or cloudy days, the CCD camera rotation angle acquisition unit acquires the CCD camera rotation angle when a bright object is captured, and the central processing unit calculates the sun and moon phases based on the geographic coordinates of the CCD camera, the CCD camera rotation angle, and the local time where the CCD camera is located. For the position of the CCD camera, match the position of the sun and the moon to the position of the CCD camera and the position of the highlighted object. If they do not coincide, the central processing unit sends an instruction to the alarm unit. If they coincide, the central processing unit will put the highlighted object. Marked as celestial light source, the CCD camera resumes rotation; The alarm unit is used to obtain the alarm instruction issued by the central processing unit and issue a fire alarm signal. 2 . The image-based fire identification system according to claim 1 , further comprising an image preprocessing unit, which is used to perform preprocessing of eliminating illumination on the image acquired by the image acquisition unit, firstly, the image acquisition unit The acquired color image is converted into a grayscale image, and then the gamma transformation method is used to remove excess light, wherein the threshold value of the gamma transformation is dynamically determined by calculating the maximum value of the pixel grayscale in the image; The basic form of gamma transform is: s=cr ^γ Convert the color image to a grayscale image, and then traverse the entire image to find the highest grayscale value g in all pixels, that is g=max{g ₁ , g ₂ , ..., g _i } Then use the deformation formula of the gamma transform: s=c(rg) ^γ and s=c(r+L-1-g) ^γ where r is the grayscale value of the input image, s is the grayscale value of the output image, c and γ are normal numbers, L is the grayscale level of the grayscale image, and _gi represents the grayscale value of the ith pixel in the image . 3. An image-based fire identification method, characterized in that, comprising the following steps; S1. Under the control of the CCD camera rotation control unit, the CCD camera rotates 360 degrees around the installation rod to collect forest images in real time; S2, the image acquisition unit acquires the current image of the video surveillance area of the CCD camera; S3, the brightness identification unit identifies whether there is a highlighted object in the current image, and when the highlighted object is identified, the CCD camera rotation control unit controls the CCD camera to stop rotating, and the CCD camera is aimed at the position of the highlighted object and continuously captures images; S4. The speed measuring unit tracks the identified highlighted object, and calculates the moving speed of the highlighted object. If the speed is greater than the threshold, it is determined that the highlighted object is not a flame, and the CCD camera resumes rotation; S5. If the moving speed of the highlighted object is less than the threshold, the real-time weather data of the location of the CCD camera in the meteorological system is retrieved; S6, if the weather data read by the weather data reading unit is cloudy or rainy, then exclude the possibility that the highlighted object is the sun or the moon, the central processing unit sends an alarm instruction to the alarm unit, and the alarm unit acquires the central processing unit The unit sends an alarm command and sends out a fire alarm signal; if the real-time weather is sunny or cloudy, the CCD camera rotation angle acquisition unit obtains the rotation angle of the CCD camera when the highlighted object is captured, and the central processing unit rotates the CCD camera according to the geographic coordinates of the CCD camera. The angle and the local time where the CCD camera is located, calculate the position of the sun and the moon relative to the position of the CCD camera, and match the position of the sun and the moon relative to the position of the CCD camera and the position of the highlighted object. If they do not coincide, the central processing unit will alert the police. The unit sends an instruction, and the alarm unit acquires the alarm instruction sent by the central processing unit and sends out a fire alarm signal. If it coincides, the central processing unit marks the highlighted object as a celestial light source, and the CCD camera resumes rotation. 4. An image-based fire identification method according to claim 3, characterized in that, between steps S2 and S3, an image preprocessing step is further included for removing the illumination from the image acquired by the image acquisition unit. Preprocessing, first convert the image acquired by the image acquisition unit into a grayscale image, and then use the gamma transformation method to eliminate excess light, wherein the threshold of the gamma transformation is dynamically determined by calculating the maximum value of the pixel grayscale in the image; The basic form of gamma transform is: s=cr ^γ Convert the color image to a grayscale image, and then traverse the entire image to find the highest grayscale value g in all pixels, that is g=max{g ₁ , g ₂ , ..., g _i } Then use the deformation formula of the gamma transform: s=c(rg) ^γ and s=c(r+L-1-g) ^γ where r is the grayscale value of the input image, s is the grayscale value of the output image, c and γ are normal numbers, L is the grayscale level of the grayscale image, and _gi represents the grayscale value of the ith pixel in the image .

Images