A system and method for early, ultra-weak flame imaging localization

Abstract

The application provides a detection system and method for early extremely weak flame imaging positioning, which comprises a horizontal light path branch, a vertical light path branch and a video circuit processing board; the horizontal light path branch is coaxially connected with a first focusing lens, a light splitting prism, an ultraviolet filter, a second focusing lens, an image intensifier and a first image detector in sequence along a horizontal direction; the vertical light path branch is connected with a third focusing lens and a second image detector in sequence along a direction perpendicular to the horizontal light path and a light splitting surface of the light splitting prism; the video circuit processing board is provided with a video circuit processing board, which is electrically connected with the first image detector and the second image detector respectively to receive image electric signals output by the two detectors. The system can image and detect early extremely weak flames at a very long distance, can amplify the spectrum of the flame alone at a high magnification, and cannot amplify the spectrum of other background objects, so that the characteristics of the flame are highlighted.

Description

Technical Field

[0001] This application relates to the field of fire monitoring and alarm technology, and in particular to a detection system and method for early-stage, extremely weak flame imaging and localization. Background Technology

[0002] Fire is one of the most common natural disasters, causing severe casualties and economic losses. For critical locations such as chemical plants, oil refineries, oil storage tanks, and hazardous chemical warehouses, the consequences of a fire are extremely serious. Therefore, in these highly flame-sensitive locations, the detection and location of early, weak flames are crucial. Timely warnings and determination of the ignition point are essential for controlling the occurrence and spread of fire.

[0003] Currently, flame monitoring is mainly divided into two categories: point flame detectors and image-based flame monitors. Point flame detectors determine the presence of a flame by detecting the intensity of light illuminating the flame in different wavelengths. Their detection range is relatively small, typically effective for flames measuring 1 cm within 1 meter. They are susceptible to external interference, have slow detection speeds, and cannot provide imaging or location information, nor can they provide information such as flame size or degree of combustion. They are only suitable for confined spaces and cannot cover large three-dimensional spaces. Image-based flame monitors, on the other hand, can detect and provide information such as the shape, location, and intensity of the flame. They are highly suitable for fire monitoring in large three-dimensional spaces such as chemical plants, oil refineries, oil storage tanks, and hazardous chemical warehouses, and have become a research hotspot in fire monitoring in recent years.

[0004] Currently, image-based flame detection technologies mainly include visible light imaging and infrared imaging. This technology images all objects within the field of view in the visible or infrared bands, and then uses specific algorithms to extract features such as flame color, brightness, texture, shape contour, area changes, number of sharp corners, and flickering, thereby identifying the flame's location and size. However, in these two bands, the flame spectrum and the background object spectrum often coexist and overlap, easily causing interference. Therefore, it is more suitable for imaging and detecting large-scale flames. Early, weak flames often produce visible light or infrared radiation far weaker than the background environment, making them almost undetectable. For example, for flames smaller than 1 cm, visible light imaging can usually accurately identify them within 10 m, and infrared thermal imaging can usually accurately identify them within 8 m. At greater distances, they are easily affected by background interference.

[0005] In the visible and infrared bands, the flame spectrum and the spectrum of other background objects overlap and coexist. Although certain features of the flame can be extracted through certain algorithms, it is easy to make misjudgments. It is only suitable for large-scale flames with obvious dynamic features, and it is difficult to extract them for early and extremely weak flames. Summary of the Invention

[0006] This application provides a detection system and method for imaging and locating early-stage, extremely weak flames. To solve the above-mentioned technical problems, this application adopts the following technical methods: In a first aspect, this application provides a detection system for imaging and locating early, extremely weak flames, the detection system comprising a horizontal optical path branch, a vertical optical path branch, and a video circuit processing board; The horizontal optical path branch is a first focusing lens, a beam splitter, an ultraviolet filter, a second focusing lens, an image intensifier, and a first image detector that are coaxially connected in the horizontal direction. The vertical optical path branch is located in a direction perpendicular to the horizontal optical path, and is sequentially connected to the beam-splitting surface of the beam-splitting prism by the third focusing lens and the second image detector. The video circuit processing board is equipped with a video circuit processing board, which is electrically connected to the first image detector and the second image detector respectively. After the image uploaded by the first image detector is shaped according to a preset ratio, it is fused with the image uploaded by the second image detector to generate a video image containing the background object and the magnified flame.

[0007] Optionally, the beam splitter divides the beam passing through the first focusing lens into a strong beam and a weak beam at an energy ratio of 8:2; The high-intensity light beam is input into the ultraviolet filter; The weak beam is input into the third focusing lens.

[0008] Optionally, the magnification of the image intensifier is 20,000 times.

[0009] Optionally, the image intensifier includes a photocathode, a microchannel plate, and a fluorescent screen connected in sequence.

[0010] Secondly, this application provides a detection method for early-stage, extremely weak flame imaging and localization, including: Acquire the light waves generated by the flame and other background objects; Based on the light waves, strong and weak beams are determined; Based on the strong beam, an ultraviolet band strong light enhancement detection image is generated; Based on the weak light beam, a weak light reflection image is generated; The ultraviolet band strong light enhancement detection image is shaped according to a preset ratio and then fused with the weak light reflection image to generate a video image that includes the background object and the magnified flame.

[0011] Optionally, determining the strong and weak light beams based on the light wave includes: After the light waves are focused and split sequentially, strong beams and weak beams are generated.

[0012] Optionally, generating an ultraviolet band intensity-enhanced detection image based on the intense light beam includes: The strong beam is sequentially filtered, focused, and magnified to generate an ultraviolet band strong light enhanced detection image.

[0013] Thirdly, this application also provides a computer system, comprising: Memory is used to store instructions that can be executed by the processor; A processor for executing the instructions to implement the method as described in any of the second aspects.

[0014] Fourthly, this application also provides a computer-readable medium storing computer program code that, when executed by a processor, implements the method as described in any of the second aspects.

[0015] This application has the following beneficial effects: The system proposed in this application has an extremely long imaging detection range for early-stage, very weak flames. It can magnify the flame spectrum at high magnification without amplifying the spectra of other background objects, thereby highlighting the flame characteristics. Attached Figure Description

[0016] Figure 1 Solar spectrum provided for embodiments of this application; Figure 2 A schematic diagram of the structure of a detection system for early-stage, extremely weak flame imaging and localization provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of an image intensifier provided in an embodiment of this application; Figure 4 Different flame images provided for embodiments of this application; Figure 4 (a) is a full-spectrum flame image; Figure 4 (b) is the ultraviolet band high-intensity light enhanced detection image output by the first image detector; Figure 4 (c) is an image of the shaped flame; Figure 5 Images of different processing procedures provided in the embodiments of this application; Figure 5 (a) is a full-spectrum image of the background object and the flame; Figure 5 (b) is a weak light reflection image output by the second image detector; Figure 5 (c) is the fused image; Figure 6 An imaging effect diagram provided for an embodiment of this application when the flame is 1 cm in size at a distance of 10 meters; Figure 6 (a) is a visible light imaging image; Figure 6 (b) is an infrared thermal imaging image; Figure 6 (c) is the final fused image; Figure 7 An imaging effect diagram provided for an embodiment of this application when the flame is 1 cm in size at a distance of 20 meters; Figure 7 (a) is a visible light imaging image; Figure 7 (b) is an infrared thermal imaging image; Figure 7 (c) is the final fused image; Figure 8 An imaging effect diagram provided for an embodiment of this application when the flame is 1 cm in size at a distance of 30 meters; Figure 8 (a) is a visible light imaging image; Figure 8 (b) is an infrared thermal imaging image; Figure 8 (c) is the final fused image. Detailed Implementation

[0017] To facilitate understanding by those skilled in the art, the present application will be further described below in conjunction with embodiments and accompanying drawings. The content mentioned in the embodiments is not intended to limit the present application.

[0018] The principles of this application are based on the following points: (1) Solar spectrum diagram as shown Figure 1 As shown, the main energy is concentrated in visible light, with a significant amount of energy also present in the infrared region. Because the atmosphere strongly absorbs ultraviolet radiation, very little ultraviolet radiation reaches the ground. While some UVA and UVB rays can penetrate the atmosphere to reach the ground, UVC rays are almost completely absorbed and rarely reach the ground. In addition, there is usually almost no UVC band ultraviolet light at night or indoors; (2) The spectral distribution of flames of different substances is different, but the energy distribution is basically similar, mainly distributed in the visible light range. When the temperature rises, the infrared range also increases significantly. At the same time, although the distribution in the ultraviolet band is weak, it can basically extend to the UVC band; (3) This application proposes to use the above spectral distribution characteristics to magnify and image the ultraviolet light in the UVC band, that is, to magnify and image the UVC light wave of the flame separately, while other background objects do not reflect light in this band. Therefore, the UVC light waves of other background objects will not interfere with the imaging effect of the flame; (4) Since the energy of the flame in the UVC band is also very weak, it basically loses the original shape of the flame after being magnified and imaged. Therefore, it is necessary to reshape the magnified flame image to restore the approximate shape of the flame so that maintenance personnel can identify it; (5) Since the UVC band only has the image of the flame and no other background object images.

[0019] To solve the above technical problems, based on the above principles, such as Figure 2As shown, this application proposes a detection system for early-stage, extremely weak flame imaging and localization, including a horizontal optical path branch, a vertical optical path branch, and a video circuit processing board: The horizontal optical path branch consists of a first focusing lens 2, a beam splitter 3, an ultraviolet filter 4, a second focusing lens 5, an image intensifier 6, and a first image detector 7, which are coaxially connected in the horizontal direction. The vertical optical path branch is located in a direction perpendicular to the horizontal optical path, and is sequentially connected to the beam-splitting surface of the beam-splitting prism by the third focusing lens 8 and the second image detector 9. The signal processing module is equipped with a video circuit processing board 10, which is electrically connected to the first image detector 7 and the second image detector 9 respectively. The image uploaded by the first image detector 7 is shaped according to a preset ratio and then fused with the image uploaded by the second image detector 9 to generate a video image containing the background object and the magnified flame.

[0020] The detection system proposed in this application will be described in detail below, taking into account the above-mentioned apparatus: Within the field of view 1, the light waves generated by the flame and other background objects are focused by the first focusing lens 2 and transmitted to the beam splitter prism 3. The beam splitter prism 3 divides the beam into a strong beam and a weak beam at an energy ratio of 8:2. The strong beam enters the ultraviolet filter 4, which only allows UVC wavelengths to pass through. This wavelength is focused by the second focusing lens 5 and then transmitted to the image intensifier 6. The image intensifier 6 can magnify the UVC image by approximately 20,000 times. The magnified image, the UVC strong light enhancement detection image, is transmitted to the first image detector (CCD) 7 for image acquisition. Because the image is magnified approximately 20,000 times, the overall outline of the flame is formed by the snowflakes generated by electron bombardment of the fluorescent screen. This magnified image is transmitted to the video processing circuit board 10 in the signal processing module, which reshapes the outline formed by the snowflakes according to a specific ratio, preparing for subsequent image fusion. The weak beam is visible light and mainly carries image information of other background objects. After being focused and imaged by the third focusing lens 8, the beam is transmitted to the second image detector (CCD) 9 for acquisition, resulting in a weak light reflection image. The video processing circuit board 10 fuses the image acquired by the second image detector (CCD) 9 with the shaped image output by the first image detector (CCD) 7, and finally outputs a video image containing the background object and the magnified flame.

[0021] The internal working principle of Intensifier 6 is as follows Figure 3As shown, when UVC photons reach the photocathode 15, they generate electrons through the photoelectric effect. The electrons are accelerated under the action of a 200 V high-voltage electric field until they reach the microchannel plate 16. The electrons collide with the inner wall of the micropores of the microchannel plate 16, resulting in electron multiplication. The multiplied electrons are accelerated under the action of a 7000 V high-voltage electric field and collide with the fluorescent screen 17. A highly magnified UVC band image is formed on the back of the fluorescent screen 17. The image is characterized by the fact that the overall outline of the flame is composed of snowflake-like dots generated by electrons bombarding the fluorescent screen. The image is then transmitted to the first image detector (CCD) 7 for further processing.

[0022] The image processing flow of the video circuit processing board is as follows: First, the ultraviolet band high-intensity light enhanced detection image output by the first image detector (CCD) 7 is subjected to shaping processing, which involves sequentially acquiring the coordinates of the flame center point, acquiring the flame area, and replacing the flame pattern. The process is as follows: Figure 4 (a) is a full-spectrum flame image. Figure 4 (b) The image output by the first image detector (CCD) 7 after the ultraviolet band of the flame has been magnified 20,000 times by the image intensifier 6. This image is obtained using the following formula (1). Figure 4 (b) Coordinates of the center point of the snowflake dot profile: (1) x and y are the leftmost x-coordinate and the bottommost y-coordinate of the snowflake dot outline, respectively; W and H are the maximum width and maximum height of the snowflake dot outline, respectively. The coordinates of the center point of the snowflake dot outline are the same as the coordinates of the center point of the flame.

[0023] Get Figure 4 (b) After determining the area of ​​the snowflake dot outline, i.e., the maximum width and maximum height values ​​W and H of the snowflake dot outline; at the coordinates of the flame center point, cover the snowflake dot outline area with a flame icon of the same area as the flame, and at the same time, transform all the black areas outside the snowflake dots into a white transparent state, i.e., the shaped image output by the first image detector 7, such as Figure 4 As shown in (c).

[0024] The low-light reflection image acquired by the second image detector (CCD) 9 is fused with the shaped image output by the first image detector (CCD) 7. The specific process is as follows: Figure 5 (a) is a full-spectrum image of the background object and the flame. Figure 5 (b) is a weak light reflection image output by the second image detector (CCD) 9. Because the flame intensity itself is very weak and the light is dispersed, therefore... Figure 5 (b) The flame is barely visible. Figure 5 (b) and Figure 4 (c) Perform image fusion, because Figure 4 (c) All white areas except for the flames are transparent; therefore, the merged image is as follows: Figure 5 As shown in (c).

[0025] Based on the above detection system, this application also proposes a detection method for early-stage, extremely weak flame imaging and localization, including: The system acquires the light waves generated by the flame and other background objects, and then determines the strong and weak light beams based on these light waves. Based on the strong light beams, it generates an ultraviolet band strong light enhancement detection image; based on the weak light beams, it generates a weak light reflection image; and based on the ultraviolet band strong light enhancement detection image and the weak light reflection image, it generates a video image containing the background objects and the magnified flame.

[0026] After focusing and splitting the light waves sequentially, strong and weak beams are generated, facilitating subsequent detection and processing.

[0027] By sequentially filtering, focusing, and high-magnification processing the strong beam, an enhanced ultraviolet light detection image can be generated. The flame image is then processed to prepare for subsequent fusion.

[0028] The ultraviolet band strong light enhancement detection image is shaped according to a preset ratio and then fused with the weak light reflection image to generate a video image that includes the background object and the magnified flame.

[0029] Simulation experiment analysis: To verify the effectiveness of the system proposed in this application, the imaging effects at different distances and scales are described in detail. (Note: To enable the three cameras to image and detect simultaneously from the same location, they are designed inside the same cavity. Due to the limited space inside the cavity and the relatively complex optical path structure, including factors such as beam splitting and dual-path focusing, the installation positions of the visible light detector and the infrared thermal imaging detector are mirrored. Therefore, their output images are also mirror images with the vertical axis as the reference. At the same time, since the focal lengths of the visible light lens and the infrared thermal lens are different, the field of view of their output images is also slightly different, but this does not affect the comparison of imaging effects.) When the flame is 1 cm in size, at a distance of 10 meters, the visible light imaging image is as follows: Figure 6 As shown in (a), only brightness features are visible; other features such as texture, shape contours, area variations, number of sharp corners, and flicker are almost invisible, making it easily susceptible to interference from brighter background objects, leading to misjudgment. Infrared thermal imaging images are as follows... Figure 6 As shown in (b), because the human body has a much larger surface area than the flame, the infrared radiation from the human body is much stronger than that from the flame, making it impossible to distinguish the position and shape of the flame. The final image obtained in this application is as follows. Figure 6 As shown in (c), since the UVC band of the flame is magnified by 20,000 times, the flame volume is magnified. After shaping, the position of the flame and its approximate size after magnification can be accurately seen.

[0030] When the flame is 1 cm in size, at a distance of 20 meters, the visible light imaging image is as follows: Figure 7 As shown in (a), the brightness characteristics are also very indistinct, making it difficult to accurately identify the flame using computer algorithms. Infrared thermal imaging images are as follows: Figure 7 As shown in (b), the infrared image of the flame is completely overwhelmed by the background infrared image. The final image of this application is as follows. Figure 7 As shown in (c), the flame in the UVC band can still be continuously displayed on the video image in real time. When the flame is 1 cm in size, at a distance of 30 meters, the visible light imaging and infrared thermal imaging images are as follows: Figure 8 As shown in (a) and 8(b), the light waves of other background objects are much larger than the light waves of the flame, therefore the flame shape cannot be detected at all. The final image obtained in this application is as follows: Figure 8 As shown in (c), due to the increase in distance, the overall area of ​​the flame in the UVC band is reduced after magnification, but it is still very obvious. After being integrated with visible light, it can display the accurate position and approximate shape in real time. In actual tests, it can accurately detect a weak flame at a scale of 1 cm at a distance of 50 meters.

[0031] In summary, the system proposed in this application has an extremely long imaging detection range for very weak early-stage flames. It can magnify the flame spectrum independently at high magnification without amplifying the spectra of other background objects, thus highlighting the flame characteristics. Furthermore, the system proposed in this application has wider applicability, especially in indoor and nighttime scenarios where UVC ultraviolet light is almost absent. Flames of different materials all radiate a certain intensity of ultraviolet light in the UVC band, allowing even extremely weak flames to be accurately detected at long distances.

[0032] In some embodiments, this application also provides a computer system including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0033] This application also provides a computer-readable storage medium for storing a computer program. This computer-readable storage medium can be applied to a computer device, and the computer program causes the computer device to execute the corresponding processes in the methods described above in the embodiments of this application; for brevity, further details are omitted here.

[0034] The above embodiments are preferred implementations of this application. In addition, this application can be implemented in other ways. Any obvious substitutions without departing from the concept of this technical solution are within the protection scope of this application.

[0035] To facilitate understanding by those skilled in the art of the improvements made by this application compared to the prior art, some of the accompanying drawings and descriptions have been simplified, and for clarity, some other elements have been omitted from this application. Those skilled in the art should realize that these omitted elements may also constitute the content of this application.

Claims

1. A detection system for early-stage, extremely weak flame imaging and localization, characterized in that, The detection system includes a horizontal optical path branch, a vertical optical path branch, and a signal processing module; The horizontal optical path branch is a first focusing lens, a beam splitter, an ultraviolet filter, a second focusing lens, an image intensifier, and a first image detector that are coaxially connected in the horizontal direction. The vertical optical path branch is located in a direction perpendicular to the horizontal optical path, and is sequentially connected to the beam-splitting surface of the beam-splitting prism by the third focusing lens and the second image detector. The signal processing module is equipped with a video circuit processing board, which is electrically connected to the first image detector and the second image detector respectively. The image uploaded by the first image detector is shaped according to a preset ratio and then fused with the image uploaded by the second image detector to generate a video image containing the background object and the magnified flame.

2. The system according to claim 1, characterized in that, The beam splitter divides the light beam passing through the first focusing lens into a strong beam and a weak beam at an energy ratio of 8:

2. The high-intensity light beam is input into the ultraviolet filter; The weak beam is input into the third focusing lens.

3. The system according to claim 2, characterized in that, The magnification of the image intensifier is 20,000 times.

4. The system according to claim 3, characterized in that, The image intensifier includes a photocathode, a microchannel plate, and a fluorescent screen connected in sequence.

5. A detection method for early-stage, extremely weak flame imaging and localization based on the system described in claim 1, characterized in that, include: Acquire the light waves generated by the flame and other background objects; Based on the light waves, strong and weak beams are determined; Based on the strong beam, an ultraviolet band strong light enhancement detection image is generated; Based on the weak light beam, a weak light reflection image is generated; The ultraviolet band strong light enhancement detection image is shaped according to a preset ratio and then fused with the weak light reflection image to generate a video image that includes the background object and the magnified flame.

6. The method according to claim 5, characterized in that, The determination of strong and weak light beams based on the light waves includes: After the light waves are focused and split sequentially, strong beams and weak beams are generated.

7. The method according to claim 5, characterized in that, The generation of an ultraviolet band intensity-enhanced detection image based on the intense light beam includes: The strong beam is sequentially filtered, focused, and magnified to generate an ultraviolet band strong light enhanced detection image.

8. A computer system, characterized in that, include: Memory is used to store instructions that can be executed by the processor; A processor for executing the instructions to implement the method as described in any one of claims 5 to 7.

9. A computer-readable medium, characterized in that, The system contains computer program code that, when executed by a processor, implements the method as described in any one of claims 5 to 7.

Images