Method and system for merging suspected areas in flame detection of pre-oxidation furnace

Abstract

The invention relates to the technical field of carbon fiber production, in particular to a suspected region merging method and system in flame detection of a pre-oxidation furnace, and the method comprises the steps: obtaining suspected regions recognized in flame detection, and taking a pixel point in each suspected region as a five-dimensional vector containing color information and spatial position information; selecting a pixel point with the minimum gradient from each suspected region as a feature pixel point of the region; taking the five-dimensional vector of the selected feature pixel point as a classification standard, and carrying out iterative classification and aggregation operation on the suspected region until region classification is completely converged; and carrying out feature recognition on the remaining suspected area after polymerization, and judging whether the suspected area is flame. According to the method, through five-dimensional vector representation and iterative classification aggregation, the number of suspected areas needing to be analyzed one by one can be reduced, so that the calculation time is shortened, the pixel point with the minimum gradient is selected as the feature, the influence of noise and interference can be reduced, and feature recognition is more reliable.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon fiber production, and in particular to a method and system for merging suspected areas in flame detection of a pre-oxidation furnace. Background Art

[0002] The process of carbon fiber pre-oxidation is to combine oxygen molecules in the air with PAN-based fibers under high temperature environment, so that the polymer chains begin to cross-link, and the fiber density increases from 1.18g / cc to 1.38g / cc. Temperature control is very important. Usually the temperature range of the pre-oxidation furnace is between 200℃ and 300℃. However, in a high temperature environment, if there are joints on the carbon fiber or the temperature field or wind field in the furnace is uneven, the local temperature of the fiber will be too high, causing the tow to catch fire, thereby causing deflagration in the furnace. Since once the flame is generated, it is a deflagration, it is necessary to detect the flame in a very short time and take fire extinguishing measures.

[0003] In the prior art, there is a method of identifying and detecting the flame in the pre-oxidation furnace by visual means. First, an industrial camera is used to capture an image of the oxidation furnace, and the corresponding suspected flame area is segmented by the corresponding threshold. Because the furnace is completely black, the threshold set during the threshold segmentation is generally small. Then, the characteristics of the pixels in the suspected area are judged, and finally it is determined whether it is a flame or interference.

[0004] However, there will be a problem when further feature judgment is performed on the suspected flame. That is, if there is only one suspected area segmented by the threshold, then only this one area needs to be feature judged subsequently. If there are many suspected areas segmented by the threshold, it is necessary to perform feature judgment on each area to determine whether these areas are flame areas. In the process of feature judgment on each area, a lot of computing time is required, which affects the judgment time of the flame and makes it impossible to identify the flame and take fire extinguishing measures in a very short time.

[0005] The information disclosed in this background technology section is only intended to deepen the understanding of the overall background technology of the present invention, and should not be regarded as acknowledging or suggesting in any form that the information constitutes the prior art already known to those skilled in the art. Summary of the invention

[0006] The present invention provides a method for merging suspected areas in flame detection of a pre-oxidation furnace, thereby effectively solving the problems in the background technology.

[0007] In order to achieve the above object, the technical solution adopted by the present invention is: a method for merging suspected areas in flame detection of a pre-oxidation furnace, comprising the following steps: Obtaining the suspected area identified in the flame detection, and treating each pixel point in the suspected area as a five-dimensional vector containing color information and spatial position information; Selecting a pixel point with the smallest gradient from each of the suspected regions as a characteristic pixel point of the region; Using the five-dimensional vector of the selected characteristic pixel point as a classification standard, performing iterative classification and aggregation operations on the suspected area until the area classification is completely converged; Perform feature recognition on the remaining suspected area after aggregation to determine whether it is a flame.

[0008] Furthermore, the pixel point in each suspected area is regarded as a five-dimensional vector including color information and spatial position information, including: The five-dimensional vector Q (R, G, B, X, Y) of the pixel point Q in the suspected area, wherein R is the red component of the pixel point, G is the green component of the pixel point, B is the blue component of the pixel point, X is the horizontal coordinate of the pixel point, and Y is the vertical coordinate of the pixel point.

[0009] Furthermore, selecting a pixel point with the smallest gradient from each of the suspected regions as a characteristic pixel point of the region includes: Calculate the grayscale difference between the pixel point in the suspected area and the pixel points in its eight neighborhoods, and use the grayscale difference of the eight neighborhoods of the pixel point as the gradient value of the pixel point; Traverse the suspected area and find the pixel with the smallest gradient value as the feature pixel.

[0010] Furthermore, the iterative classification and aggregation operation is performed on the suspected region until the region classification is completely converged, comprising the following steps: The feature pixel points q1, q2, ...qn selected in the suspected area are represented as five-dimensional vectors respectively; Select the characteristic pixels q1 and q2 of the first two suspected areas as the initial classification centers; For other suspected areas, they are classified according to the Euclidean distance between their characteristic pixels and the two initial classification centers, and the suspected areas are initially divided into two categories; After the preliminary classification, the new centers of the two types of regions are calculated, and the suspected regions are classified by Euclidean distance using the two new centers, and the suspected regions are reclassified into two categories; The center calculation and classification process is repeated until the classification result no longer changes.

[0011] Furthermore, the Euclidean distance is calculated as follows: ; In the formula, d ij For the iThe feature pixel and j The Euclidean distance between the centers, i From 3 to n, j is 1 to 2; R i , G i , B i , X i , Y i Respectively i The five-dimensional vector of feature pixels, R i , G i , B i , X i , Y i Respectively j A five-dimensional vector with a center.

[0012] Furthermore, the classification according to the Euclidean distance between the characteristic pixel point and the two initial classification centers includes: Determine the size of the Euclidean distance between the feature pixel and the two initial classification centers, and divide the suspected area into the classification center with the smallest Euclidean distance.

[0013] Furthermore, after the preliminary classification, new centers of the two types of regions are calculated, including: The new central five-dimensional vector is the mean of the five-dimensional vectors of the feature pixels of all suspected areas in the classification.

[0014] The present invention also includes a system for merging suspected areas in flame detection of a pre-oxidation furnace, using the above method, the system comprising: An acquisition unit, used to acquire the suspected area identified in the flame detection, and regard each pixel point in the suspected area as a five-dimensional vector containing color information and spatial position information; A feature pixel point extraction unit, used to select a pixel point with the smallest gradient from each of the suspected areas as a feature pixel point of the area; A classification aggregation unit, used for taking the five-dimensional vector of the selected characteristic pixel point as a classification standard, and performing iterative classification aggregation operation on the suspected area until the area classification is completely converged; The flame recognition unit is used to perform feature recognition on the suspected area remaining after aggregation to determine whether it is a flame.

[0015] The present invention also includes a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method described above is implemented.

[0016] The present invention also includes a storage medium on which a computer program is stored. When the computer program is executed by a processor, the method described above is implemented.

[0017] The beneficial effects of the present invention are as follows: the present invention can reduce the number of suspected areas that need to be analyzed one by one through five-dimensional vector representation and iterative classification aggregation, thereby shortening the calculation time, and selecting pixel points with the smallest gradient as features, which can reduce the influence of noise and interference, making feature recognition more reliable, and can achieve the purpose of rapid response through efficient suspected area merging and flame detection in complex environments, and is suitable for pre-oxidation furnace application scenarios that require real-time flame detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 is a flow chart of the method of the present invention; Figure 2 It is a structural schematic diagram of the system of the present invention; Figure 3 It is a schematic diagram of the structure of the computer device of the present invention. DETAILED DESCRIPTION

[0020] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0021] like Figure 1 As shown: A method for merging suspected areas in flame detection of a pre-oxidation furnace comprises the following steps: Obtain the suspected area identified in the flame detection, and regard the pixel points in each suspected area as a five-dimensional vector containing color information and spatial position information; Select the pixel with the smallest gradient from each suspected area as the feature pixel of the area; The five-dimensional vector of the selected feature pixel points is used as the classification standard, and the suspected area is iteratively classified and aggregated until the regional classification is completely converged; Perform feature recognition on the remaining suspected area after aggregation to determine whether it is a flame.

[0022] Through five-dimensional vector representation and iterative classification aggregation, the number of suspected areas that need to be analyzed one by one can be reduced, thereby shortening the calculation time. Selecting the pixel point with the smallest gradient as the feature can reduce the influence of noise and interference, making feature recognition more reliable. In a complex environment, it can achieve the purpose of rapid response through efficient suspected area merging and flame detection. It is suitable for pre-oxidation furnace application scenarios that require real-time flame detection.

[0023] In this embodiment, the pixel point in each suspected area is regarded as a five-dimensional vector containing color information and spatial position information, including: The five-dimensional vector Q (R, G, B, X, Y) of the pixel point Q in the suspected area, where R is the red component of the pixel point, G is the green component of the pixel point, B is the blue component of the pixel point, X is the horizontal coordinate of the pixel point, and Y is the vertical coordinate of the pixel point.

[0024] The three-dimensional color information of RGB can accurately describe the color of the pixel and help identify the difference between the flame area and the background. Since flames usually have unique color characteristics in images (such as orange-red), these color components play a key role in the classification process. Spatial position information (X, Y) can help the algorithm determine whether adjacent pixels may belong to the same area. Combined with color information, spatial position can further reduce misjudgment, such as separating non-flame areas with similar colors but located in different positions.

[0025] This preferred embodiment uses a five-dimensional vector to represent each pixel, combining color and spatial information, making the aggregation and classification of suspected flame areas more accurate and effective. This method can significantly reduce the possibility of misclassification and improve the efficiency and accuracy of flame detection, and is particularly suitable for pre-oxidation furnace environments that require rapid flame detection and fire extinguishing measures.

[0026] Among them, the pixel point with the smallest gradient is selected from each suspected area as the characteristic pixel point of the area, including: Calculate the grayscale difference between the pixel point in the suspected area and its eight neighboring pixels, and use the grayscale difference of the eight neighboring pixels of the pixel point as the gradient value of the pixel point; Traverse the suspected area and find the pixel with the smallest gradient value as the feature pixel.

[0027] Compare the grayscale value of the pixel with the grayscale values ​​of its eight neighboring pixels and calculate the difference in each direction. The formula for grayscale difference is as follows: ; Where I(P) represents the gray value of pixel P, I(Pi ) represents the grayscale value of its eight neighboring pixels, and i represents the index of the neighboring pixel. This gradient value reflects the degree of change of the grayscale around the pixel. The smaller the gradient value, the smaller the grayscale change of the pixel is, and it is more likely to be a stable pixel in the region, representing the overall characteristics of the region.

[0028] Selecting the pixel with the smallest gradient can reduce the amount of subsequent calculations. Compared with performing complex feature analysis on the entire suspected area, directly extracting features from pixels with smaller gradient values ​​can significantly improve calculation efficiency.

[0029] As a preferred embodiment of the above, performing iterative classification and aggregation operations on the suspected regions until the region classification is completely converged includes the following steps: The feature pixel points q1, q2, ...qn selected in the suspected area are represented as five-dimensional vectors respectively; Select the characteristic pixels q1 and q2 of the first two suspected areas as the initial classification centers; For other suspected areas, they are classified according to the Euclidean distance between their characteristic pixels and the two initial classification centers, and the suspected areas are initially divided into two categories; After the preliminary classification, the new centers of the two types of regions are calculated, and the suspected regions are classified by Euclidean distance using the two new centers, and the suspected regions are reclassified into two categories; The center calculation and classification process is repeated until the classification result no longer changes.

[0030] Through iteration, the classification results can be dynamically adjusted according to the characteristics of the data. The initial classification may not be very accurate, but through continuous iteration, it can gradually converge to a more reasonable classification. In flame detection, quickly and accurately identifying the flame area is the key. By clustering and merging the suspected areas first, the number of areas that require further feature judgment can be greatly reduced, thereby accelerating the flame detection process.

[0031] Among them, the Euclidean distance is calculated as follows: ; In the formula, d ij For the i The feature pixel and j The Euclidean distance between the centers, i From 3 to n, j is 1 to 2; R i , G i , B i , X i , Yi Respectively i The five-dimensional vector of feature pixels, R i , G i , B i , X i , Y i Respectively j A five-dimensional vector with a center.

[0032] In this embodiment, classification is performed based on the Euclidean distance between its characteristic pixel points and two initial classification centers, including: Determine the size of the Euclidean distance between the feature pixel and the two initial classification centers, and divide the suspected area into the classification center with the smallest Euclidean distance.

[0033] By calculating and comparing the distances between the characteristic pixels of the suspected area and multiple classification centers, it can be ensured that each area belongs to the closest center, reducing the possibility of misclassification and improving the accuracy of classification.

[0034] As a preferred embodiment of the above, after the preliminary classification, the new centers of the two types of regions are calculated, including: The new central five-dimensional vector is the mean of the five-dimensional vectors of the feature pixels of all suspected areas in the classification.

[0035] Using the new classification center, repeat the previous Euclidean distance classification steps, that is, calculate the distance between the characteristic pixel points of each suspected area and the two new classification centers, and reclassify them; by continuously calculating and updating the classification center, ensure that the characteristic pixel points of each type of suspected area are close to the center, thereby enhancing the accuracy of classification; through mean calculation, the new classification center can better reflect the overall characteristics of the suspected area of ​​this type, avoid the deviation of the initial classification center selection, and ensure that the final classification is more representative; the classification center can be automatically adjusted to ensure that all suspected areas are accurately clustered.

[0036] like Figure 2 As shown, this embodiment also includes a suspected area merging system in the pre-oxidation furnace flame detection, using the above method, the system includes: An acquisition unit is used to acquire the suspected area identified in the flame detection, and regard the pixel points in each suspected area as a five-dimensional vector containing color information and spatial position information; A feature pixel point extraction unit is used to select a pixel point with the smallest gradient from each suspected area as a feature pixel point of the area; A classification aggregation unit is used to use the five-dimensional vector of the selected feature pixel points as a classification standard to perform iterative classification aggregation operations on the suspected area until the area classification is completely converged; The flame recognition unit is used to perform feature recognition on the suspected area remaining after aggregation to determine whether it is a flame.

[0037] See also Figure 3 A computer device 400 provided in an embodiment of the present application includes: a processor 410 and a memory 420, wherein the memory 420 stores a computer program executable by the processor 410, and when the computer program is executed by the processor 410, the above method is executed.

[0038] The embodiment of the present application further provides a storage medium 430 on which a computer program is stored. When the computer program is run by the processor 410, the above method is executed.

[0039] Among them, the storage medium 430 can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable red-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, disk or optical disk.

[0040] In the description of the present invention, the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more of the features. "Multiple" means two or more, unless otherwise clearly and specifically defined.

[0041] In the present invention, unless otherwise clearly specified and limited, the terms "installed", "connected", "connected", "fixed" and the like should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, it can be the internal connection of two elements or the interaction relationship between two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0042] In the description of this specification, the description with reference to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, those skilled in the art may combine and combine the different embodiments or examples described in this specification and the features of the different embodiments or examples, unless they are contradictory.

[0043] Any process or method description in a flowchart or otherwise described herein may be understood to represent a module, segment or portion of code that includes one or more executable instructions for implementing the steps of a specific logical function or process, and the scope of the preferred embodiments of the present invention includes alternative implementations in which functions may not be performed in the order shown or discussed, including performing functions in a substantially simultaneous manner or in the reverse order depending on the functions involved, which should be understood by those skilled in the art to which the embodiments of the present invention belong.

[0044] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in combination with these instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in combination with these instruction execution systems, devices or apparatuses. More specific examples (non-exhaustive list) of computer-readable media include the following: an electrical connection with one or more wires (electronic device), a portable computer disk box (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium and then editing, interpreting or processing in other suitable ways if necessary, and then stored in a computer memory.

[0045] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0046] A person skilled in the art may understand that all or part of the steps in the method for implementing the above-mentioned embodiment may be completed by instructing related hardware through a program, and the program may be stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiment.

[0047] The storage medium mentioned above may be a read-only memory, a magnetic disk or an optical disk, etc. Although the embodiments of the present invention have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limiting the present invention. A person of ordinary skill in the art may change, modify, replace and modify the above embodiments within the scope of the present invention.

Claims

1. A method for merging suspected areas in flame detection of a pre-oxidation furnace, characterized in that: The steps include: Obtaining the suspected area identified in the flame detection, and treating each pixel point in the suspected area as a five-dimensional vector containing color information and spatial position information; Selecting a pixel point with the smallest gradient from each of the suspected regions as a characteristic pixel point of the region; Using the five-dimensional vector of the selected characteristic pixel point as a classification standard, performing iterative classification and aggregation operations on the suspected area until the area classification is completely converged; Perform feature recognition on the remaining suspected area after aggregation to determine whether it is a flame.

2. The method for merging suspected areas in flame detection of a pre-oxidation furnace according to claim 1, characterized in that: The pixel point in each suspected area is regarded as a five-dimensional vector including color information and spatial position information, including: The five-dimensional vector Q (R, G, B, X, Y) of the pixel point Q in the suspected area, wherein R is the red component of the pixel point, G is the green component of the pixel point, B is the blue component of the pixel point, X is the horizontal coordinate of the pixel point, and Y is the vertical coordinate of the pixel point.

3. The method for merging suspected areas in flame detection of a pre-oxidation furnace according to claim 1, characterized in that: The step of selecting a pixel point with the smallest gradient from each of the suspected regions as a characteristic pixel point of the region includes: Calculate the grayscale difference between the pixel point in the suspected area and the pixel points in its eight neighborhoods, and use the grayscale difference of the eight neighborhoods of the pixel point as the gradient value of the pixel point; Traverse the suspected area and find the pixel with the smallest gradient value as the feature pixel.

4. The method for merging suspected areas in flame detection of a pre-oxidation furnace according to claim 2, characterized in that: The iterative classification and aggregation operation is performed on the suspected region until the region classification is completely converged, comprising the following steps: The feature pixels q1, q2, ...q selected in the suspected area n Respectively represented as five-dimensional vectors; Select the characteristic pixels q1 and q2 of the first two suspected areas as the initial classification centers; For other suspected areas, they are classified according to the Euclidean distance between their characteristic pixels and the two initial classification centers, and the suspected areas are initially divided into two categories; After the preliminary classification, the new centers of the two types of regions are calculated, and the suspected regions are classified by Euclidean distance using the two new centers, and the suspected regions are reclassified into two categories; The center calculation and classification process is repeated until the classification result no longer changes.

5. The method for merging suspected areas in flame detection of a pre-oxidation furnace according to claim 4, characterized in that: The Euclidean distance is calculated as follows: ； In the formula, d ij For the i The feature pixel and j The Euclidean distance between the centers, i From 3 to n, j is 1 to 2; R i , G i , B i , X i , Y i Respectively i The five-dimensional vector of feature pixels, R i , G i , B i , X i , Y i Respectively j A five-dimensional vector with a center.

6. The method for merging suspected areas in flame detection of a pre-oxidation furnace according to claim 4, characterized in that: The classification is performed based on the Euclidean distance between the characteristic pixel point and the two initial classification centers, including: Determine the size of the Euclidean distance between the feature pixel and the two initial classification centers, and divide the suspected area into the classification center with the smallest Euclidean distance.

7. The method for merging suspected areas in flame detection of a pre-oxidation furnace according to claim 4, characterized in that: After the preliminary classification, new centers of two types of regions are calculated, including: The new central five-dimensional vector is the mean of the five-dimensional vectors of the feature pixels of all suspected areas in the classification.

8. A suspected area merging system in pre-oxidation furnace flame detection, characterized in that: Using the method according to any one of claims 1 to 7, the system comprises: An acquisition unit, used to acquire the suspected area identified in the flame detection, and regard each pixel point in the suspected area as a five-dimensional vector containing color information and spatial position information; A feature pixel point extraction unit, used to select a pixel point with the smallest gradient from each of the suspected areas as a feature pixel point of the area; A classification aggregation unit, used for taking the five-dimensional vector of the selected characteristic pixel point as a classification standard, and performing iterative classification aggregation operation on the suspected area until the area classification is completely converged; The flame recognition unit is used to perform feature recognition on the suspected area remaining after aggregation to determine whether it is a flame.

9. A computer device comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images