Electrical equipment infrared spectrum temperature coding method based on color space transformation

Abstract

The invention relates to the technical field of electrical equipment infrared spectrum temperature coding, in particular to an electrical equipment infrared spectrum temperature coding method based on color space transformation, which improves the traditional color space temperature coding according to the characteristics of a measured image, and carries out temperature coding transformation and gray function transformation based on color space, so as to obtain a temperature coding result. And an adaptive threshold theory is applied to gray segments of color space temperature coding, and adaptive selection of a color transformation interval is realized by using a maximum between-class method. According to the specific characteristics of the infrared measurement image, the traditional color space temperature coding is transformed, the adaptive color space temperature coding of the infrared measurement image is realized, the problem that the color space temperature coding quality is restricted by the gray scale imbalance of the infrared measurement image is solved, the temperature change of the electrical equipment can be truly reflected, and the accuracy of the color space temperature coding is improved. Moreover, the method has good self-adaptability, and lays a foundation for the efficient and accurate thermal fault detection of the electrical equipment.

Description

Technical Field

[0001] The invention relates to the technical field of infrared spectrum temperature coding of electrical equipment, and in particular to an infrared spectrum temperature coding method of electrical equipment based on color space transformation. Background Art

[0002] The operation and maintenance process of electrical equipment is mainly to intuitively grasp the relevant data of the equipment under the operating state, and timely and effectively perceive the equipment's faults and hidden dangers, etc., and use this as a basis to evaluate the working conditions of the electrical equipment and its operating development trend, so as to extend the service life of the equipment and ensure that it can give full play to its expected effect in the operating environment. With the construction of smart grids, the power system has put forward higher requirements for remote monitoring of the safe operation of electrical equipment. In order to ensure the safe production of electrical equipment and improve the reliability of electrical equipment operation, it is necessary to remotely monitor the temperature of overheated parts of electrical equipment online to achieve temperature limit alarm. As a common type of fault, thermal fault of electrical equipment can be diagnosed in a timely and accurate manner using infrared spectrum, which has important practical significance for the safe and stable operation of electrical equipment.

[0003] At present, the main status and problems of infrared spectrum technology for electrical equipment at home and abroad are summarized as follows:

[0004] 1) Based on infrared images, the fault diagnosis method of transmission lines is studied. First, the LS D line segment detection method is used to extract the conductors in the infrared image, and the deep convolutional neural network is used to extract the insulators in the infrared image, so as to obtain the line component area. The temperature and grayscale analysis are performed in the line component area to extract the hot spots, and the flood filling algorithm is used to segment the overheating area, extract the skeleton scanning points, effective convex defects, and drainage lines to identify the defect type. However, this method often has the defects of low recognition accuracy and low generalization level of the model.

[0005] 2) Based on the model of extracting the directional gradient histogram features of the target to be tested by support vector machine, a transmission line insulator fault diagnosis method combining binary support vector machine (SVM) classifier and Bayesian optimization (BOA) is proposed for the classification and recognition of infrared spectra during insulator flashover. By extracting the directional gradient histogram features in the infrared spectra of insulators, the Bayesian optimization algorithm is used to obtain the optimal hyperparameters of the diagnosis model to improve the accuracy of the classification algorithm, and the principal component analysis method is used to reduce the dimension of the extracted features to improve the efficiency of the classification algorithm. However, there is no clear conclusion on the recognition effect of this method for multiple targets.

[0006] 3) Based on the improved watershed algorithm of morphological labeling and MSRM, the image reconstruction is realized by using morphological hybrid opening and closing operations, and the image boundary is strengthened by Laplace sharpening. Then, the target and background are marked, the amplitude image is corrected, and the image is roughly segmented by watershed transformation. Then, the effective segmentation of the target and background is realized by using the MSRM regional merging method. This method has a good segmentation effect on infrared images of electrical equipment with large temperature difference and simple background, but it is easy to over-segment when the contrast is low or the fault is in a small area.

[0007] At present, the main problems of infrared spectrum technology for electrical equipment are concentrated on the single target device identification in the process of abnormal heating fault diagnosis of power equipment, the large number of infrared data set samples, the low average recognition accuracy and the low recognition rate. Summary of the invention

[0008] In order to solve the problems of single target device identification, huge number of infrared data set samples, low average recognition accuracy and low recognition rate in the process of abnormal heating fault diagnosis of power equipment, the technical focus of the present invention is to improve the traditional color space temperature encoding according to the characteristics of the measured image, and apply the adaptive threshold theory to the grayscale segmentation of color space temperature encoding, and propose a temperature encoding method for infrared spectrum of electrical equipment based on color space transformation.

[0009] In order to achieve the purpose of the present invention, the technical scheme adopted is: an infrared spectrum temperature encoding method for electrical equipment based on color space transformation, including: improving the traditional color space temperature encoding according to the characteristics of the measured image, performing temperature encoding transformation and grayscale function transformation based on color space, and applying the adaptive threshold theory to the grayscale segmentation of color space temperature encoding, and using the maximum between-class method to achieve adaptive selection of color transformation intervals. The color image transformed by this method can truly reflect the temperature changes of electrical equipment and has good adaptability.

[0010] As an optimization scheme of the present invention, improving the traditional color space temperature encoding means setting the area with low gray level near blue and the area with high gray level near red when performing pseudo-color transformation on the infrared measurement image, thus forming an encoding method in the order of red, orange, yellow, green and blue. This encoding method of transitioning from warm colors to cold colors better reflects the process of the target changing from high temperature to low temperature.

[0011] As an optimization scheme of the present invention, performing temperature coding transformation based on color space means: based on traditional electrical equipment temperature coding, dividing the area of ​​the infrared measurement image into four parts, corresponding to: low-temperature objects are blue, medium-low temperature objects are green, medium-temperature objects are yellow and high-temperature objects are red, performing electrical equipment temperature coding transformation based on color space, and converting the temperature of the electrical equipment into corresponding color distribution.

[0012] As an optimization scheme of the present invention, the grayscale function transformation refers to: the low temperature area is a blue primary color transformation; the first half of the medium and low temperature area is a change with the blue primary color as the main color and the green primary color as the auxiliary color, and the second half is a change with the green primary color as the main color and the blue primary color as the auxiliary color; the medium temperature area is a red primary color transformation, the green primary color remains unchanged, and a yellow coding value is generated; the high temperature area is a red primary color, and the green primary color input linearly decreases to generate a coding value for the high temperature area, which can simply distinguish the target from the background.

[0013] As an optimization scheme of the present invention, the adaptive threshold theory is applied to the grayscale segmentation of color space temperature coding, and the maximum inter-class method is used to realize the adaptive selection of color conversion intervals, which means: the grayscale range of the four fixed areas of the temperature coding of electrical equipment based on color space transformation is modified into four adaptive thresholds. The specific method is to first use the maximum inter-class variance method to obtain the binarization threshold of the image as the second area l2, and then respectively obtain the thresholds for the two parts as the first threshold l1 and the third threshold l3.

[0014] As an optimization scheme of the present invention, the red primary color transformation method is to map any grayscale level lower than the l2 area to the darkest red, linearly increase the red primary color between the l2 area and the l3 area, and keep the red primary color mapping from the l3 area to the 255 area unchanged, which is equivalent to the brightest red tone, to obtain the transformation function of the red primary color, and synthesize the transformation function of the red primary color, the transformation function of the green primary color, and the transformation function of the blue primary color to obtain a synthetic transformation function, and use the synthetic transformation function to perform pseudo color transformation on the infrared measurement image.

[0015] As an optimization scheme of the present invention, using the maximum inter-class variance method to obtain the binarization threshold of the image includes: performing binary segmentation on the image to achieve the division of high and low temperature areas, and secondly, performing secondary segmentation on the high and low temperature areas respectively. Performing binary segmentation in the high temperature area can achieve the division of the target high temperature center and the target surrounding area, and performing binary segmentation in the low temperature area can achieve the fine division of background content and background, so as to adaptively correspond different grayscale areas of the infrared measurement image to different temperature areas.

[0016] The present invention has positive effects:

[0017] (1) The method of adaptively dividing the color change area proposed in the present invention makes the color space of the infrared measurement image clearly layered, and it is easy to distinguish the color change law of the image in different temperature areas, which conforms to the trend of target temperature change and better completes the color space transformation of the infrared measurement image.

[0018] (2) The present invention projects the three primary colors in the encoding process onto the grayscale image respectively, and analyzes the correctness of the encoding from the grayscale image. The infrared color image formed by the adaptive electrical equipment temperature encoding has clear layers and is easy to distinguish different temperature areas, which conforms to the temperature change trend. At the same time, the temperature edges are clearly distinguished, which can truly reflect the trend of temperature changes inside the electrical equipment.

[0019] (3) According to the specific characteristics of the infrared measurement image, the present invention transforms the traditional color space temperature coding and successfully uses the maximum inter-class method to realize the adaptive color space temperature coding of the infrared measurement image, thereby solving the problem that the grayscale imbalance of the infrared measurement image restricts the quality of the color space temperature coding and conforms to the temperature change law of the electrical equipment under test. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] 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 embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying creative work.

[0021] Figure 1 This is the grayscale function transformation curve provided in the second embodiment of the present invention. DETAILED DESCRIPTION

[0022] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0023] The terms "first" and "second" and the like in the specification and claims of the present invention and the above drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device including a series of steps or units is not limited to the listed steps or units, but may include steps or units that are not listed.

[0024] The invention discloses a method for temperature coding of infrared spectrum of electrical equipment based on color space transformation, comprising: improving the traditional color space temperature coding according to the characteristics of the measured image, performing temperature coding transformation and grayscale function transformation based on color space, applying adaptive threshold theory to the grayscale segmentation of color space temperature coding, and realizing adaptive selection of color transformation interval by using the maximum between-class method.

[0025] 1. Grayscale color conversion method, generally color mode refers to the color matching method of image color, and its theoretical basis is the three-primary color principle in chromaticity. The present invention maps each grayscale value of the image into a color gradient according to a certain functional relationship. Different grayscales correspond to different colors. Since any color can be synthesized by the three primary colors of red, green and blue in different proportions, the pseudo-color processing of the image first sets the functional relationship of the three primary colors of red, green and blue, so that each corresponding function has corresponding red, green and blue outputs, and then the three are synthesized into one color, namely pseudo-color.

[0026] 2. Color selection for pseudo-color temperature coding of electrical equipment. In order to make the obtained pseudo-color accurately reflect the temperature variation range in the infrared image and the transformed pseudo-color image conform to human visual habits, it is necessary to use the transformation function to perform pseudo-color transformation on the infrared measurement image, set the area with low grayscale near blue, and the area with high grayscale near red. The larger the grayscale value, the higher the corresponding temperature. In the area with low grayscale value, blue is used as the main color, and in the area with high grayscale value, red is used as the main color. In this way, a coding method in the order of red, orange, yellow, green and blue is formed. This temperature coding method with transition from warm colors to cold colors better reflects the process of target changing from high temperature to low temperature.

[0027] 3. Grayscale region adaptive segmentation principle. Since the infrared measurement image mainly processes the measurement target, the overall display effect of the image is not high. Therefore, the present invention adopts the grayscale segmentation principle. First, the image is binary segmented, which can basically achieve the purpose of segmenting the target and the background into two regions. The actual segmentation can generally achieve the division of high and low temperature regions. Secondly, the high and low temperature regions are segmented twice respectively. The binary segmentation is performed in the high temperature region to achieve the division of the target high temperature center and the target surrounding area. The binary segmentation is performed in the low temperature region to achieve the division of the background content and the background. In this way, the different grayscale regions of the infrared measurement image can be adaptively corresponded to different temperature regions.

[0028] Embodiment 1

[0029] Although the traditional electrical equipment temperature coding transformation image is bright and colorful, the main purpose of measuring image processing is to timely and accurately find high temperature targets among normal temperature targets. Therefore, this transformation method has too many levels and it is difficult to find the target. The grayscale distribution of infrared measurement images is also difficult to balance. Therefore, the transformed pseudo-color image cannot achieve the purpose of enhancing the image target. Therefore, the traditional electrical equipment temperature coding must be improved. When the infrared measurement image is pseudo-colored, the low grayscale area is set near blue, and the high grayscale area is set near red. In this way, a coding method in the order of red, orange, yellow, green, and blue is formed. This coding method of transition from warm colors to cold colors better reflects the process of the target changing from high temperature to low temperature. Based on the above principles and based on the traditional electrical equipment temperature coding, the area of ​​the infrared measurement image is divided into four parts, corresponding to: low temperature objects l0~l1 (blue), medium and low temperature objects l1~l2 (green), medium temperature objects l2~l3 (yellow) and high temperature objects l3~l4 (red). The electrical equipment temperature coding transformation based on the color space is performed to convert the temperature of the electrical equipment into the corresponding color distribution. The transformation function is as follows:

[0030]

[0031] Among them: R(x,y), G(x,y), B(x,y) are the values ​​corresponding to the three primary colors red, green and blue respectively, l0, l1, l2, l3, l4 are the areas corresponding to different temperatures respectively, and f(x) is a certain area in the infrared measurement image of the electrical equipment to be temperature coded.

[0032] Embodiment 2

[0033] Corresponding to the temperature encoding transformation function based on color space disclosed in the first embodiment of the present invention, the second embodiment of the present invention also provides a specific implementation method of grayscale function transformation, see Figure 1 , the method comprising:

[0034] After completing the temperature coding transformation, it is necessary to perform grayscale function transformation. The transformation curve is as follows: Figure 1 shown.

[0035] Figures (a), (b) and (c) are the transformation function curves of the three primary colors of red, green and blue respectively, and Figure (d) is a graph that plots the three transformation curves into one graph. As can be seen from Figure (a), the transformation method of the red primary color is to map any grayscale level below l2 to the darkest red, linearly increase the red primary color between l2 and l3, and keep the red primary color mapping in the area from l3 to 255 unchanged, which is equivalent to the brightest red tone. The mapping method of other primary colors is similar. The transformation function of the red primary color in Figure (a), the transformation function of the green primary color in Figure (b), and the transformation function of the blue primary color in Figure (c) are synthesized to obtain the synthetic transformation function of Figure (d). It can be obtained that the low temperature area is a blue primary color transformation; the first half of the medium and low temperature area is a change with the blue primary color as the main color and the green primary color as the auxiliary color, and the second half is a change with the green primary color as the main color and the blue primary color as the auxiliary color; the medium temperature area is a linear change of the red primary color, and the green primary color remains unchanged, generating a yellow coding value; the high temperature area is a red primary color as the main color, and the green primary color input is linearly reduced to generate the coding value of the high temperature area. The advantage of these transformations is that once red appears in the image, it indicates the presence of a high-temperature object, and the target can be easily distinguished from the background.

[0036] Embodiment 3

[0037] Corresponding to the temperature encoding transformation function based on color space disclosed in the first embodiment of the present invention, the third embodiment of the present invention further provides a specific implementation method of the grayscale region adaptive segmentation principle, the method comprising:

[0038] The present invention adopts the maximum inter-class variance algorithm to calculate the binarization threshold of the image. Suppose the number of pixels with gray value i in the image f is n. i , the gray level of the image is L, and the total number of pixels is The probability of each gray level appearing is Assume that the grayscale k is used as the threshold to divide the image into two regions. The pixels with grayscales from 1 to k belong to the foreground region, denoted as A, and the pixels with grayscales from k+1 to L belong to the background region, denoted as B. Then the probabilities of region A and region B are Define the average grayscale of area A and area B as The average grayscale of image f Then the inter-class variance of the foreground area and the background area is σ 2 =W A (U A ―U) 2 +W B (U B ―U) 2 , according to the maximum inter-class variance criterion, change k from 1 to L and calculate the inter-class variance σ 2 , k that maximizes the above formula is the threshold l for region segmentation j , j is the number of segmented regions.

[0039] Then, the grayscale ranges of the four fixed areas of the temperature coding of electrical equipment based on color space transformation are modified to the four adaptive thresholds obtained according to the above formula. The specific method is: first, the maximum inter-class variance method is used to obtain the binarization threshold of the image as the second area l2, and then the thresholds are obtained for the two parts respectively as the first threshold l1 and the third threshold l3. Binary segmentation of the image can achieve more precise division of high and low temperature areas. Secondly, secondary segmentation is performed on the high and low temperature areas respectively. Binary segmentation is performed in the high temperature area to achieve the division of the target high temperature center and the target surrounding area. Binary segmentation is performed in the low temperature area to achieve the fine division of background content and background. In this way, different grayscale areas of the infrared measurement image can be adaptively corresponded to different temperature areas, and high-precision infrared spectrum temperature coding of electrical equipment based on color space transformation can be achieved.

[0040] The present invention uses temperature coding to diagnose thermal faults based on the color space transformation and distribution characteristics of infrared measurement images of electrical equipment. At the same time, the maximum class interval method is used to achieve adaptive selection of color transformation intervals, and an infrared measurement image adaptive electrical equipment color space temperature coding method is proposed. The color image transformed by this method can truly reflect the temperature change of electrical equipment and has good adaptability.

[0041] The present invention realizes a temperature coding method of infrared spectrum of electrical equipment based on color space transformation. Infrared thermal imaging technology is a basic and effective method for diagnosing thermal faults of electrical equipment due to its non-contact and non-destructive advantages, and the imaging environment is not restricted by weather or lighting conditions. The temperature of electrical equipment is encoded by color space transformation and infrared temperature measurement, which lays a foundation for realizing efficient and accurate thermal fault diagnosis of electrical equipment.

[0042] The specific embodiments described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for encoding temperature of infrared spectrum of electrical equipment based on color space transformation, characterized in that: include: According to the characteristics of the measured image, the traditional color space temperature coding is improved, and temperature coding transformation and grayscale function transformation based on color space are carried out. The adaptive threshold theory is applied to the grayscale segmentation of color space temperature coding. The maximum between-class method is used to realize the adaptive selection of color transformation interval.

2. The method for temperature encoding of infrared spectrum of electrical equipment based on color space transformation according to claim 1, characterized in that: The improvement of the traditional color space temperature coding refers to setting the area with low gray level near blue and the area with high gray level near red when performing pseudo color transformation on the infrared measurement image, thus forming a coding method in the order of red, orange, yellow, green and blue. The coding method of transition from warm colors to cold colors reflects the process of the target changing from high temperature to low temperature.

3. The method for temperature encoding of infrared spectrum of electrical equipment based on color space transformation according to claim 2, characterized in that: The temperature coding transformation based on color space refers to: based on the traditional electrical equipment temperature coding, the area of ​​the infrared measurement image is divided into four parts, corresponding to: low-temperature objects are blue, medium-low temperature objects are green, medium-temperature objects are yellow and high-temperature objects are red, and the electrical equipment temperature coding transformation based on color space is performed to convert the temperature of the electrical equipment into corresponding color distribution.

4. The method for temperature encoding of infrared spectrum of electrical equipment based on color space transformation according to claim 2, characterized in that: The grayscale function transformation refers to: the low temperature area is a blue primary color transformation; the first half of the medium and low temperature area is a change with the blue primary color as the main color and the green primary color as the auxiliary color, and the second half is a change with the green primary color as the main color and the blue primary color as the auxiliary color; the medium temperature area is a red primary color transformation, the green primary color remains unchanged, and a yellow coding value is generated; the high temperature area is a red primary color as the main color, and the green primary color input linearly decreases to generate the coding value of the high temperature area, which distinguishes the target from the background.

5. The method for temperature encoding of infrared spectrum of electrical equipment based on color space transformation according to claim 2, characterized in that: The application of the adaptive threshold theory to the grayscale segmentation of the color space temperature coding and the use of the maximum inter-class method to achieve the adaptive selection of the color transformation interval refers to: modifying the grayscale range of the four fixed areas of the temperature coding of the electrical equipment based on the color space transformation into four adaptive thresholds. The specific method is to first use the maximum inter-class variance method to obtain the binarization threshold of the image as the second area l2, and then respectively obtain the thresholds for the two parts as the first threshold l1 and the third threshold l3.

6. The method for temperature encoding of infrared spectrum of electrical equipment based on color space transformation according to claim 5, characterized in that: include: The red primary color transformation method is to map any gray level lower than the l2 area to the darkest red, linearly increase the red primary color between the l2 area and the l3 area, and keep the red primary color mapping from the l3 area to the 255 area unchanged, which is equivalent to the brightest red tone, to obtain the transformation function of the red primary color, and synthesize the transformation function of the red primary color, the transformation function of the green primary color, and the transformation function of the blue primary color to obtain a synthetic transformation function, and use the synthetic transformation function to perform pseudo color transformation on the infrared measurement image.

7. The method for temperature encoding of infrared spectrum of electrical equipment based on color space transformation according to claim 5, characterized in that: The method of obtaining the binarization threshold of the image using the maximum inter-class variance method includes: performing binary segmentation on the image to divide the high and low temperature areas, and then performing secondary segmentation on the high and low temperature areas respectively. Binary segmentation is performed in the high temperature area to achieve the division of the target high temperature center and the target surrounding area, and binary segmentation is performed in the low temperature area to achieve the fine division of the background content and the background, so that different grayscale areas of the infrared measurement image are adaptively matched to different temperature areas.

Images