A method for detecting abnormal surface temperature of an electric heater based on AI recognition

Abstract

The present invention belongs to the technical field of abnormal surface temperature detection of electric heaters, and particularly relates to a method for detecting abnormal surface temperature of electric heaters based on AI recognition. First, images are collected through infrared thermal imaging technology, and images from different perspectives are fused by a weighted average method optimized by weights. The weights are determined according to the evaluation of image quality in terms of clarity and contrast, and a comprehensive infrared thermal image is obtained through weighted averaging. Then, adaptive contrast adjustment, edge optimization, and sharpening processing are performed on the fused image. The local contrast enhancement algorithm is used to increase the dynamic range, multi-scale high-pass filtering is used to sharpen the edges, and non-linear sharpening is used to adjust the image according to gradient information, enhancing the resolution of temperature changes. Finally, data collected from multiple dimensions by sensors are weighted and fused after standardization, and the weights are dynamically adjusted through error evaluation. The present invention improves the accuracy and reliability of detecting abnormal surface temperature of electric heaters, ensures the safe and efficient operation of equipment, and effectively avoids potential safety hazards and performance losses caused by abnormal temperature.

Description

Technical Field

[0001] The present invention belongs to the technical field of abnormal surface temperature detection of electric heaters, and particularly relates to a method for detecting abnormal surface temperature of electric heaters based on AI recognition. Background Art

[0002] In many scenarios such as industrial production and household life, electric heaters are widely used. During the use of electric heaters, abnormal surface temperature may cause safety problems such as fires and equipment damage, and may also affect their working efficiency and service life. Therefore, accurate detection of abnormal surface temperature of electric heaters is crucial. Existing temperature detection methods have certain limitations. For example, some detection methods only rely on a single sensor to obtain data, with limited detection accuracy and unable to comprehensively reflect the actual situation of the surface temperature of electric heaters. Moreover, in terms of image acquisition, effective fusion processing of images from different perspectives may not be carried out, resulting in incomplete and inaccurate temperature information obtained. In addition, there is a lack of effective technical means for image processing to enhance the distinguishability of temperature changes, thus affecting the judgment of temperature abnormalities. Existing detection methods based on AI recognition still need to be improved in aspects such as image and sensor data fusion to further improve detection accuracy and meet actual requirements. The present invention aims to solve these problems and provide a more accurate and efficient method for detecting abnormal surface temperature of electric heaters. Summary of the Invention

[0003] The present invention addresses the technical problems existing in the above background art and proposes a method for detecting abnormal surface temperature of electric heaters based on AI recognition.

[0004] To achieve the above object, the technical solution adopted by the present invention includes the following steps:

[0005] S1. First, collect images through infrared thermal imaging technology and fuse images from different perspectives based on the weighted average method optimized by weights;

[0006] S2. Then, for the fused images, adopt the adaptive contrast adjustment technology and perform edge information optimization and sharpening processing to optimize image details and enhance the distinguishability of temperature changes;

[0007] S3. Finally, combine sensors to achieve multi-dimensional data acquisition, and perform weighted fusion on sensor data through a weighted algorithm, and improve the overall detection accuracy through the collaborative work of images and sensors;

[0008] The specific implementation steps of adopting the adaptive contrast adjustment technology and performing edge information optimization and sharpening processing in step S2 are as follows:

[0009] S21. First, for each local region of the image, calculate the contrast of this region. Using the local contrast enhancement algorithm, enhance the dynamic range of the image region through local calculation. In the local region range of the original image, the image after adaptive contrast enhancement where I enhanced1 (x, y) is the image after contrast optimization, and μ local (x, y) is the mean value of the local region, and σ local (x, y) is the standard deviation of the local region. C is the contrast enhancement factor, which is dynamically adjusted according to the contrast requirement of the image;

[0010] S22. Then, optimize the edge information in the infrared image to make the edges of the regions with larger temperature differences clearer. Decompose the I enhanced1 (x, y) image into images of multiple scales I scale (x, y), and then apply a high-pass filter to each scale to enhance the image edge information. The obtained local edge-enhanced image is: where α s is the weighting coefficient of each scale, H s represents the operation at scale s, and N is the number of scales;

[0011] S23. Finally, process the image through a non-linear sharpening method, and adjust the sharpening degree of the image through local gradient information: where I final (x, y) is the final processed image, is the Laplacian operator of the image, representing local edge information, is the gradient magnitude of the image, is the maximum gradient value, which is used for non-linear adjustment.

[0012] Preferably, the specific implementation of step S1 for fusing images from different perspectives based on the weighted average method optimized by weights is as follows:

[0013] S11. Perform quality assessment on the images I1, I2, I3...I n collected from each perspective, set a local weighting function, and calculate the local weighting value according to the quality of the images from each perspective. The weight ω i is the comprehensive score calculated according to the image sharpness C(I i ) and contrast D(I i ), and the formula is: where n is the total number of images;

[0014] S12. Use the weighted average method to perform image fusion, and perform weighted fusion according to the weights and qualities of each perspective, and finally generate a comprehensive infrared thermal image. The weighting coefficient ω i of each image Ii It is determined according to the calculation of the previous step, and the weighted fusion formula is: where I final (x, y) is the temperature value of the fused image at the coordinate point (x, y).

[0015] Preferably, the specific implementation of the contrast enhancement factor C in the step S21 is:

[0016] Preferably, the weighted coefficient α s for each scale in the step S22 is in the specific form of: where σ s is the standard deviation at scale s.

[0017] Preferably, the operation H s at scale s in the step S22 is a Gaussian filter at scale s:

[0018] Preferably, in the step S3, the sensor data is weighted and fused through a weighted algorithm. The implementation method for improving the overall detection accuracy by the collaborative work of the image and the sensor is as follows: First, ensure that the sensor and the image acquisition device work within the same time window, and perform standardization processing on them so that they can be jointly input into the fusion model, and then perform weighted average fusion: where is the fused result, is the standardized data of the i-th sensor, is the standardization of the image data, γ i is the weight of the i-th sensor, γ I is the weight of the image data; finally, error evaluation is performed by comparing the weighted fusion result with the actual situation, and the weights of each sensor and the image data are adjusted to ensure the final detection accuracy.

[0019] Compared with the prior art, the advantages and positive effects of the present invention are that multi-view image fusion obtains comprehensive information, solves the limitation of a single view, and has a wider detection coverage. The adaptive image optimization technology enhances the temperature resolution, clearly presents subtle changes, and helps to accurately judge abnormalities. The multi-dimensional data fusion coordinates the sensor and the image, and the comprehensive advantages improve the accuracy. Compared with single-source detection, the accuracy is significantly improved, and the weights can be dynamically adjusted to maintain stability. Overall, it effectively improves the accuracy and reliability of the surface temperature abnormality detection of the electric heater, and ensures the safe and efficient operation of the equipment. Specific Embodiments

[0020] To better understand the above objects, features, and advantages of the present invention, the present invention will be further described below in conjunction with embodiments. It should be noted that, without conflict, the embodiments of the present application and the features in the embodiments can be combined with each other.

[0021] In the following description, many specific details are set forth in order to provide a thorough understanding of the present invention. However, the present invention may be practiced in other ways different from those described herein. Therefore, the present invention is not limited by the specific embodiments disclosed in the following specification.

[0022] Embodiment: In industrial production and household life, electric heaters are widely used, but abnormal surface temperature may cause safety problems and affect their performance. Existing temperature detection methods have many limitations. For example, some rely only on a single sensor, with limited detection accuracy and unable to comprehensively reflect the actual temperature situation. During image acquisition, the fusion processing of images from different perspectives is poor, and the obtained temperature information is inaccurate. There is a lack of image enhancement technology, which affects the resolution of temperature changes and the judgment of abnormalities. The detection method based on AI recognition also needs improvement in the fusion of images and sensor data and is difficult to meet the requirements of accurate detection.

[0023] First, in order to obtain comprehensive and accurate surface temperature image information of the electric heater, the weighted average method based on weight optimization is used to fuse images from different perspectives. Existing image acquisition often obtains information only from a single perspective, unable to comprehensively reflect the surface temperature condition of the electric heater and easily missing local abnormally high or low temperature areas. The present invention uses the weighted average method based on weight optimization to fuse images from different perspectives, determines the weights through quality evaluation, and can highlight important perspective information. First, the images I1, I2, I3...I n collected from each perspective are subjected to quality evaluation, a local weighting function is set, and the local weighting value is calculated according to the quality of each perspective image. The weight ω i is the comprehensive score calculated according to the image clarity C(I i ) and the contrast D(I i ), and the formula is: where n is the total number of images; then the weighted average method is used for image fusion, and weighted fusion is performed according to the weights and qualities of each perspective, and finally a comprehensive infrared thermal image is generated. The weighting coefficient ω i of each image I i is determined according to the calculation of the previous step, and the weighted fusion formula is: where I final (x, y) is the temperature value of the fused image at the coordinate point (x, y). In this way, a comprehensive infrared thermal image is generated, effectively integrating information from different perspectives. For example, in the detection of large industrial heating equipment, the fused image can clearly present the overall surface temperature distribution of the equipment, avoiding the omission of local high-temperature points and providing a basis for subsequent accurate judgment of temperature abnormalities.

[0024] Then, in order to enhance image details and improve the distinguishability of temperature changes, the fused image is subjected to adaptive contrast adjustment, edge information optimization, and sharpening processing. This avoids problems such as unclear display of temperature changes in the image, low contrast, blurred edges, and missing details, which are not conducive to accurately judging temperature anomalies. First, for each local region of the image, calculate the contrast of this region, use the local contrast enhancement algorithm, and enhance the dynamic range of the image region through local calculation. In the range of the local region of the original image, the image after adaptive contrast enhancement where, I enhanced1 (x,y) is the image after contrast optimization, μ local (x,y) is the mean value of the local region, σ local (x,y) is the standard deviation of the local region, and C is the contrast enhancement factor, which is dynamically adjusted according to the contrast requirement of the image; then optimize the edge information in the infrared image to make the edges of regions with larger temperature differences clearer. Decompose the I enhanced1 (x,y) image into images I scale (x,y) at multiple scales, and then apply a high-pass filter to each scale to enhance the edge information of the image. The resulting local edge-enhanced image is: where α s is the weighting coefficient for each scale where σ s is the standard deviation at scale s. H s represents the operation at scale s N is the number of scales; finally, process the image through a non-linear sharpening method. The sharpening degree of the image can be adjusted through local gradient information: where, I final (x,y) is the final processed image, is the Laplacian operator of the image, representing local edge information, is the magnitude of the gradient of the image, is the maximum value of the gradient, used for non-linear adjustment.

[0025] Finally, to improve the overall detection accuracy, sensors are combined to achieve multi-dimensional data acquisition, and a weighted algorithm is used to perform weighted fusion on sensor data and image data. Existing detections mostly rely on a single data source, such as only using sensor data or image data, resulting in limited detection accuracy. The present invention combines sensors and image acquisition to achieve multi-dimensional data acquisition. During data fusion, through weighted average fusion, weights are reasonably allocated according to the characteristics of the data to give full play to the advantages of high-precision measurement by sensors and the overall distribution presented by images. First, ensure that the sensor and the image acquisition device work within the same time window, and perform standardization processing on them so that they can be jointly input into the fusion model, and then perform weighted average fusion: Wherein is the result after fusion, is the standardized data of the i-th sensor, is the standardization of the image data, γ i is the weight of the i-th sensor, γ I is the weight of the image data; finally, error evaluation is performed by comparing the weighted fusion result with the actual situation, and the weights of each sensor and image data are adjusted to ensure the final detection accuracy.

[0026] The above are only the preferred embodiments of the present invention, and do not limit the present invention in other forms. Any person skilled in the art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An abnormal surface temperature detection method for an electric heater based on AI recognition, characterized in that, Including the following steps: S1. First, collect images through infrared thermal imaging technology, and fuse images from different perspectives based on the weighted average method optimized by weights; S2. Then, adopt the adaptive contrast adjustment technology for the fused images and perform edge information optimization and sharpening processing to optimize image details and enhance the resolvability of temperature changes; S3. Finally, combine sensors to achieve multi-dimensional data collection, and perform weighted fusion on sensor data through a weighted algorithm, and improve the overall detection accuracy through the collaborative work of images and sensors; The specific implementation steps of adopting the adaptive contrast adjustment technology and performing edge information optimization and sharpening processing in step S2 are as follows: S21. First, for each local region of the image, calculate the contrast of this region. Using the local contrast enhancement algorithm, enhance the dynamic range of the image region through local calculation. For the original image within the local region range, the image after adaptive contrast enhancement where, I enhanced1 (x, y) is the image after contrast optimization, μ local (x, y) is the mean value of the local region, σ local (x, y) is the standard deviation of the local region, and C is the contrast enhancement factor, which is dynamically adjusted according to the contrast requirement of the image; S22. Next, optimize the edge information in the infrared image to make the edges in areas with larger temperature differences clearer. Decompose the I enhanced1 (x, y) image into images of multiple scales I scale (x, y), and then apply a high-pass filter to each scale to enhance the edge information of the image. The resulting locally edge-enhanced image is: where α s is the weighting coefficient for each scale, H s represents the operation at scale s, and N is the number of scales; S23. Finally, the image is processed by a non-linear sharpening method, and the sharpening degree of the image is adjusted by local gradient information: where I final (x, y) is the final processed image, is the Laplacian operator of the image, representing local edge information, is the gradient magnitude of the image, is the maximum gradient value for non-linear adjustment; The specific implementation of fusing images from different perspectives based on the weighted average method optimized by weights in step S1 is as follows: S11. For each of the images I1, I2, I3... I collected from each perspective n perform quality assessment, set a local weighting function, and calculate the local weighting value according to the quality of each perspective image. The weight ω i is a comprehensive score calculated based on the image sharpness C(I i ) and contrast D(I i ). The formula is: where n is the total number of images; S12. The weighted average method is adopted for image fusion, and weighted fusion is carried out according to the weights and qualities of each perspective, and finally a comprehensive infrared thermal image is generated, where each image I i The weight ω i of is determined according to the calculation of the previous step, and the weighted fusion formula is: where I final (x, y) is the temperature value of the fused image at the coordinate point (x, y).

2. The method for detecting abnormal surface temperature of an electric heater based on AI recognition according to claim 1, wherein, The specific implementation of the contrast enhancement factor C in the step S21 is as follows:

3. The method for detecting abnormal surface temperature of an electric heater based on AI recognition according to claim 1, wherein, The weighting coefficient α at each scale in the step S22 s has the specific form of: where σ s is the standard deviation at the scale s.

4. The surface temperature abnormal detection method of an electric heater based on AI recognition according to claim 3, characterized in that The operation H at scale s in step S22 s is a Gaussian filter at scale s:

5. The method for detecting abnormal surface temperature of an electric heater based on AI recognition according to claim 1, wherein In step S3, the sensor data is weighted and fused through a weighted algorithm. The implementation method of improving the overall detection accuracy by the collaborative work of the image and the sensor is as follows: First, ensure that the sensor and the image acquisition device work within the same time window, and perform standardization processing on them so that they can be jointly input into the fusion model, and then perform weighted average fusion: where is the result after fusion, is the standardized data of the i-th sensor, is the standardization of the image data, γ i is the weight of the i-th sensor, γ I is the weight of the image data; finally, error evaluation is performed by comparing the weighted fusion result with the actual situation, and the weights of each sensor and the image data are adjusted to ensure the final detection accuracy.

Images