Discharge detection method and system for adaptive registration of ultraviolet and visible light images

Abstract

The invention discloses a discharge detection method and system for adaptive registration of ultraviolet and visible light images, and belongs to the technical field of image processing. An ultraviolet image and a visible light image of a discharge area of power equipment are respectively captured through an imaging module; motion vectors of a visible light image sequence are calculated according to optical flow motion estimation in a motion scene through a registration module and mapped to an ultraviolet image for displacement compensation, translation compensation is determined by adopting histogram matching in a static scene, and the motion vectors of the visible light image sequence are calculated according to the compensated image. Zooming the ultraviolet image according to the zoom scale calculated by the focal length and the central point of the ultraviolet image, and zooming the effective registration area of the cut ultraviolet image to a target size; and the electric detection module carries out electric discharge detection on the discharge point area according to the registered image. According to the method, the robustness is improved, boundary offset and nonlinear distortion are avoided, edge linear fitting is realized while accurate alignment of the center points is ensured, and the calculation efficiency is remarkably improved in cooperation with integer operation.

Description

Technical Field

[0001] The present invention belongs to the technical field of image processing, and in particular relates to a discharge detection method and system for adaptive registration of ultraviolet and visible light images. Background Art

[0002] In fields such as power equipment testing and high-voltage discharge monitoring, UV imaging technology is widely used to detect defects such as corona discharge and partial discharge. UV instruments capture the UV light signal from the discharge area to generate a corresponding UV image, while visible light cameras simultaneously capture visible light images of the same area. However, in practical applications, there is a spatial misalignment between the UV and visible light images. The main reasons include:

[0003] 1) Optical system differences: The focal length, distortion, and installation position of UV lenses and visible light lenses are different, resulting in incomplete correspondence between the positions of the same target in the two images.

[0004] 2) Dynamic tracking problem: When the detection target (such as the discharge point on a high-voltage device) moves, the visible light image can be tracked in real time, but the UV image may not move synchronously due to different imaging mechanisms, resulting in the inability to accurately match the UV-visible light image.

[0005] Currently, common image registration methods include feature point matching (such as SIFT and ORB) and affine transformation (such as translation, rotation, and scaling). However, these methods have the following problems in UV-visible image alignment:

[0006] 1) Failure of feature point matching: Ultraviolet images usually have low signal-to-noise ratio and little texture information, which makes feature point extraction difficult and the matching accuracy low.

[0007] 2) Affine transformation is computationally complex: Traditional methods require the calculation of a global transformation matrix (such as a homography matrix), which is computationally intensive and difficult to adapt to dynamic adjustments at different focal lengths.

[0008] 3) Boundary alignment error: During the zoom process, the scaling ratios of the UV and visible light images may be inconsistent, resulting in center point alignment but boundary offset, affecting detection accuracy. Summary of the Invention

[0009] To address the deficiencies of the prior art and achieve the goal of precise alignment of ultraviolet images and visible light images, the present invention adopts the following technical solutions:

[0010] A method for adaptive registration of ultraviolet and visible light images comprises the following steps:

[0011] Step S1: Initial registration of the UV image and the visible light image;

[0012] For moving scenes, the motion vector of the visible light image sequence is calculated through optical flow motion estimation and mapped to the UV image for displacement compensation; for static scenes, histogram matching is used to determine the translation compensation;

[0013] Step S2: adaptive zooming of focal length;

[0014] Calculating the scaling ratio through the focal length, and scaling the ultraviolet image according to the scaling ratio and the center point of the ultraviolet image;

[0015] Step S3: image scaling;

[0016] Scale the cropped UV image effective registration area to the target size.

[0017] Furthermore, the optical flow motion estimation in step S1 obtains the optimal displacement for alignment compensation of the ultraviolet image by minimizing the pixel brightness difference within a local window between adjacent frames of visible light.

[0018] Furthermore, by setting a window area, the pixel intensity value of the visible light image in the window area at the current moment and spatial position, as well as the pixel intensity value of the corresponding spatial position after the ultraviolet image displacement compensation at the next moment are calculated, and the square of the pixel intensity difference between the current moment and the next moment after displacement compensation is calculated, and the squares of all pixel intensity differences in the window area are summed to obtain the ultraviolet image displacement compensation when the sum of the squares of the pixel intensity differences is minimized.

[0019] The formula for optical flow motion estimation is as follows:

[0020]

[0021] in, represents the two-dimensional displacement vector to be solved, and They represent the displacement of the ultraviolet image that needs to be compensated in the horizontal coordinate x and vertical coordinate y directions respectively. Arg min represents the optimal displacement parameter that minimizes the objective function. The optimal displacement can be solved by minimizing the error function. Indicates the sum of all pixels in the specified window area W. represents the optical flow calculation window, represents a visible light image, Represents the pixel intensity value of the visible light image at time t and spatial position (x, y), Indicates the next moment Time displacement The pixel intensity value at the corresponding position.

[0022] Furthermore, the histogram matching in step S1 calculates the normalized cross-correlation value between the visible light image and the corresponding ultraviolet light image, uses the offset corresponding to the maximum normalized cross-correlation value as the alignment parameter, translates the ultraviolet image, and aligns the visible light image.

[0023] Furthermore, the normalized cross-correlation value is obtained by multiplying the pixel value of the visible light image with the pixel value of the corresponding position of the ultraviolet image after the offset, and summing all the products in the overlapping area of ​​the visible light image and the ultraviolet image as the numerator; summing the squares of all visible light image pixel values ​​in the image overlapping area, summing the squares of all ultraviolet image pixel values ​​at the corresponding position after the offset, and then multiplying the sum of the squares of all visible light image pixel values ​​with the sum of the squares of all ultraviolet image pixel values ​​at the corresponding position after the offset, and opening them as the denominator.

[0024] The formula for the normalized cross-correlation value is as follows:

[0025]

[0026] in, Indicates displacement The normalized cross-correlation value under , is used to measure the similarity between two images under a given displacement, Represent the displacement to be evaluated, represents the horizontal displacement of x, represents the displacement of y in the vertical direction, Indicates the summation of all pixel coordinates (x, y), and the summation range is all valid pixel points in the overlapping area of ​​the image. Represents the pixel value of the visible light reference image at coordinate (x, y), Represents the square of the pixel value of the visible light image at the coordinate (x, y), is the pixel value of the corresponding position of the ultraviolet target image after the offset (dx, dy), indicating the alignment of the ultraviolet image with the visible light image after the offset (dx, dy). Represents the square of the pixel value at the corresponding position of the ultraviolet image after the offset (dx, dy).

[0027] Furthermore, in the ultraviolet-visible light image registration process, when the optical system is zoomed, the scaling ratio of the ultraviolet and visible light images will change with the focal length. For this reason, in the step S2 of the present invention, the linear relationship between the focal length and the horizontal and / or vertical scaling factor is experimentally calibrated, and the position of the cropping window is dynamically adjusted according to the scaling factor and the coordinates of the center point of the ultraviolet image matching area to obtain the effective matching area of ​​the ultraviolet image. The linear relationship calibrated through the experiment is used to achieve precise alignment of the center point and linear fitting of the boundary, which effectively solves the boundary offset and nonlinear distortion problems caused by the traditional fixed ratio method and can adapt to the matching requirements under different focal lengths.

[0028] Furthermore, in step S3, a bilinear interpolation algorithm is used to allocate the nearest horizontal and vertical integer coordinates of the pixel points within the effective matching area of ​​the ultraviolet image based on weights to adjust the pixel values ​​of the pixel points so that the effective matching area of ​​the ultraviolet image is scaled to the target size to obtain the target image.

[0029] Furthermore, the bilinear interpolation algorithm calculates the nearest reference pixel integer coordinates of the pixel point's coordinates in the horizontal and / or vertical directions based on the coordinates of the pixel point in the effective matching area of ​​the ultraviolet image, and obtains the horizontal and / or vertical difference weights based on the ratio of the distance between the pixel point's coordinates and the reference pixel integer coordinates in the horizontal and / or vertical directions. The pixel value of the pixel point's coordinates is obtained based on the sum of the products of the reference pixel integer coordinates and the corresponding difference weights, so as to generate a scaled target image of the effective matching area of ​​the ultraviolet image.

[0030] The pixel value of the target image at coordinate (x, y) is as follows:

[0031]

[0032] Among them, dst represents the target image after scaling, Represents the pixel value of the target image at the coordinate (x, y), src represents the effective registration area of ​​the original UV image, 、 、 、 They represent the four nearest pixels to be calculated in the target image of the pixel in the UV image registration area, u and v are the interpolation weights in the horizontal and vertical directions (0 ≤ u, v ≤ 1), which are used to represent the distance between the target point and and The horizontal distance ratio, and are the integer coordinates of the reference pixel in the source image (via calculate).

[0033] A discharge detection method using adaptive registration of ultraviolet and visible light images obtains ultraviolet and visible light images of a discharge point area of ​​an electric power device. The ultraviolet and visible light images of the discharge point area of ​​the electric power device are registered according to the adaptive registration method for ultraviolet and visible light images for discharge detection.

[0034] A discharge detection system with adaptive registration of ultraviolet and visible light images, comprising an imaging module, a registration module and a discharge detection module;

[0035] The imaging module includes an ultraviolet imaging module and a visible light imaging module, which are respectively used to capture ultraviolet images and visible light images of the discharge area of ​​the power equipment;

[0036] The registration module registers the ultraviolet and visible light images of the power equipment discharge point area according to the ultraviolet and visible light image adaptive registration method;

[0037] The discharge detection module performs discharge detection on the discharge point area according to the registered image.

[0038] The advantages and beneficial effects of the present invention are:

[0039] The present invention provides a discharge detection method and system for adaptive registration of ultraviolet and visible light images. By dynamically adjusting the cropping area through focal length parameters and combining an improved interpolation method, ultraviolet images and visible light images are precisely aligned through step-by-step optimization, achieving high-precision image matching. The method is particularly suitable for application scenarios such as alignment of ultraviolet and visible light images in discharge detection of power equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] Figure 1 This is the effect diagram of the simultaneous comparison of ultraviolet and visible light in the traditional method.

[0041] Figure 2 is a flow chart of a method in an embodiment of the present invention.

[0042] Figure 3 This is a synchronous comparison effect diagram of ultraviolet and visible light after initial alignment in an embodiment of the present invention.

[0043] Figure 4 This is a synchronous comparison effect diagram of ultraviolet and visible light after adaptive focal length scaling in an embodiment of the present invention.

[0044] Figure 5 Schematic diagram of the structure of the system in the embodiment of the present invention. DETAILED DESCRIPTION

[0045] The following describes the specific embodiments of the present invention in detail with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only used to illustrate and explain the present invention and are not intended to limit the present invention.

[0046] like Figure 1 As shown in Figure 2, there is a spatial position deviation between the UV image and the visible light image in actual application, which makes the UV-visible light image registration impossible. In order to solve the problem that the traditional feature matching method (such as SIFT / SURF) is computationally complex and has a high registration failure rate due to the scarcity of UV image features, this paper proposes an adaptive registration method for UV and visible light images, such as Figure 2 As shown,

[0047] Step S1: Initial registration of UV image and visible light image; fast coarse alignment by fusing optical flow motion estimation and histogram cross-correlation matching.

[0048] First, the motion vector of the visible light image sequence is calculated using the Lucas-Kanade optical flow method and mapped to the ultraviolet image for displacement compensation. For static scenes, a histogram matching algorithm based on normalized cross correlation (NCC) is used to determine the optimal translation parameters, thereby achieving fast and robust initial alignment, laying the foundation for subsequent fine alignment.

[0049] (1) Displacement estimation based on optical flow

[0050] If there is motion in the scene (such as camera movement or target displacement), the Lucas-Kanade sparse optical flow method is used to calculate the pixel motion vector of the visible light image and map it to the ultraviolet image:

[0051]

[0052] in, represents the two-dimensional displacement vector to be solved, and They represent the displacement of the UV image that needs to be compensated in the x and y directions respectively. Argmin represents the optimal displacement parameter that minimizes the objective function. The optimal displacement can be solved by minimizing the error function. Indicates the sum of all pixels in the specified window area W. is the optical flow calculation window (such as 5x5 pixels), represents a visible light image, Represents the pixel intensity value of the visible light image at time t and spatial position (x, y), Indicates the next moment Time displacement The pixel intensity value at the corresponding position.

[0053] The core idea of ​​this equation is to solve the optimal displacement vector ( ). In the UV and visible light registration, the solved displacement vector will be used for the initial alignment compensation of the UV image.

[0054] (2) Translation compensation based on histogram matching

[0055] For static scenes, the normalized cross correlation (NCC) between the visible and UV images is calculated to determine the optimal translation offset (Δx, Δy):

[0056]

[0057] in, Indicates displacement The normalized cross-correlation value under , the value range is: [-1, 1], 1 means perfect match, used to measure the similarity of two images under a given displacement, Represent the displacement to be evaluated, where Refers to the x-direction (horizontal) displacement (unit: pixel), Refers to the y-direction (vertical) displacement (unit: pixel), usually searched within a certain range (such as ±20 pixels). Indicates the summation of all pixel coordinates (x, y), and the summation range is all valid pixels in the overlapping area of ​​the image. Represents the pixel value of the visible light reference image at coordinate (x, y), usually a grayscale value (0-255), and It represents the square of the pixel value of the visible light image at (x, y), which is used as the denominator of the normalization calculation. Indicates the pixel value of the corresponding position of the UV target image after the offset (dx, dy), indicating the alignment of the UV image with the visible light image after the UV image is translated (dx, dy). Represents the square of the pixel value of the UV image at the offset position, also used for normalization calculation.

[0058] Take the offset corresponding to the maximum NCC value as the coarse alignment parameter:

[0059]

[0060] in, The optimal displacement vectors are represented by (3.2, -1.5) and (3.2, -1.5), respectively. For example, if (3.2, -1.5) is obtained, the optimal registration requires a right shift of 3.2 pixels and an upward shift of 1.5 pixels. It is a mathematical operator used to find the independent variable that maximizes the objective function NCC. .

[0061] Output: Rough alignment result of the UV image after translation transformation:

[0062]

[0063] in represents the original UV image, represents the image after translation transformation, represents the target image coordinates, Indicates the translation displacement.

[0064] Table 1 Performance comparison of UV-visible image initial registration algorithms

[0065] index Traditional SIFT method Fusion algorithm of the present invention Improvement Feature point extraction success rate 32% No feature points required ∞ Single frame processing time (ms) 120 18 85% Translation error (pixels) ±3.5 ±0.8 77% Robustness to lighting changes Difference excellent - Dynamic scene adaptability Not supported support -

[0066] Table 1 quantitatively compares the performance differences between the traditional SIFT method and the proposed optical flow-NCC fusion algorithm. The data shows that the proposed method significantly outperforms the traditional method in terms of feature point dependence, processing speed, and registration accuracy:

[0067] Feature point extraction: The success rate of traditional methods is only 32%, while this method does not require feature points and directly solves the problem of scarce features in ultraviolet images;

[0068] Processing speed: Single-frame processing time is reduced from 120ms to 18ms, with an efficiency improvement of 85%, meeting real-time requirements.

[0069] Registration accuracy: Translation error is optimized from ±3.5 pixels to ±0.8 pixels, achieving sub-pixel alignment;

[0070] Adaptability: New support for dynamic scenes and excellent robustness to lighting changes.

[0071] The data verified the comprehensive advantages of the fusion strategy in speed, accuracy and scene coverage.

[0072] The core idea of ​​this equation is to search for the optimal translation parameter within a preset displacement range by maximizing the normalized cross-correlation value (NCC) between the visible light and ultraviolet images. In the UV-visible registration of static scenes, the displacement vector solved will achieve the pixel-level coarse alignment of the two images, such as Figure 3 As shown, this lays the foundation for subsequent fine registration. The advantage of this method is that it is robust to illumination changes and does not rely on feature point extraction. It is particularly suitable for UV image matching scenarios with sparse features.

[0073] Step S2: adaptive zooming of focal length;

[0074] During UV-visible image registration, when the optical system zooms, Figure 3 The scaling ratio of mid-UV and visible light images changes with focal length. The traditional method of fixing the scaling factor will result in:

[0075] 1) Center points are aligned but boundaries are offset: the edge areas of the UV image cannot be matched with the visible light image;

[0076] 2) Non-linear distortion: During zooming, the zoom ratio and focal length are not in a simple linear relationship.

[0077] Therefore, based on experimental data, the present invention proposes a dynamic zoom method with adaptive focal length, which achieves precise matching through the following steps:

[0078] Step S2.1: scaling parameter calculation;

[0079] According to the experimental calibration data, the linear relationship between the focal length x and the scaling coefficients a and b is established:

[0080]

[0081] Where x represents the focal length (or magnification) of the current optical system, a represents the scaling factor in the x-direction (horizontal scaling ratio), and b represents the scaling factor in the y-direction (vertical scaling ratio).

[0082] Step S2.2: dynamic cropping window calculation;

[0083] Determine the valid matching area of ​​the UV image based on the scaling factor:

[0084]

[0085] Among them, (x1, y1) represents the coordinates of the upper left corner of the cropping window, (x2, y2) represents the coordinates of the lower right corner of the cropping window, and (360, 288) represents the coordinates of the center point of the UV image (assuming the original image size is 720×576).

[0086] The algorithm achieves precise center point alignment and boundary linear fitting through experimentally calibrated linear relationships, effectively solving the boundary offset and nonlinear distortion problems caused by traditional fixed-ratio methods, and can adapt to matching requirements at different focal lengths.

[0087] Step S3: Image scaling

[0088] Use the improved bilinear interpolation algorithm to scale the effective registration area of ​​the cropped UV image to the target size (e.g. 720×576):

[0089]

[0090] Among them, dst represents the target image after scaling, Represents the pixel value of the target image at the coordinate (x, y), src represents the effective registration area of ​​the original UV image, 、 、 、 They represent the four nearest pixels to be calculated in the target image of the pixel in the UV image registration area, u and v are the interpolation weights in the horizontal and vertical directions (0 ≤ u, v ≤ 1), which are used to represent the distance between the target point and and The horizontal distance ratio, and are the integer coordinates of the reference pixel in the source image (via calculate).

[0091] In the embodiment of the present invention, an ethanol burning flame is used to simulate the ultraviolet radiation characteristics of the discharge of power equipment to construct a standardized test environment. Experimental data show that the ultraviolet radiation intensity of the ethanol flame produced by the alcohol lamp in the 300-400nm band is highly consistent with the ultraviolet spectrum characteristics of typical corona discharge, and its relative error is controlled within the range of ±5%, which can be used as a high-fidelity physical simulation source. Figure 1 As shown, through the image processing algorithm of the present invention, the dynamic response characteristics of the simulated flame in the ultraviolet imaging system can achieve synchronous displacement tracking with the visible light image, providing a reliable experimental basis for the subsequent verification of the registration algorithm.

[0092] Table 2 Performance comparison of focal length adaptive scaling algorithms

[0093] index Traditional fixed ratio method Dynamic scaling algorithm of the present invention Improvement Edge registration error (pixels) 4.8±2.1 0.9±0.3 81% Zoom response delay (ms) 45 12 73% Nonlinear distortion coefficient 0.18 0.03 83% Computing resource usage (MIPS) 320 150 53% Dynamic zoom support Not supported support -

[0094] As shown in Table 2, the dynamic scaling algorithm of the present invention achieves significant improvements over traditional methods: edge registration error is reduced by 81% to 0.9 pixels, response delay is shortened by 73% to 12ms, nonlinear distortion is improved by 83% to 0.03, computing resources are reduced by 53%, and dynamic zoom support is newly added.

[0095] like Figure 4 As shown, the present invention dynamically adjusts the zoom factor through the focal length parameter x, adopts a dynamic cropping window and an optimized bilinear interpolation algorithm, achieves edge linear fitting while ensuring precise alignment of the center point, and significantly improves computational efficiency by combining integer operations. It effectively solves the nonlinear matching problem of ultraviolet-visible light images during zooming, and has the comprehensive advantages of strong adaptability, high precision, and good real-time performance.

[0096] A discharge detection method using adaptive registration of ultraviolet and visible light images obtains ultraviolet and visible light images of a discharge point area of ​​an electric power device. The ultraviolet and visible light images of the discharge point area of ​​the electric power device are registered according to the adaptive registration method for ultraviolet and visible light images for discharge detection.

[0097] like Figure 5 As shown, a discharge detection system with adaptive registration of ultraviolet and visible light images includes an imaging module, a registration module and a discharge detection module; the imaging module includes an ultraviolet imaging module and a visible light imaging module, which are respectively used to capture ultraviolet images and visible light images of the discharge point area of ​​the power equipment; the registration module aligns the ultraviolet and visible light images of the discharge point area of ​​the power equipment according to the described ultraviolet and visible light image adaptive registration method; the discharge detection module performs discharge detection on the discharge point area based on the registered images.

[0098] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for adaptive registration of ultraviolet and visible light images, characterized in that The steps include: Step S1: Initial registration of the UV image and the visible light image; For moving scenes, the motion vector of the visible light image sequence is calculated through optical flow motion estimation and mapped to the UV image for displacement compensation; for static scenes, histogram matching is used to determine the translation compensation; Step S2: adaptive zooming of focal length; Calculating the scaling ratio through the focal length, and scaling the ultraviolet image according to the scaling ratio and the center point of the ultraviolet image; Step S3: image scaling; Scale the cropped UV image effective registration area to the target size.

2. The method for adaptive registration of ultraviolet and visible light images according to claim 1, characterized in that: The optical flow motion estimation in step S1 obtains the optimal displacement for alignment compensation of the ultraviolet image by minimizing the pixel brightness difference within a local window between adjacent frames of visible light.

3. The method for adaptive registration of ultraviolet and visible light images according to claim 2, wherein: By setting a window area, calculating the pixel intensity value of the visible light image at the current moment and spatial position in the window area, as well as the pixel intensity value of the corresponding spatial position after the ultraviolet image displacement compensation at the next moment, and calculating the square of the pixel intensity difference between the current moment and the next moment after displacement compensation, the squares of all pixel intensity differences in the window area are summed to obtain the ultraviolet image displacement compensation when the sum of the squares of the pixel intensity differences is minimized.

4. The method for adaptive registration of ultraviolet and visible light images according to claim 1, wherein: The histogram matching in step S1 calculates the normalized cross-correlation value between the visible light image and the corresponding ultraviolet light image, uses the offset corresponding to the maximum normalized cross-correlation value as the alignment parameter, translates the ultraviolet image, and aligns the visible light image.

5. The method for adaptive registration of ultraviolet and visible light images according to claim 4, characterized in that: The normalized cross-correlation value is obtained by multiplying the pixel value of the visible light image with the pixel value of the corresponding position of the ultraviolet image after the shift, and summing all the products in the overlapping area of ​​the visible light image and the ultraviolet image as the numerator; In the overlapping area of ​​the images, the squares of all visible light image pixel values ​​are summed, and the squares of all ultraviolet image pixel values ​​at the corresponding positions after the offset are summed. Then, the sum of the squares of all visible light image pixel values ​​is multiplied by the sum of the squares of all ultraviolet image pixel values ​​at the corresponding positions after the offset, and the result is opened as the denominator.

6. The method for adaptive registration of ultraviolet and visible light images according to claim 1, characterized in that: In step S2, the linear relationship between the focal length and the horizontal and / or vertical scaling factor is experimentally calibrated, and the position of the cropping window is dynamically adjusted according to the scaling factor and the coordinates of the center point of the ultraviolet image matching area to obtain the effective matching area of ​​the ultraviolet image.

7. The method for adaptive registration of ultraviolet and visible light images according to claim 6, characterized in that: In step S3, a bilinear interpolation algorithm is used to allocate the nearest horizontal and vertical integer coordinates of the pixel points within the effective matching area of ​​the ultraviolet image based on the weights to adjust the pixel values ​​of the pixel points so that the effective matching area of ​​the ultraviolet image is scaled to the target size to obtain the target image.

8. The method for adaptive registration of ultraviolet and visible light images according to claim 7, characterized in that: The bilinear interpolation algorithm calculates the horizontal and / or vertical nearest reference pixel integer coordinates of the pixel coordinates in the effective matching area of ​​the ultraviolet image, and obtains horizontal and / or vertical difference weights based on the ratio of the distance between the pixel coordinates and the reference pixel integer coordinates in the horizontal and / or vertical directions. The pixel value of the pixel coordinates is obtained based on the sum of the products of the reference pixel integer coordinates and the corresponding difference weights to generate a scaled target image of the effective matching area of ​​the ultraviolet image.

9. A discharge detection method using adaptive registration of ultraviolet and visible light images, characterized by: Acquire ultraviolet images and visible light images of the discharge point area of ​​the power equipment, and align the ultraviolet and visible light images of the discharge point area of ​​the power equipment according to the adaptive registration method of ultraviolet and visible light images according to any one of claims 1 to 8 for discharge detection.

10. A discharge detection system with adaptive registration of ultraviolet and visible light images, comprising an imaging module, a registration module, and a discharge detection module, characterized in that: The imaging module includes an ultraviolet imaging module and a visible light imaging module, which are respectively used to capture ultraviolet images and visible light images of the discharge area of ​​the power equipment; The registration module registers the ultraviolet and visible light images of the power equipment placement area according to the method for adaptive registration of ultraviolet and visible light images according to any one of claims 1 to 8; The discharge detection module performs discharge detection on the discharge point area according to the registered image.

Images