Image post-processing optimization management method for electric power materials of intelligent receiving cabinet

Abstract

The invention provides an image post-processing optimization management method for electric power materials of an intelligent receiving cabinet, and belongs to the technical field of electric power material management. S2, collecting a coverage image; s3, feature extraction; s4, identifying and matching; s5, the whole process of inventory management and the like. According to the method, the multi-angle images of the materials are collected through a high-resolution camera, a multi-direction convolution frame is constructed by utilizing a space gradient energy kernel, and the extraction capability of geometric contours, surface textures and identification features is enhanced. The core technology adopts three-dimensional differential operation to analyze edge strength, curvature change and deformation features step by step, combines multi-scale texture analysis and a self-adaptive illumination compensation algorithm, eliminates environmental interference, generates a unique feature coupling factor, matches a pre-stored feature fingerprint database in real time through an elliptical envelope detection algorithm, accurately recognizes the types and the number of materials, and improves the accuracy of material identification. And closed-loop management from authority verification, accurate identification to intelligent early warning is formed.

Description

Technical Field

[0001] The present invention relates to the technical field of power material management, and in particular to an image post-processing optimization management method for power materials in an intelligent collection cabinet. Background Art

[0002] With the rapid development of the power industry, the demand for intelligent management of power materials (such as composite insulators, cable terminals, etc.) is becoming increasingly urgent. Traditional power material collection and management relies on manual registration and inventory, which has problems such as low efficiency, prone to errors, and poor real-time performance. In recent years, intelligent collection cabinets based on image recognition have been gradually introduced to optimize the process by automatically collecting and identifying material information.

[0003] For example, the patent with announcement number CN119006764A discloses a smart medicine cabinet drug management method and system, which uses a camera inside the cabinet to collect drug entry and exit orders and drug images, obtains entry and exit information through image recognition, and then verifies its compliance in combination with a real-time inventory table; reduces manual intervention through automated processes, and improves the speed and accuracy of drug entry and exit. However, in actual applications, power materials are of various types and complex structures (such as anti-slip threads, umbrella skirt structures, etc.), and traditional image feature extraction methods (such as single gradient or texture analysis) are difficult to fully capture their multi-dimensional features, resulting in difficulty in distinguishing similar materials, a high misrecognition rate, and insufficient image recognition accuracy.

[0004] Therefore, an image post-processing optimization management method for power materials in intelligent distribution cabinets was introduced. Summary of the invention

[0005] The technical problem to be solved by the present invention is to overcome the defects of the image feature extraction method in the prior art, such as the difficulty in distinguishing similar materials and the high misrecognition rate when facing the situation where there are many types of electric power materials with complex structures. The present invention proposes an image post-processing optimization management method for electric power materials in an intelligent dispensing cabinet.

[0006] In order to solve the above technical problems, the technical solution adopted by the present invention is: an image post-processing optimization management method for power materials in an intelligent collection cabinet, comprising the following steps:

[0007] S1: Pick-up and drop-off registration: the user clicks the function button on the touch screen of the smart collection cabinet and selects the function of collecting or storing power materials. After the facial recognition authority is determined, the door lock is opened and the operator registration is completed;

[0008] S2: Coverage image acquisition. After the user takes out or places electrical materials, the door lock is closed. During this process, the camera inside the cabinet takes all-round images of the objects taken out or placed and the electrical materials inside the cabinet, completes the original image data acquisition, and constructs a multi-directional convolution framework containing a spatial gradient energy kernel.

[0009] S3: Feature extraction: construct a multi-feature dynamic coupling model for electric power materials to extract the features of the original image, use the nonlinear multi-feature dynamic coupling model to calculate and process the original image, and calculate the characteristic coupling factor of the electric power materials based on the original image;

[0010] S4: Identification and matching: using the characteristic coupling factor to establish a characteristic fingerprint library of power materials, using the elliptical envelope detection method to match the calculated characteristic coupling factor with the characteristic fingerprint library in real time, and determine and identify the type and quantity of power materials to be taken in and out;

[0011] S5: The entire inventory management process updates the inventory status of materials in the smart collection cabinet based on the matching results and triggers an early warning signal.

[0012] Furthermore, in S3, the steps of calculating the characteristic coupling factor of the multi-characteristic dynamic coupling model are as follows:

[0013] Step 1: Perform three-dimensional differential operation on the original image data, and calculate the 1st to 3rd order partial derivatives along the X-axis direction to extract the geometric characteristics of the power materials;

[0014] Step 2: Calculate the multi-scale differential texture density of power materials, use 12 directional kernels to cover 360-degree space, and calculate the adaptive illumination compensation factor based on the mean and variance of each directional kernel based on the local image block;

[0015] Step 3: nonlinearly couple the differential result of step 1 with the output of step 2 to generate characteristic numerator terms;

[0016] Step 4: Calculate the cubic gradient integral of the second-order mixed partial derivative of the image as the noise suppression denominator, and divide the numerator by the denominator to generate the characteristic coupling factor of the power material.

[0017] Furthermore, the formula for calculating the characteristic coupling factor of the multi-characteristic dynamic coupling model is as follows:

[0018]

[0019] Where: I represents the input power material image, which is a two-dimensional matrix, in which each element I(i, j) corresponds to the grayscale value of the pixel, i represents the row coordinate, and j represents the column coordinate; k is the loop index used to traverse n feature dimensions, and the value of k is 1 to n; m represents the order used for derivation under each feature dimension k, and the value of m is 1 to 3; E θk Represents the convolution output of the k-th directional gradient energy kernel; represents the multi-scale differential texture density; Ψ k represents the adaptive illumination compensation factor; (W, H) represents the image size; Represents the image gradient energy kernel E θkPerform 1st to 3rd order partial derivative operations along the x direction; It means calculating the second-order mixed partial derivatives of the image in the x and y directions.

[0020] Further, in S4, the elliptical envelope detection method verifies the characteristic coupling factor of the power material in the following manner:

[0021] Let the power material feature fingerprint library Λ={Γ(I 1 ), Γ(I 2 ),…,Γ(I m )};

[0022] Real-time calculation of the material image Γ(Ix) to be identified;

[0023] Matching is done using the elliptical envelope detection method: Where ε is the dynamic tolerance threshold, which is a constant. When the characteristic coupling factor Γ(Ix) of the material image to be identified is equal to Γ(I k )satisfy The match is determined to be successful.

[0024] Furthermore, the calculation formula of the illumination compensation factor in step 2 is as follows:

[0025]

[0026] Among them, μ k represents the local image block mean of the power material image, Represents the local image block variance of the power material image.

[0027] Furthermore, the calculation method of the local image block mean of the power material image is as follows:

[0028]

[0029] The local image block variance of the power material image is calculated as follows:

[0030]

[0031] Among them, Ω k represents the local image block centered at pixel k; N is the total number of pixels in the local block.

[0032] Furthermore, when performing three-dimensional differential operations on the original image in step one: a 5×5 Sobel convolution kernel is used to calculate the first-order, second-order, and third-order partial derivatives along the x-axis direction, the first-order partial derivative extracts the edge strength of the power material, the second-order partial derivative extracts the curvature change of the power material, and the third-order partial derivative extracts the high-order deformation characteristics of the power material.

[0033] Furthermore, in S2, the key information of the original power material image collected by the camera in the distribution cabinet includes: geometric features including the shape, edge and curvature of the power material, texture features including the surface material of the power material, and semantic features including nameplate text and model identification. When the built-in camera in the distribution cabinet obtains the original image of the power material, the image resolution is at least 1920×1080 pixels or above, and the frame rate is not less than 15fps.

[0034] Furthermore, in S1, the power materials include but are not limited to: composite insulators, cable terminal heads, wiring hardware and lightning arrester components. When calculating the characteristic coupling factor, the surface characteristics of the power materials include at least two of anti-slip threads, umbrella skirt structures and metal flanges.

[0035] Furthermore, in S5, the specific implementation steps of the entire inventory management process are as follows:

[0036] S501, Inventory status update: adjust the inventory data of smart requisition cabinets in real time according to the matching results;

[0037] S502, intelligent early warning: When the inventory level is lower than the lower limit set by the intelligent collection cabinet or higher than the upper limit, an audible and visual alarm signal is triggered and pushed to the operation and maintenance management system through the MQTT protocol;

[0038] S503, dynamic adjustment of threshold: automatically optimize the inventory threshold of the smart collection cabinet based on historical collection data, with an adjustment cycle of 24 hours.

[0039] Compared with the prior art, the beneficial effects of the present invention include: a multi-directional convolution framework constructed through high-resolution image acquisition and spatial gradient energy kernel can accurately extract the geometric contours, surface textures and identification features of power materials, and combine three-dimensional differential operations to analyze edge intensity, curvature changes and deformation information step by step, and effectively eliminate environmental interference through adaptive lighting compensation technology, enhance feature stability in complex scenes, and use a multi-feature dynamic coupling model to generate unique feature factors by fusing multi-scale texture density and high-order differential features, and combine the elliptical envelope detection algorithm to achieve real-time high-precision matching, thereby ensuring the accuracy of identifying the types and quantities of materials during the management of power materials in the intelligent distribution cabinet. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] The disclosure of the present invention is described with reference to the accompanying drawings. It should be understood that the drawings are for illustrative purposes only and are not intended to limit the scope of protection of the present invention. In the accompanying drawings, the same reference numerals are used to refer to the same components. Among them:

[0041] Figure 1 The overall flow chart of the optimization management method proposed according to one embodiment of the present invention is schematically shown;

[0042] Figure 2 A flowchart of feature extraction implementation steps of an optimization management method proposed according to an embodiment of the present invention is schematically shown;

[0043] Figure 3 A flowchart of the steps for implementing the entire inventory management process of an optimization management method proposed according to an embodiment of the present invention is schematically shown. DETAILED DESCRIPTION

[0044] It is easy to understand that according to the technical solution of the present invention, without changing the essential spirit of the present invention, a person skilled in the art can propose a variety of interchangeable structural modes and implementation modes. Therefore, the following specific implementation modes and drawings are only exemplary descriptions of the technical solution of the present invention, and should not be regarded as the whole of the present invention or as a limitation or restriction to the technical solution of the present invention.

[0045] According to one embodiment of the present invention, Figure 1-Figure 3 An image post-processing optimization management method for power materials in an intelligent receiving cabinet includes the following implementation steps:

[0046] Step 1: User operation and permission verification

[0047] The user selects the "collect" or "deposit" function through the touch interface of the smart collection cabinet, and the system calls the face recognition module to verify the user's authority. After the verification is passed, the cabinet door is automatically unlocked and the operator's information is recorded.

[0048] Step 2: Image acquisition and data preprocessing

[0049] After the user completes the material collection and closes the cabinet door, the high-definition camera installed in the cabinet collects multi-angle images of the cabinet environment and power materials. The collected raw image data is processed by a multi-directional convolution framework, which integrates the spatial gradient energy kernel to enhance the recognizability of image features. The image resolution is not less than 1920×1080 pixels, the frame rate is ≥15fps, and the original power material image collected by the smart cabinet camera. It contains the following key information:

[0050] Geometric features: material shape, edge, curvature, etc. (such as the shed structure of the insulator)

[0051] Texture characteristics: surface material (such as brushed texture of metal flanges, granular texture of rubber parts)

[0052] Semantic features: key information such as nameplate text and model identification.

[0053] Step 3: Multi-dimensional feature extraction and coupling

[0054] Geometric feature extraction: Perform three-dimensional differential operations on the original image along the X-axis, use a 5×5 Sobel convolution kernel to calculate the 1st-3rd order partial derivatives respectively, and extract the edge strength (1st order), curvature change (2nd order) and high-order deformation features (3rd order) of the data.

[0055] Texture and lighting compensation: Based on 12 directional kernels covering 360 degrees of space, multi-scale differential texture density is calculated And through the mean of the local image block (μ k ) and variance Generate adaptive lighting compensation factors Eliminating the interference of light, the calculation method of the local image block mean of the power material image is as follows:

[0056]

[0057] The local image block variance of the power material image is calculated as follows:

[0058]

[0059] Among them, Ω k represents the local image block centered at pixel k; N is the total number of pixels in the local block.

[0060] Feature fusion: The differential operation result is nonlinearly coupled with the texture density to generate the feature numerator term, and combined with the cubic gradient integral of the second-order mixed partial derivative of the image (noise suppression denominator term), the feature coupling factor (Γ(I)) is finally output. The formula for calculating the feature coupling factor of the multi-feature dynamic coupling model is:

[0061]

[0062] Where: I represents the input power material image, which is a two-dimensional matrix, in which each element I(i, j) corresponds to the grayscale value of the pixel, i represents the row coordinate, and j represents the column coordinate; k is the loop index used to traverse n feature dimensions, and the value of k is 1 to n; m represents the order used for derivation under each feature dimension k, and the value of m is 1 to 3; E θk Represents the convolution output of the k-th directional gradient energy kernel; represents the multi-scale differential texture density; Ψ k represents the adaptive illumination compensation factor; (W, H) represents the image size; Represents the image gradient energy kernel E θk Perform 1st to 3rd order partial derivative operations along the x direction; It means calculating the second-order mixed partial derivatives of the image in the x and y directions.

[0063] By the mean (μ k ) and variance The dynamic ratio of , to achieve the following functions:

[0064] Light compensation factor design principle

[0065] Light Balance: Suppress overexposure in highlight areas and enhance details in shadow areas.

[0066] Noise suppression: In low contrast areas ( Small) to increase the effective signal weight, in high noise areas ( Large) to reduce interference effects.

[0067] Mechanism of action of light compensation factor

[0068] Highlight area (μ k big, Small), k The value increases significantly → through log(1+Ψ k ) compresses the dynamic range to prevent loss of details in metal reflective areas.

[0069] Shadow area (μ k Small, Large), k Decrease the value → retain the texture of dark areas (such as nameplate text) while suppressing noise amplification.

[0070] Complex texture area (μ k medium, Large), k Moderate value → Balances texture enhancement and noise suppression (such as anti-slip texture on cable terminal heads).

[0071] Step 4: Feature matching and material identification

[0072] Construction of feature fingerprint library: The feature coupling factors of historical power material images are stored in the database, denoted as Λ = {Γ(I1), Γ(I2), ..., Γ(Im)}.

[0073] Real-time matching: The characteristic coupling factor Γ(Ix) of the material image to be identified is matched using the elliptical envelope detection method to meet the conditions Where ε is the dynamic tolerance threshold, which is a constant. When the characteristic coupling factor Γ(Ix) of the material image to be identified is equal to Γ(I k )satisfy The match is determined to be successful and the type and quantity of materials are determined.

[0074] Step 5: Dynamic inventory management and early warning

[0075] Inventory update: adjust the inventory data of materials in the cabinet in real time according to the matching results.

[0076] Threshold warning: When the inventory level is lower than the preset lower limit or exceeds the upper limit, an audible and visual alarm is triggered, and warning information is pushed to the operation and maintenance system through the MQTT protocol.

[0077] Adaptive optimization: Based on historical collection data within 24 hours, the system automatically optimizes inventory thresholds to improve management efficiency.

[0078] In this embodiment, when the power materials in the intelligent collection cabinet are taken and placed based on user authority verification, the camera in the cabinet is used for multi-angle image acquisition, and the geometric contour, surface texture and identification information of the power materials are captured by a high-resolution camera. The multi-directional convolution framework is constructed through the spatial gradient energy kernel to enhance the feature extraction capability. In the core processing stage, three-dimensional differential operations are used to analyze the edge intensity, curvature change and deformation characteristics step by step, and the environmental interference is eliminated by combining multi-scale texture analysis and adaptive illumination compensation technology. Finally, characteristic factors are generated through nonlinear coupling to form the "characteristic fingerprint" of the materials. The matching link uses the dynamic elliptical envelope algorithm to dynamically compare the real-time characteristic factors with the pre-stored fingerprint library. The threshold judgment of the ratio and derivative deviation is used to accurately identify the type and quantity of materials. The inventory management module updates the data in real time according to the recognition results, and dynamically adjusts the inventory threshold. Through multi-dimensional feature fusion and dynamic optimization mechanism, the recognition accuracy and robustness under complex illumination and deformation conditions are significantly improved, and the real-time monitoring and automatic management of power material inventory are realized.

[0079] The technical scope of the present invention is not limited to the contents in the above description. Those skilled in the art can make various deformations and modifications to the above embodiments without departing from the technical idea of ​​the present invention, and these deformations and modifications should all fall within the protection scope of the present invention.

Claims

1. An image post-processing optimization management method for power materials in an intelligent cabinet, characterized in that: The following steps are involved: S1: Pick-up and drop-off registration: the user clicks the function button on the touch screen of the smart collection cabinet and selects the function of collecting or storing power materials. After the facial recognition authority is determined, the door lock is opened and the operator registration is completed; S2: Coverage image acquisition. After the user takes out or places electrical materials, the door lock is closed. During this process, the camera inside the cabinet takes all-round images of the objects taken out or placed and the electrical materials inside the cabinet, completes the original image data acquisition, and constructs a multi-directional convolution framework containing a spatial gradient energy kernel. S3: Feature extraction: construct a multi-feature dynamic coupling model for electric power materials to extract the features of the original image, use the nonlinear multi-feature dynamic coupling model to calculate and process the original image, and calculate the characteristic coupling factor of the electric power materials based on the original image; S4: Identification and matching: using the characteristic coupling factor to establish a characteristic fingerprint library of power materials, using the elliptical envelope detection method to match the calculated characteristic coupling factor with the characteristic fingerprint library in real time, and determine and identify the type and quantity of power materials to be taken in and out; S5: The entire inventory management process updates the inventory status of materials in the smart collection cabinet based on the matching results and triggers an early warning signal.

2. The image post-processing optimization management method for power materials in the intelligent cabinet according to claim 1 is characterized in that: In S3, the steps of calculating the characteristic coupling factor of the multi-characteristic dynamic coupling model are as follows: Step 1: Perform three-dimensional differential operation on the original image data, and calculate the 1st to 3rd order partial derivatives along the X-axis direction to extract the geometric characteristics of the power materials; Step 2: Calculate the multi-scale differential texture density of power materials, use 12 directional kernels to cover 360-degree space, and calculate the adaptive illumination compensation factor based on the mean and variance of each directional kernel based on the local image block; Step 3: nonlinearly couple the differential result of step 1 with the output of step 2 to generate characteristic numerator terms; Step 4: Calculate the cubic gradient integral of the second-order mixed partial derivative of the image as the noise suppression denominator, and divide the numerator by the denominator to generate the characteristic coupling factor of the power material.

3. The image post-processing optimization management method for the intelligent power supply cabinet according to claim 2 is characterized in that: The formula for calculating the characteristic coupling factor of the multi-characteristic dynamic coupling model is as follows: Where: I represents the input power material image, which is a two-dimensional matrix, in which each element I(i, j) corresponds to the grayscale value of the pixel, i represents the row coordinate, and j represents the column coordinate; k is the loop index used to traverse n feature dimensions, and the value of k is 1 to n; m represents the order used for derivation under each feature dimension k, and the value of m is 1 to 3; E θk Represents the convolution output of the k-th directional gradient energy kernel; represents the multi-scale differential texture density; Ψ k represents the adaptive illumination compensation factor; (W, H) represents the image size; Represents the image gradient energy kernel E θk Perform 1st to 3rd order partial derivative operations along the x direction; It means calculating the second-order mixed partial derivatives of the image in the x and y directions.

4. The image post-processing optimization management method for power materials in the intelligent collection cabinet according to claim 3 is characterized in that: In S4, the elliptical envelope detection method verifies the characteristic coupling factor of the power material in the following manner: Let the power material feature fingerprint library Λ = {Γ(I1), Γ(I2), ..., Γ(I m )}; Real-time calculation of the material image Γ(Ix) to be identified; Matching is done using the elliptical envelope detection method: Where ε is the dynamic tolerance threshold, which is a constant. When the characteristic coupling factor Γ(Ix) of the material image to be identified is equal to Γ(I k )satisfy The match is determined to be successful.

5. The image post-processing optimization management method for the intelligent power supply cabinet according to claim 3 is characterized in that: The calculation formula of the illumination compensation factor in step 2 is as follows: Among them, μ k represents the local image block mean of the power material image, Represents the local image block variance of the power material image.

6. The image post-processing optimization management method for power materials in the intelligent collection cabinet according to claim 5 is characterized in that: The calculation method of the local image block mean of the power material image is as follows: The local image block variance of the power material image is calculated as follows: Among them, Ω k represents the local image block centered at pixel k; N is the total number of pixels in the local block.

7. The image post-processing optimization management method for the intelligent power supply cabinet according to claim 2 is characterized in that: When performing three-dimensional differential operations on the original image in the step 1: a 5×5 Sobel convolution kernel is used to calculate the first-order, second-order, and third-order partial derivatives along the x-axis direction, the first-order partial derivative extracts the edge strength of the power material, the second-order partial derivative extracts the curvature change of the power material, and the third-order partial derivative extracts the high-order deformation characteristics of the power material.

8. The image post-processing optimization management method for power materials in the intelligent collection cabinet according to claim 1 is characterized in that: In S2, the key information of the original power material image collected by the camera in the distribution cabinet includes: geometric features including the shape, edge and curvature of the power material, texture features including the surface material of the power material, and semantic features including the nameplate text and model identification. When the built-in camera in the distribution cabinet obtains the original image of the power material, the image resolution is at least 1920×1080 pixels or above, and the frame rate is not less than 15fps.

9. The image post-processing optimization management method for the intelligent power supply cabinet according to claim 1 is characterized in that: In S1, the power materials include but are not limited to: composite insulators, cable terminal heads, wiring hardware and lightning arrester components. When calculating the characteristic coupling factor, the surface features of the power materials include at least two of anti-slip threads, umbrella skirt structures and metal flanges.

10. The image post-processing optimization management method for power materials in the intelligent collection cabinet according to claim 1 is characterized in that: In S5, the specific implementation steps of the entire inventory management process are as follows: S501, Inventory status update: adjust the inventory data of smart requisition cabinets in real time according to the matching results; S502, intelligent early warning: When the inventory level is lower than the lower limit set by the intelligent collection cabinet or higher than the upper limit, an audible and visual alarm signal is triggered and pushed to the operation and maintenance management system through the MQTT protocol; S503, dynamic adjustment of threshold: automatically optimize the inventory threshold of the smart collection cabinet based on historical collection data, with an adjustment cycle of 24 hours.

Images