A method and system for identifying Zisha teapots based on image processing

Abstract

This invention relates to the field of image processing technology, and more particularly to a method and system for identifying Zisha (purple clay) teapots based on image processing. The system includes: a data acquisition module for acquiring images of the Zisha teapot and environmental data; a preprocessing module for color normalization of the Zisha teapot image to obtain a corrected image; a region division module for dividing the corrected image into regions based on the horizontal and vertical gradients of each pixel to obtain Zisha teapot regions; a color feature analysis module for constructing color coefficients for the Zisha teapot regions; a texture feature analysis module for analyzing the texture features of the Zisha teapot regions; a gloss feature analysis module for analyzing the gloss features of the Zisha teapot regions; and a judgment module for analyzing the type of clay used in the Zisha teapot. This invention effectively improves the accuracy of Zisha teapot clay type analysis and increases the efficiency of Zisha teapot identification.

Description

Technical Field

[0001] This invention relates to the field of image processing technology, and in particular to a method and system for identifying Zisha teapots based on image processing. Background Technology

[0002] Zisha ware is a treasure of traditional Chinese arts and crafts. Its unique clay composition and production process endow each piece with unique artistic value. The main types of clay used in Zisha ware include purple clay, red clay, and green clay, each exhibiting significant differences in color, texture, and luster. Therefore, accurately identifying the type of clay used in Zisha ware is crucial for determining its craftsmanship, collectible value, and authenticity.

[0003] Chinese Patent Publication No. CN110208303A discloses a method for detecting Zisha (purple clay) ware, comprising the following steps: testing a standard raw material using energy-dispersive X-ray fluorescence spectrometry (EDXRF) to obtain the fluorescence intensity of each analyte in the standard raw material, and calculating the total fluorescence intensity of all analytes and the relative proportion of each fluorescence intensity in the total fluorescence intensity; testing a sample to be tested using EDXRF to obtain the fluorescence intensity of each analyte in the sample to be tested, and calculating the total fluorescence intensity of all analytes and the relative proportion of each fluorescence intensity in the total fluorescence intensity; comparing the relative proportion of each analyte in the standard raw material with the relative proportion of each analyte in the sample to be tested. This method for detecting Zisha ware can detect whether chemical raw materials have been added to the sample, and the testing process does not require damage to the sample. However, this method does not analyze the type of clay used in Zisha ware identification, resulting in low identification efficiency. Summary of the Invention

[0004] The purpose of this invention is to provide an image processing-based Zisha teapot identification system to solve at least one of the problems existing in the prior art.

[0005] To achieve the above objectives, the present invention adopts the following technical solution:

[0006] A Zisha teapot identification system based on image processing, comprising:

[0007] The data acquisition module is used to collect images of Zisha teapots and environmental data;

[0008] The preprocessing module is used to perform color normalization on the Zisha teapot image to obtain a corrected image;

[0009] The region segmentation module is used to segment the corrected image into regions based on the gradient of each pixel in the horizontal direction and the gradient in the vertical direction, so as to obtain the Zisha pottery region.

[0010] The color feature analysis module is used to extract color features from the HSV image of Zisha teapots and construct the color coefficients of the Zisha teapot area based on the extraction results.

[0011] The texture feature analysis module is used to analyze the texture state of the Zisha teapot area based on the contrast and gray-level entropy of the Zisha teapot area, and to analyze the texture features of the Zisha teapot area based on the analysis results.

[0012] The gloss feature analysis module is used to classify pixel types and analyze the gloss features of the Zisha teapot area based on the classification results.

[0013] The judgment module is used to analyze the type of Zisha clay based on the construction results of the color coefficient of the Zisha area, texture characteristics, gloss characteristics, and the collected ambient light intensity and humidity.

[0014] Furthermore, the region segmentation module calculates the gradient magnitude Pi of the i-th pixel in the corrected image based on the gradient Tix in the horizontal direction and the gradient Tiy in the vertical direction of the i-th pixel in the corrected image, and analyzes the relative position of the pixel based on the gradient magnitude Pi of the i-th pixel in the corrected image and the preset gradient magnitude threshold p0. The relative position of the pixel includes non-edge pixels and edge pixels.

[0015] The region division module uses the curve formed by connecting each edge pixel as the outline of the Zisha pottery, and the area within the outline of the Zisha pottery as the Zisha pottery area.

[0016] Furthermore, the color feature analysis module sets the hue of the HSV image of the Zisha pottery to Hu and the hue coefficient to SD;

[0017] The color feature analysis module sets the saturation of the HSV image of the Zisha pottery to BH and the saturation coefficient to BHD.

[0018] The color feature analysis module sets the color coefficient of Zisha pottery as Y, and sets Y = r1 × SD + r2 × BHD;

[0019] Where r1 is the hue weight and r2 is the saturation weight.

[0020] Furthermore, the texture feature analysis module includes a contrast analysis unit, which is used to analyze the contrast state of the Zisha teapot area based on the contrast d0 of the Zisha teapot area, and construct contrast coefficients based on the analysis results. The contrast coefficients include D1, D2 and D3.

[0021] The texture feature analysis module also includes a gray-level entropy analysis unit. The gray-level entropy analysis unit is used to analyze the gray-level entropy state of the Zisha pottery area based on the image gray-level entropy s0 and the preset gray-level entropy s1. The gray-level entropy state of the Zisha pottery area includes a normal state and an abnormal state. If the gray-level entropy state of the Zisha pottery area is an abnormal state, the gray-level entropy coefficient is set to S1. If the gray-level entropy state of the Zisha pottery area is a normal state, the gray-level entropy coefficient is set to S2.

[0022] Furthermore, the texture feature analysis module also includes a texture analysis unit, which is used to analyze the texture state of the Zisha pottery area based on the construction result of the contrast coefficient. The texture state of the Zisha pottery area includes a normal state and an abnormal state. If the texture state of the Zisha pottery area is an abnormal state, the texture feature of the Zisha pottery area is set to WT1. If the texture state of the Zisha pottery area is a normal state, the texture feature of the Zisha pottery area is set to WT2.

[0023] Furthermore, the gloss feature analysis module includes a type analysis unit, which is used to classify the pixel type according to the gray value Lk of the kth pixel in the Zisha pottery area and the preset gray value threshold YZ. The pixel type includes a first-class pixel and a second-class pixel.

[0024] The gloss feature analysis module also includes a gloss feature analysis unit, which is used to calculate the average gray value PG of the two types of pixels, and analyze the gloss coefficient GX of the Zisha teapot based on the average gray value PG of the two types of pixels and the average pixel value Pu of the Zisha teapot area.

[0025] The gloss feature analysis unit analyzes the gloss features of the Zisha teapot area based on the gloss coefficient GX of the Zisha teapot and the preset gloss coefficient GX0. If GX < GX0, the gloss feature analysis unit sets the gloss feature of the Zisha teapot area to GT1; otherwise, the gloss feature analysis unit sets the gloss feature of the Zisha teapot area to GT2.

[0026] Furthermore, the judgment module includes a judgment unit, which is used to analyze the type of Zisha clay based on the construction results of the color coefficient of the Zisha area, the texture characteristics of the Zisha area, and the gloss characteristics of the Zisha area, wherein:

[0027] If w1×color coefficient of Zisha area + w2×texture feature of Zisha area + w3×gloss feature of Zisha area ≤ u0, the judgment unit determines that the Zisha clay type is Qingni (green clay); otherwise, the judgment unit determines that the Zisha clay type is not Qingni (green clay).

[0028] Where w1 is the color weight, w2 is the texture weight, w3 is the gloss weight, and u0 is the preset judgment threshold.

[0029] Furthermore, the judgment module includes a light analysis unit, which is used to analyze the light state based on the collected ambient light intensity g0 and the preset light intensity g1. The light state includes a normal state and a strong light state.

[0030] The illumination analysis unit processes the analysis of Zisha pottery types when the illumination is strong light, and sets the preset judgment threshold after processing to u1.

[0031] Furthermore, the judgment module also includes a humidity analysis unit, which is used to analyze the humidity status based on the collected ambient humidity hs0 and the preset humidity hs1. The humidity status includes a normal state and a high humidity state.

[0032] The humidity analysis unit processes the light intensity analysis when the humidity is high, and sets the pre-set light intensity after processing to g2.

[0033] On the other hand, the present invention also provides a method for identifying Zisha teapots based on image processing, comprising:

[0034] Step S1: Collect images of the Zisha teapot and environmental data;

[0035] Step S2: Perform color normalization processing on the Zisha teapot image to obtain a corrected image;

[0036] Step S3: Divide the corrected image into regions based on the gradient of each pixel in the horizontal direction and the gradient in the vertical direction to obtain the Zisha pottery region.

[0037] Step S4: Extract color features from the HSV image of the Zisha teapot, and construct the color coefficients of the Zisha teapot region based on the extraction results;

[0038] Step S5: Analyze the texture state of the Zisha pottery area based on the contrast and gray-scale entropy of the Zisha pottery area, and analyze the texture features of the Zisha pottery area based on the analysis results.

[0039] Step S6: Divide the pixel types and analyze the gloss characteristics of the Zisha pottery area based on the division results.

[0040] Step S7: Analyze the types of Zisha clay based on the construction results of the color coefficient of the Zisha area, texture characteristics, gloss characteristics, and the collected ambient light intensity and humidity.

[0041] The beneficial effects of this invention are as follows: The data acquisition module, through high-resolution image acquisition and environmental data recording, provides a reliable foundation for subsequent analysis, ensuring data accuracy. In the preprocessing module, the color normalization process eliminates the influence of different shooting conditions on color, making image analysis more consistent; the region division module accurately extracts the outline of the Zisha pottery, allowing subsequent analysis to focus only on the core area of ​​the Zisha pottery; the color feature analysis module accurately captures the unique color of the Zisha pottery by constructing color coefficients, while the texture feature analysis module reveals the details of the surface craftsmanship of the Zisha pottery through the calculation of contrast and grayscale entropy, enhancing the objectivity of identification; the gloss feature analysis module deeply analyzes the surface characteristics of the Zisha pottery through pixel type division and gloss coefficient calculation, providing more dimensions of information for judging authenticity; the judgment module, through comprehensive analysis of color, texture, and gloss features, combined with ambient light and humidity, ensures accurate judgment under different conditions. Through the collaborative work of the above modules, the entire system can achieve efficient and accurate identification of Zisha pottery clay. Attached Figure Description

[0042] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0043] Figure 1 This is a schematic diagram of the structure of the Zisha teapot identification system based on image processing in this embodiment.

[0044] Figure 2 This is a schematic diagram of the texture feature analysis module in this embodiment.

[0045] Figure 3 This is a schematic diagram of the gloss feature analysis module in this embodiment.

[0046] Figure 4 This is a schematic diagram of the judgment module in this embodiment.

[0047] Figure 5 This is a flowchart illustrating the image processing-based method for identifying Zisha teapots in this embodiment. Detailed Implementation

[0048] To more clearly illustrate the present invention, the following description, in conjunction with preferred embodiments and accompanying drawings, further explains the invention. Similar components in the drawings are indicated by the same reference numerals. Those skilled in the art should understand that the specific description below is illustrative rather than restrictive and should not be construed as limiting the scope of protection of the present invention.

[0049] It should be noted that although the terms first, second, third, etc., may be used in the embodiments of this application for description, these descriptions should not be limited to these terms. These terms are only used to distinguish the descriptions. For example, without departing from the scope of the embodiments of this application, first can also be referred to as second, and similarly, second can also be referred to as first.

[0050] Please see Figure 1 As shown, this is a schematic diagram of the image processing-based Zisha teapot identification system of this embodiment. The system includes:

[0051] The data acquisition module is used to acquire images of the Zisha teapot and environmental data. The Zisha teapot images include grayscale images and HSV images of the Zisha teapot. The environmental data includes ambient light intensity and ambient humidity. In this embodiment, the acquisition method of the Zisha teapot images and environmental data is not specifically limited. Those skilled in the art can set it freely, as long as the acquisition requirements of ambient light intensity and ambient humidity are met. Specifically, the Zisha teapot image can be taken under conditions of white natural light or fluorescent light, color temperature controlled at 5500K~6000K, and resolution higher than 1080p. The image is then read and converted into grayscale and HSV images using the OpenCV image processing library. The environmental data can be acquired by a smart sensor when taking images.

[0052] Please continue reading. Figure 1 As shown, the system includes:

[0053] The preprocessing module, which is connected to the data acquisition module, is used to perform color normalization processing on the Zisha pottery image to obtain a corrected image.

[0054] Specifically, the preprocessing module sets the grayscale value of the i-th pixel after color normalization to LQi, where LQi = (Ni / 255). γ ×255, γ is the correction factor, and Ni is the gray value of the i-th pixel in the Zisha pottery image before color normalization. The preprocessing module eliminates color differences caused by shooting conditions through color normalization, making subsequent image analysis more robust and ensuring that the analysis results match the actual color of the Zisha pottery.

[0055] It is understood that this embodiment does not impose specific limitations on the setting of the correction factor. Those skilled in the art can set it freely, as long as the setting requirements of the correction factor are met. The optimal value of the correction factor is 2.2.

[0056] Please continue reading. Figure 1 As shown, the system also includes:

[0057] A region segmentation module, connected to the preprocessing module, is used to calculate the gradient magnitude Pi of the i-th pixel in the corrected image based on the gradient Tix in the horizontal direction and the gradient Tiy in the vertical direction of the i-th pixel. The module sets...

[0058] Pi = (Tix 2 +T iy 2 ) 0.5 ;

[0059] The region segmentation module analyzes the relative position of pixels based on the gradient magnitude Pi of the i-th pixel in the corrected image and a preset gradient magnitude threshold p0, wherein:

[0060] If Pi≤p0, the region division module determines that the pixel is a non-edge pixel; otherwise, the region division module determines that the pixel is an edge pixel.

[0061] The region segmentation module uses the curve formed by connecting each edge pixel as the outline of the Zisha pottery, and the area within the outline as the Zisha pottery region. By calculating the gradient magnitude of each pixel in the image, the region segmentation module can effectively distinguish the outline of the Zisha pottery from the background, extract the key Zisha pottery regions for further analysis, and improve the accuracy of target recognition.

[0062] It is understood that this embodiment does not specifically limit the method of obtaining the gradient and grayscale value of a pixel. Those skilled in the art can set them freely, as long as the requirements for obtaining the gradient and grayscale value of a pixel are met. The gradient of a pixel can be obtained by applying a 3x3 convolution kernel using the Sobe edge detection algorithm. The grayscale value of a pixel can be obtained by reading and converting the image using the OpenCV library in Python. This embodiment does not specifically limit the setting of the preset gradient magnitude threshold. Those skilled in the art can set it freely, as long as the requirements for setting the preset gradient magnitude threshold are met. The optimal value of p0 is 0.3.

[0063] Please continue reading. Figure 1 As shown, the system also includes:

[0064] A color feature analysis module, connected to the data acquisition module, is used to extract color features from the HSV image of the Zisha teapot and construct color coefficients for the Zisha teapot region based on the extraction results.

[0065] The color feature analysis module sets the hue of the HSV image of the Zisha pottery to Hu and the hue coefficient to SD, setting SD = 1 - |Hu - H0| / H0;

[0066] The color feature analysis module sets the saturation of the HSV image of the Zisha pottery to BH and the saturation coefficient to BHD, and sets BHD = 1 - |BH - BH0| / BH0.

[0067] The color feature analysis module sets the color coefficient of Zisha ware to Y, and sets Y = r1 × SD + r2 × BHD. By analyzing the color characteristics of Zisha ware and constructing the color coefficient, the color feature analysis module can effectively capture the unique color expression of Zisha ware, thereby improving the accuracy of Zisha ware clay identification.

[0068] Where r1 is the hue weight, r2 is the saturation weight, r1+r2=1, H0 is the preset hue, BHO is the preset saturation, and when Y<0, the value of Y is 0.

[0069] Specifically, this embodiment does not impose specific limitations on the method of obtaining the hue and saturation of HSV images. Those skilled in the art can freely set the method, as long as the requirements for obtaining the hue and saturation of HSV images are met. The hue and saturation can be obtained through OpenCV's built-in conversion function from RGB to HSV.

[0070] It is understood that this embodiment does not specifically limit the settings of each weight, preset hue and preset saturation. Those skilled in the art can set them freely, as long as the setting requirements of each weight, preset hue and preset saturation are met. Among them, the optimal value of r1 is 0.7, the optimal value of r2 is 0.3, the optimal value of H0 is 10 and the optimal value of BH0 is 40%.

[0071] Please continue reading. Figure 1 As shown, the system also includes:

[0072] The texture feature analysis module, connected to the region division module, is used to analyze the texture state of the Zisha teapot region based on its contrast and image grayscale entropy, and to analyze the texture features of the Zisha teapot region based on the analysis results. It is understood that this embodiment does not specifically limit the acquisition methods of the contrast and image grayscale entropy of the Zisha teapot region; those skilled in the art can freely set them, as long as the acquisition requirements of the contrast and image grayscale entropy of the Zisha teapot region are met. These can be obtained using NumPy and OpenCV in Python.

[0073] Please see Figure 2 As shown, the texture feature analysis module includes:

[0074] The contrast analysis unit is used to analyze the contrast state of the Zisha teapot area based on the contrast d0 of the area, and to construct a contrast coefficient based on the analysis results to quantify the contrast state of the Zisha teapot area, wherein:

[0075] If d1≤d0≤d2, the contrast analysis unit determines that the contrast state of the Zisha pottery area is normal and sets the contrast coefficient to D1, with D1 = 1. Conversely, if d0 < d1, the contrast analysis unit determines that the contrast state of the Zisha pottery area is abnormal. When d0 < d1, the contrast analysis unit sets the contrast coefficient to D2, with D2 = (d1 - d0) / Δd. When d0 > d2, the contrast analysis unit sets the contrast coefficient to D3, with D3 = (d0 - d2) / Δd. By evaluating the contrast state of the Zisha pottery area, the contrast analysis unit helps to analyze surface details, determine its manufacturing process and quality, and thus improve the accuracy of texture feature analysis.

[0076] Where △d is the preset contrast difference, △d=(d1+d2) / 2, d1 is the first preset contrast, d2 is the second preset contrast, and d1<d2.

[0077] It is understood that no specific limitations are made on the setting of each preset contrast ratio in this embodiment. Those skilled in the art can set them freely, as long as the setting requirements of each preset contrast ratio are met. The optimal value of d1 is 7 and the optimal value of d2 is 12.

[0078] Please continue reading. Figure 2 As shown, the texture feature analysis module further includes:

[0079] A gray-scale entropy analysis unit analyzes the gray-scale entropy state of the Zisha pottery region based on the image gray-scale entropy s0 and the preset gray-scale entropy s1, and constructs a gray-scale entropy coefficient based on the analysis results to quantify the gray-scale entropy state of the Zisha pottery region, wherein:

[0080] If s0≤s1, the grayscale entropy analysis unit determines that the grayscale entropy state of the Zisha pottery area is an abnormal state and sets the grayscale entropy coefficient to S1, with S1=0; otherwise, the grayscale entropy analysis unit determines that the grayscale entropy state of the Zisha pottery area is a normal state and sets the grayscale entropy coefficient to S2, with S2=exp[3×(s0-s1) / (s0+s1)-3]; the grayscale entropy analysis unit can quantify the complexity of the texture through grayscale entropy analysis, providing quantitative support for the surface characteristics of Zisha pottery.

[0081] It is understood that no specific limitation is made to the setting of the preset grayscale entropy in this embodiment. Those skilled in the art can set it freely, as long as the setting requirements of the preset grayscale entropy are met. The optimal value of s1 is 1.8.

[0082] Please continue reading. Figure 2 As shown, the texture feature analysis module further includes:

[0083] A texture analysis unit, connected to the contrast analysis unit and the grayscale entropy analysis unit, analyzes the texture state of the Zisha pottery area based on the construction results of the contrast coefficient, and analyzes the texture features of the Zisha pottery area based on the analysis results, wherein:

[0084] If x1×contrast coefficient + x2×grayscale entropy coefficient ≤ WL, the texture analysis unit determines the texture state of the Zisha pottery area to be abnormal and sets the texture feature of the Zisha pottery area to WT1, setting WT1 = 0; otherwise, the texture analysis unit determines the texture state of the Zisha pottery area to be normal and sets the texture feature of the Zisha pottery area to WT2, setting WT2 = lg(x1×contrast coefficient + x2×grayscale entropy coefficient - WL + 1) / lg2; the texture analysis unit determines the texture state of the Zisha pottery by comprehensively analyzing the contrast and grayscale entropy, which helps to improve the scientificity and accuracy of identification, thereby improving the accuracy of Zisha pottery clay type analysis;

[0085] Where x1 is the contrast weight, x2 is the grayscale entropy weight, x1+x2=1, and WL is the preset texture coefficient.

[0086] It is understood that no specific limitations are made on the settings of each weight and the preset texture coefficient in this implementation. Those skilled in the art can set them freely, as long as the setting requirements of each weight and the preset texture coefficient are met. Among them, the optimal value of x1 is 0.4, the optimal value of x2 is 0.6, and the optimal value of WL is 0.2.

[0087] Please continue reading. Figure 1 As shown, the system also includes:

[0088] A gloss feature analysis module, which is connected to the texture feature analysis module, is used to classify pixel types and analyze the gloss features of the Zisha pottery area based on the classification results.

[0089] Please see Figure 3 As shown, the gloss feature analysis module includes:

[0090] The type analysis unit is used to classify the pixel type based on the grayscale value Lk of the k-th pixel in the Zisha pottery area and the preset grayscale threshold YZ, where:

[0091] If Lk≤YZ, the type analysis unit determines that the pixel is a type I pixel; otherwise, the type analysis unit determines that the pixel is a type II pixel. By classifying pixels into different types, the type analysis unit can analyze different features and improve the effectiveness of gloss feature extraction.

[0092] Specifically, this embodiment does not impose specific limitations on the setting of the preset grayscale threshold. Those skilled in the art can set it freely, as long as the setting requirements of the preset grayscale threshold are met. The optimal value of YZ is 150.

[0093] Please continue reading. Figure 3 As shown, the gloss feature analysis module further includes:

[0094] A gloss feature analysis unit, connected to the type analysis unit, calculates the average value PG of the grayscale values ​​of the two types of pixels and sets... LEm is the gray value of the m-th binary pixel, and M is the number of binary pixels;

[0095] The gloss feature analysis unit analyzes the gloss coefficient GX of the Zisha teapot based on the average gray value PG of the two types of pixels and the average pixel value Pu of the Zisha teapot area. The gloss coefficient GX is set as GT = ln(PG / Pu) / ln2. Zn is the grayscale value of the nth pixel in the Zisha pottery area, and N is the number of pixels in the Zisha pottery area.

[0096] The gloss feature analysis unit analyzes the gloss characteristics of the Zisha teapot area based on the gloss coefficient GX and the preset gloss coefficient GX0. If GX < GX0, the gloss feature analysis unit sets the gloss characteristic of the Zisha teapot area to GT1, and sets GT1 = (GX0 - GX) / GX0. Conversely, the gloss feature analysis unit sets the gloss characteristic of the Zisha teapot area to GT2, and sets GT2 = 0. By analyzing the gloss coefficient, the gloss feature analysis unit can reveal the details of the surface treatment process of the Zisha teapot, thereby improving the accuracy of the gloss feature analysis.

[0097] It is understood that this embodiment does not specifically limit the setting of the preset gloss coefficient. Those skilled in the art can set it freely, as long as the setting requirements of the preset gloss coefficient are met. The optimal value of GX0 is 0.3.

[0098] Please continue reading. Figure 1 As shown, the system also includes:

[0099] The judgment module, which is connected to the color feature analysis module and the gloss feature analysis module, is used to analyze the type of Zisha clay based on the construction results of the color coefficient of the Zisha area, the texture features of the Zisha area, the gloss features of the Zisha area, and the collected ambient light intensity and ambient humidity.

[0100] Please see Figure 4 As shown, the judgment module includes:

[0101] The judgment unit analyzes the type of Zisha clay based on the constructed color coefficients, texture characteristics, and gloss characteristics of the Zisha area, and outputs the results to the user.

[0102] If w1×color coefficient of Zisha area + w2×texture feature of Zisha area + w3×gloss feature of Zisha area ≤ u0, the judgment unit determines that the Zisha clay type is Qingni (green clay); otherwise, the judgment unit determines that the Zisha clay type is not Qingni (green clay).

[0103] Wherein, w1 is the color weight, w2 is the texture weight, w3 is the gloss weight, w1+w2+w3=1, and u0 is the preset judgment threshold; the judgment unit combines color coefficient, texture feature and gloss feature for comprehensive analysis, and makes a scientific judgment on the type of Zisha clay from a global perspective, thereby improving the comprehensiveness and accuracy of the identification.

[0104] It is understood that this embodiment does not specifically limit the setting of each weight and the preset judgment threshold. Those skilled in the art can set them freely, as long as the setting requirements of each weight and the preset judgment threshold are met. Among them, the optimal value of w1 is 0.5, the optimal value of w2 is 0.3, the optimal value of w3 is 0.2, and the optimal value of u0 is 0.6.

[0105] Please continue reading. Figure 4 As shown, the judgment module includes:

[0106] A light analysis unit, connected to the judgment unit, is used to analyze the light state based on the collected ambient light intensity g0 and the preset light intensity g1, and to process the analysis of the Zisha teapot type based on the analysis results, wherein:

[0107] If g0≤g1, the illumination analysis unit determines the illumination state as normal; otherwise, the illumination analysis unit determines the illumination state as strong light.

[0108] The light analysis unit processes the analysis of Zisha pottery types under strong light conditions and sets the preset judgment threshold after processing as u1. The threshold is set as u1 = u0 × {1 - β × ln[3 × (g0 - g1) / (g0 + g1) + 1] / ln5}, where β is a preset adjustment ratio coefficient. The light analysis unit can adjust the identification process of Zisha pottery according to the ambient light intensity, ensuring the validity of the analysis results under different light conditions, thereby improving the accuracy of Zisha pottery clay type analysis.

[0109] It is understood that this embodiment does not specifically limit the setting of preset light intensity and preset adjustment ratio coefficient. Those skilled in the art can set them freely, as long as the setting requirements of preset light intensity and preset adjustment ratio coefficient are met. The optimal value of g1 is 800 Lux, and the optimal value of β is 0.36.

[0110] Please continue reading. Figure 4 As shown, the judgment module further includes:

[0111] A humidity analysis unit, connected to the light analysis unit, is used to analyze the humidity status based on the collected ambient humidity hs0 and the preset humidity hs1, and to process the light status analysis process based on the analysis results, wherein:

[0112] If hs0≤hs1, the humidity analysis unit determines the humidity state as normal; otherwise, the humidity analysis unit determines the humidity state as high humidity.

[0113] The humidity analysis unit processes the light intensity analysis when the humidity is high, and sets the pre-set light intensity after processing as g2. The formula is g2 = g1 × [1 - η × (hs0 - hs1) / hs1], where η is a pre-set correction coefficient. By evaluating the impact of ambient humidity on the analysis process, the humidity unit can more accurately adjust the judgment results of Zisha teapots, ensuring their stability and reliability when humidity changes, thereby improving the accuracy of Zisha clay type analysis.

[0114] It is understood that this embodiment does not specifically limit the setting of preset humidity and preset correction ratio coefficient. Those skilled in the art can set them freely, as long as the setting requirements of preset humidity and preset correction ratio coefficient are met. The optimal value of hs1 is 70%, and the optimal value of η is 0.42.

[0115] Please see Figure 5 As shown, it is a flowchart illustrating the image processing-based method for identifying Zisha teapots in this embodiment, including:

[0116] Step S1: Collect images of the Zisha teapot and environmental data;

[0117] Step S2: Perform color normalization processing on the Zisha teapot image to obtain a corrected image;

[0118] Step S3: Divide the corrected image into regions based on the gradient of each pixel in the horizontal direction and the gradient in the vertical direction to obtain the Zisha pottery region.

[0119] Step S4: Extract color features from the HSV image of the Zisha teapot, and construct the color coefficients of the Zisha teapot region based on the extraction results;

[0120] Step S5: Analyze the texture state of the Zisha pottery area based on the contrast and gray-scale entropy of the Zisha pottery area, and analyze the texture features of the Zisha pottery area based on the analysis results.

[0121] Step S6: Divide the pixel types and analyze the gloss characteristics of the Zisha pottery area based on the division results.

[0122] Step S7: Analyze the types of Zisha clay based on the construction results of the color coefficient of the Zisha area, texture characteristics, gloss characteristics, and the collected ambient light intensity and humidity.

[0123] Specifically, the image processing-based Zisha teapot identification method and system described in this embodiment is applied to the identification of Zisha clay types. It involves acquiring images of the Zisha teapot and environmental data, followed by color normalization and contour extraction. Through comprehensive analysis of color coefficients, texture, and gloss characteristics, combined with ambient light and humidity, the type of Zisha clay is accurately determined. This approach significantly improves the efficiency and accuracy of Zisha teapot identification through a scientific and standardized method.

[0124] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is impossible to exhaustively list all the implementation methods here. All obvious variations or modifications derived from the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. A Zisha teapot identification system based on image processing, characterized in that, include: The data acquisition module is used to collect images of Zisha teapots and environmental data; The preprocessing module is used to perform color normalization on the Zisha teapot image to obtain a corrected image; The region segmentation module is used to segment the corrected image into regions based on the gradient of each pixel in the horizontal direction and the gradient in the vertical direction, so as to obtain the Zisha pottery region. The color feature analysis module is used to extract color features from the HSV image of Zisha teapots and construct the color coefficients of the Zisha teapot area based on the extraction results. The texture feature analysis module is used to analyze the texture state of the Zisha teapot area based on the contrast and gray-level entropy of the Zisha teapot area, and to analyze the texture features of the Zisha teapot area based on the analysis results. The gloss feature analysis module is used to classify pixel types and analyze the gloss features of the Zisha teapot area based on the classification results. The judgment module is used to analyze the type of clay material of Zisha teapot based on the construction results of the color coefficient of the Zisha teapot area, texture characteristics, gloss characteristics, and the collected ambient light intensity and humidity. The color feature analysis module sets the hue of the HSV image of the Zisha pottery to Hu and the hue coefficient to SD, setting SD=1-|Hu-H0| / H0; The color feature analysis module sets the saturation of the HSV image of the Zisha pottery to BH and the saturation coefficient to BHD, and sets BHD=1-|BH-BH0| / BH0; The color feature analysis module sets the color coefficient of Zisha pottery as Y, and sets Y=r1×SD+r2×BHD; Where r1 is the hue weight, r2 is the saturation weight, H0 is the preset hue, and BHO is the preset saturation. The judgment module includes a judgment unit, which is used to analyze the type of Zisha clay based on the construction results of the color coefficient of the Zisha area, the texture characteristics of the Zisha area, and the gloss characteristics of the Zisha area, wherein: If w1×color coefficient of Zisha area + w2×texture feature of Zisha area + w3×gloss feature of Zisha area ≤ u0, the judgment unit determines that the Zisha clay type is Qingni (green clay); otherwise, the judgment unit determines that the Zisha clay type is not Qingni (green clay). Where w1 is the color weight, w2 is the texture weight, w3 is the gloss weight, and u0 is the preset judgment threshold; The judgment module includes a light analysis unit, which is used to analyze the light state based on the collected ambient light intensity g0 and the preset light intensity g1, and to process the analysis process of the Zisha teapot type based on the analysis results, wherein: If g0≤g1, the illumination analysis unit determines the illumination state as normal; otherwise, the illumination analysis unit determines the illumination state as strong light. The light analysis unit processes the analysis of Zisha ware types when the light is strong, and sets the preset judgment threshold after processing as u1. The preset threshold is set as u1=u0×{1-β×ln[3×(g0-g1) / (g0+g1)+1] / ln5}, where β is a preset adjustment ratio coefficient. The judgment module further includes a humidity analysis unit, which is used to analyze the humidity status based on the collected ambient humidity hs0 and the preset humidity hs1, and to process the analysis process of the light status based on the analysis results, wherein: If hs0≤hs1, the humidity analysis unit determines the humidity state as normal; otherwise, the humidity analysis unit determines the humidity state as high humidity. The humidity analysis unit processes the light intensity analysis when the humidity is high, and sets the preset light intensity after processing as g2, where g2 = g1 × [1 - η × (hs0 - hs1) / hs1], and η is a preset correction ratio coefficient.

2. The image processing-based Zisha teapot identification system according to claim 1, characterized in that, The region division module calculates the gradient magnitude Pi of the i-th pixel in the corrected image based on the gradient Tix in the horizontal direction and the gradient Tiy in the vertical direction of the i-th pixel in the corrected image, and analyzes the relative position of the pixel based on the gradient magnitude Pi of the i-th pixel in the corrected image and the preset gradient magnitude threshold p0. The relative position of the pixel includes non-edge pixels and edge pixels. The region division module uses the curve formed by connecting each edge pixel as the outline of the Zisha pottery, and the area within the outline of the Zisha pottery as the Zisha pottery area.

3. The image processing-based Zisha teapot identification system according to claim 2, characterized in that, The texture feature analysis module includes a contrast analysis unit, which analyzes the contrast state of the Zisha teapot area based on the contrast d0 of the Zisha teapot area, and constructs a contrast coefficient based on the analysis results to quantify the contrast state of the Zisha teapot area, wherein: If d1≤d0≤d2, the contrast analysis unit determines that the contrast state of the Zisha pottery area is normal and sets the contrast coefficient to D1, with D1=1; otherwise, the contrast analysis unit determines that the contrast state of the Zisha pottery area is abnormal. When d0<d1, the contrast analysis unit sets the contrast coefficient to D2, with D2=(d1-d0) / △d; when d0>d2, the contrast analysis unit sets the contrast coefficient to D3, with D3=(d0-d2) / △d. Where △d is the preset contrast difference, △d=(d1+d2) / 2, d1 is the first preset contrast, and d2 is the second preset contrast; The texture feature analysis module also includes a gray-level entropy analysis unit. The gray-level entropy analysis unit is used to analyze the gray-level entropy state of the Zisha pottery area based on the image gray-level entropy s0 and the preset gray-level entropy s1. The gray-level entropy state of the Zisha pottery area includes a normal state and an abnormal state. If the gray-level entropy state of the Zisha pottery area is an abnormal state, the gray-level entropy coefficient is set to S1. If the gray-level entropy state of the Zisha pottery area is a normal state, the gray-level entropy coefficient is set to S2. S2 is set to exp[3×(s0-s1) / (s0+s1)-3].

4. The image processing-based Zisha teapot identification system according to claim 3, characterized in that, The texture feature analysis module includes a texture analysis unit, which analyzes the texture state of the Zisha pottery area based on the construction result of the contrast coefficient. If x1×contrast coefficient + x2×grayscale entropy coefficient ≤ WL, the texture analysis unit determines that the texture state of the Zisha pottery area is abnormal and sets the texture feature of the Zisha pottery area to WT1, setting WT1=0; otherwise, the texture analysis unit determines that the texture state of the Zisha pottery area is normal and sets the texture feature of the Zisha pottery area to WT2, setting WT2=lg(x1×contrast coefficient + x2×grayscale entropy coefficient - WL+1) / lg2, where x1 is the contrast weight, x2 is the grayscale entropy weight, x1+x2=1, and WL is the preset texture coefficient.

5. The image processing-based Zisha teapot identification system according to claim 4, characterized in that, The gloss feature analysis module includes a type analysis unit, which is used to classify the pixel type based on the grayscale value Lk of the k-th pixel in the Zisha pottery area and a preset grayscale value threshold YZ, wherein: If Lk≤YZ, the type analysis unit determines that the pixel is a type 1 pixel; otherwise, the type analysis unit determines that the pixel is a type 2 pixel. The gloss feature analysis module further includes a gloss feature analysis unit, which is used to calculate the average value PG of the grayscale values ​​of the two types of pixels and set... LEm is the gray value of the m-th binary pixel, and M is the number of binary pixels; The gloss feature analysis unit analyzes the gloss coefficient GX of the Zisha teapot based on the average gray value PG of the two types of pixels and the average pixel value Pu of the Zisha teapot area, setting GX=ln(PG / Pu) / ln2. Zn is the gray value of the nth pixel in the Zisha pottery area, and N is the number of pixels in the Zisha pottery area. The gloss feature analysis unit analyzes the gloss features of the Zisha teapot area based on the gloss coefficient GX of the Zisha teapot and the preset gloss coefficient GX0. If GX < GX0, the gloss feature analysis unit sets the gloss feature of the Zisha teapot area to GT1, and sets GT1 = (GX0 - GX) / GX0. Otherwise, the gloss feature analysis unit sets the gloss feature of the Zisha teapot area to GT2, and sets GT2 = 0.

6. A method for identifying Zisha teapots based on image processing, applied to the Zisha teapot identification system based on image processing as described in any one of claims 1-5, characterized in that, include: Step S1: Collect images of the Zisha teapot and environmental data; Step S2: Perform color normalization processing on the Zisha teapot image to obtain a corrected image; Step S3: Divide the corrected image into regions based on the gradient of each pixel in the horizontal direction and the gradient in the vertical direction to obtain the Zisha pottery region. Step S4: Extract color features from the HSV image of the Zisha teapot, and construct the color coefficients of the Zisha teapot region based on the extraction results; Step S5: Analyze the texture state of the Zisha pottery area based on the contrast and gray-scale entropy of the Zisha pottery area, and analyze the texture features of the Zisha pottery area based on the analysis results. Step S6: Divide the pixel types and analyze the gloss characteristics of the Zisha pottery area based on the division results; Step S7: Analyze the types of Zisha clay based on the construction results of the color coefficient of the Zisha area, texture characteristics, gloss characteristics, and the collected ambient light intensity and humidity.

Images