Material sorting method, apparatus, device, and storage medium

Abstract

This disclosure relates to the field of ore identification technology, specifically to a material sorting method, apparatus, device, and storage medium. The material sorting method includes: acquiring a ray profile image of the material to be sorted and a corresponding color image; determining the local content of the target component in the material to be sorted layer by layer based on the ray transmittance of each pixel in the color image and the target geometric dimension, through a preset first correspondence; determining the local purity of the target component in the material to be sorted layer by layer based on a comparison between the actual color distribution and the standard color distribution of each layer of the color image, through a second correspondence; determining the total content of the target component in the material to be sorted based on the local content and corresponding local purity of each layer; and sorting the material to be sorted based on the total content. This method can reduce data processing latency, reduce error interference from single-dimensional detection, and make the determination process more accurate and reliable, thereby effectively improving the efficiency and performance of material sorting.

Description

Technical Field

[0001] This disclosure relates to the field of ore identification technology, specifically to a material sorting method, a material sorting device, a material sorting equipment, and a computer-readable storage medium. Background Technology

[0002] With the development of technology, ore sorting technology has gradually evolved from traditional physical sorting methods such as manual sorting, gravity sorting, and magnetic separation to automated sorting based on "sensors + algorithms".

[0003] While ore sorting can be achieved by fusing ray and visual data in related technologies, these methods are mostly simple superpositions and rely on 3D modeling. This not only increases the cost of equipment deployment but also leads to a decrease in sorting efficiency and equipment stability due to the complexity of the algorithm logic. Summary of the Invention

[0004] To overcome the problems existing in related technologies, an exemplary embodiment of this disclosure provides a material sorting method, including: acquiring a ray profile map of the material to be sorted and a corresponding color image; determining the local content of the target component in the material to be sorted layer by layer based on the ray transmittance of each pixel in the color image and the target geometric dimension through a preset first correspondence relationship, wherein the first correspondence relationship is used to characterize the relationship between ray transmittance, a fixed distance geometric parameter range, and the content of the target component; determining the local purity of the target component in the material to be sorted layer by layer based on the comparison result between the actual color distribution and the standard color distribution of each layer of the color image through a second correspondence relationship, wherein the second correspondence relationship is used to characterize the relationship between color difference and the purity of the target component; determining the total content of the target component in the material to be sorted based on the local content and the corresponding local purity of each layer; and sorting the material to be sorted based on the total content.

[0005] In some embodiments, determining the total content of the target component in the material to be sorted based on the local content and corresponding local purity of each layer includes: determining the local component content of the target component in each layer based on the local content and corresponding local purity of each layer; determining the local volume of the material in each layer based on multiple pixels and a fixed distance geometric parameter range; and determining the total content of the target component in the material to be sorted based on the local volume of the material in each layer and the corresponding local component content.

[0006] In some embodiments, sorting materials to be sorted based on total content includes: determining the number of effective layers based on the local component content of each layer and a target threshold; determining the percentage of effective layers based on the total number of layers and the number of effective layers of the materials to be sorted; determining the category of the materials to be sorted based on the total content and the percentage of effective layers; and sorting the materials to be sorted according to the category.

[0007] In some embodiments, based on the ray transmittance of each pixel in the color image and the target geometric dimension, the local content of the target component in the material to be sorted is determined layer by layer through a preset first correspondence, including: determining the target geometric parameter range of the material to be sorted in the target geometric dimension based on the ray transmittance of each pixel in the color image; determining the minimum and maximum geometric parameters of each layer according to the interval of the geometric parameters; and determining the local content of the target component in the material to be sorted layer by layer according to the minimum and maximum geometric parameters of each layer, the ray transmittance of each pixel, and the first correspondence.

[0008] In some embodiments, the geometric parameter is a height value; the local content of the target component in the material to be sorted is determined layer by layer based on the minimum and maximum geometric parameters of each layer, the ray transmittance of each pixel, and a first correspondence, including: determining multiple target pixels corresponding to the current layer based on the minimum and maximum height values ​​of the current layer; determining the first ray transmittance corresponding to the minimum height value, the second ray transmittance corresponding to the maximum height value, and the neighborhood content of the target component between the minimum and maximum height values ​​through the first correspondence; and determining the local content of the target component in the current layer of the material to be sorted based on the first ray transmittance, the second ray transmittance, the neighborhood content, the ray transmittance of each target pixel, and the corresponding height value.

[0009] In some embodiments, based on the comparison results between the actual color distribution and the standard color distribution of each layer of color image, the local purity of the target component in the material to be sorted is determined layer by layer through a second correspondence relationship, including: determining the actual color distribution of the current layer based on the processing result after converting the color space of the current layer color image; comparing the actual color distribution and the standard color distribution to obtain the actual color difference; and determining the local purity of the target component in the current layer of the material to be sorted through the second correspondence relationship based on the actual color difference.

[0010] In some embodiments, the method further includes: updating the first correspondence based on the sorting results of the materials to be sorted.

[0011] Secondly, this disclosure also provides a material sorting device, comprising: an acquisition module for acquiring a ray profile map and a corresponding color image of the material to be sorted; a first processing module for determining the local content of a target component in the material to be sorted layer by layer based on the ray transmittance and target geometric dimension of each pixel in the color image, through a preset first correspondence relationship, wherein the first correspondence relationship is used to characterize the relationship between ray transmittance, a fixed distance geometric parameter range, and the content of the target component; a second processing module for determining the local purity of the target component in the material to be sorted layer by layer based on the comparison result between the actual color distribution and the standard color distribution of each layer of the color image, through a second correspondence relationship, wherein the second correspondence relationship is used to characterize the relationship between color difference and the purity of the target component; a third processing module for determining the total content of the target component in the material to be sorted based on the local content and the corresponding local purity of each layer; and a fourth processing module for sorting the material to be sorted based on the total content.

[0012] Thirdly, this disclosure also provides a material sorting device, including: a memory and a processor, which are interconnected and communicate with each other. The memory stores computer instructions, and the processor executes the computer instructions to perform the material sorting method provided in any of the above aspects.

[0013] Fourthly, this disclosure also provides a computer-readable storage medium storing computer instructions, which are used to cause a computer to perform the material sorting method provided in any of the above aspects.

[0014] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure.

[0015] The technical solutions provided by the embodiments of this disclosure can include the following beneficial effects: According to the material sorting method provided by this disclosure, by using a pre-determined first and second correspondence relationship, combined with the dual-modal features of the ray profile map and the color image, the difficulty of determining the local content and local purity of the target component in each layer can be significantly reduced, data processing delay can be reduced, and the identification efficiency and sorting process can be effectively improved. Furthermore, by jointly determining the total content of the target component based on the local content and local purity of each layer, the feature information provided by the ray profile map and the corresponding color image can be fully utilized, reducing the error interference of single-dimensional detection, making the determination process more accurate and reliable, thereby effectively improving the efficiency and performance of material sorting. Attached Figure Description

[0016] This disclosure can be better understood by describing exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, in which: Figure 1This is a schematic diagram of the architecture of a material sorting device according to an exemplary embodiment disclosed in a book. Figure 2 This is a schematic flowchart illustrating a material sorting method according to an exemplary embodiment of a published document; Figure 3 This is a schematic flowchart illustrating another material sorting method according to an exemplary embodiment of a published document; Figure 4 This is a schematic diagram of the structure of a material sorting device according to an exemplary embodiment disclosed in a book. Detailed Implementation

[0017] The following describes specific embodiments of this disclosure. It should be noted that, in order to provide a concise description, this specification cannot exhaustively describe all features of the actual embodiments. It should be understood that, in the actual implementation of any embodiment, just as in any engineering or design project, various specific decisions are often made to achieve the developer's specific goals and to meet system-related or business-related constraints, and this can change from one embodiment to another. Furthermore, it is understood that although the efforts made in this development process may be complex and lengthy, for those skilled in the art related to the content disclosed in this disclosure, changes in design, manufacturing, or production based on the technical content disclosed in this disclosure are merely conventional technical means and should not be construed as insufficient content of this disclosure.

[0018] Unless otherwise defined, the technical or scientific terms used in this disclosure shall have the ordinary meaning understood by one of ordinary skill in the art to which this disclosure pertains. The terms “first,” “second,” and similar terms used in this disclosure do not indicate any order, quantity, or importance, but are merely used to distinguish different components. The terms “a” or “one,” etc., do not indicate a quantity limitation, but rather indicate the presence of at least one. The terms “comprising,” “including,” etc., mean that the element or object preceding “comprising” or “including” encompasses the elements or objects listed following “comprising” or “including” and their equivalents, and do not exclude other elements or objects. The terms “connected,” “linked,” etc., are not limited to physical or mechanical connections, nor are they limited to direct or indirect connections.

[0019] With technological advancements, ore sorting technology has evolved from traditional physical sorting methods such as manual sorting, gravity sorting, and magnetic separation to automated sorting based on "sensors + algorithms." Among these technologies, material sorting techniques are mainly categorized into three types: traditional sorting technologies, single-sensor sorting technologies, and multi-source fusion technologies.

[0020] Traditional sorting technologies rely primarily on human experience or differences in physical properties, but suffer from low efficiency and poor accuracy, making them suitable only for coarse sorting operations. Single-sensor sorting technologies use data from a single dimension collected by a sensor for sorting, but are limited by the one-sidedness of the data dimension, making it difficult to guarantee the final sorting accuracy. Multi-source fusion technology offers superior sorting accuracy compared to the previous two types. Its core idea is to fuse X-ray and visual data for material sorting. For example, during ore sorting operations, equipment can combine collected ore images to complete target identification and classification, thereby implementing targeted sorting. Figure 1 As shown, the ore sorting equipment 100 may include a feeding mechanism 110, a conveying mechanism 120, a detection mechanism 130, and a sorting device 140. The feeding mechanism 110 feeds the ore to be sorted into the conveying mechanism 120. The conveying mechanism 120 may be a conveyor belt or a chute, etc., used to transport the ore fed by the feeding mechanism 110. The detection mechanism 130 performs laser beam detection on the ore conveyed on the conveying mechanism 120 to detect whether the ore is to be rejected. The ore to be rejected refers to the ore that will be separated by the sorting device. The ore to be rejected can be desired ore or undesirable ore, as long as ore sorting can be achieved. The sorting device 140 is used to remove the ore to be rejected. However, in actual data processing, multi-source fusion technology often involves simple superposition of ray and visual data rather than deep linkage, and it requires the assistance of laser 3D modeling technology. This not only increases the overall deployment cost of the equipment, but also leads to reduced sorting efficiency and decreased equipment operation stability due to the complexity of the algorithm logic.

[0021] To address the aforementioned problems, this disclosure provides a material sorting method. For example... Figure 2 As shown, the material sorting method provided in this disclosure may include the following steps: Step S210: Obtain the ray profile of the material to be sorted and the corresponding color image.

[0022] The X-ray profile of the material to be sorted can be obtained by scanning the material with an X-ray camera during material transport and then processing it for profile recognition. The material to be sorted can include, but is not limited to, ores, metals, or other solid objects. The profile extraction algorithm can be determined according to requirements.

[0023] For example, the contour extraction algorithm could be the Douglas-Peucker algorithm. The transported material is scanned using an X-ray camera to obtain a grayscale image. This grayscale image is preprocessed to determine an initial contour point set. Baseline line segments are generated by piecewise linear fitting according to the point set order. The perpendicular distance between each point and its corresponding line segment is calculated using the point-to-line formula to quantify the degree of deviation. For example, the first and last points of the contour are... , Midpoint Then the perpendicular distance from P to line segment AB is for: A threshold is set based on the distance standard deviation to remove noise points with excessive deviation and redundant points with excessively small deviations, thus simplifying the point set. The filtered point set is then recursively iterated, repeatedly removing non-critical nodes that deviate from the fitted line based on the threshold rules of the Douglas-Peucker algorithm, retaining contour feature points and reducing the complexity of the point set. Background pseudo-contours can be filtered out by setting area, perimeter, and other thresholds, ultimately outputting a ray contour map that matches the actual shape of the material to be sorted. For example, the simplified contour vertices are: The formula for determining the contour area S can be: The validity of a contour is determined by comparing its area S with an area threshold. If the area S is greater than or equal to the area threshold, the contour is considered valid, thus obtaining the desired ray contour map.

[0024] By mapping pixel coordinates, the ray profile map is matched with the complete color image of the material to be sorted in the field of view, thereby obtaining a color image corresponding to the material to be sorted, so as to avoid interference from cross-material data in subsequent processing.

[0025] Step S220: Based on the ray transmittance of each pixel in the color image and the target geometric dimension, the local content of the target component in the material to be sorted is determined layer by layer through a preset first correspondence.

[0026] The target component refers to a specific element or substance contained in the desired material category. For example, if the material to be sorted is ore, the target component could be metals or non-metals such as silicon, copper, and tin. Different elements have inherent differences in X-ray transmittance, and variations in the content of the same element will correspond to gradient fluctuations in transmittance. Therefore, to improve sorting efficiency, a first correspondence is established in advance. Then, based on the X-ray transmittance of each pixel in the determined color image and the target geometric dimension, the distribution of X-ray transmittance of each layer of pixels is determined within a fixed-distance geometric parameter range. Based on the matching results, the local content of the target component in the corresponding layer of the material to be sorted is determined. The first correspondence is used to characterize the relationship between X-ray transmittance, the fixed-distance geometric parameter range, and the content of the target component. The target geometric dimension can include, but is not limited to, height, width, or area. For example, if the target geometric dimension is height, the fixed-distance geometric parameter range is a specified height difference, and each layer is an equal-height range of the specified height difference. If the target geometric dimension is width, the fixed-distance geometric parameter range is a specified width difference, and each layer is an equal-width range of the specified height difference. If the target geometric dimension is area, then the interval of the fixed-distance geometric parameter is the specified area, and each layer is a local image region of the specified area.

[0027] Determining the local content of the target component using this method effectively simplifies the determination process, improves efficiency, and facilitates the sorting process, thereby enhancing sorting response capabilities. Furthermore, determining the local content of the target component layer by layer based on a fixed-distance geometric parameter range reduces error interference and improves the accuracy of content determination.

[0028] Step S230: Based on the comparison results between the actual color distribution and the standard color distribution of each layer of color image, the local purity of the target component in the material to be sorted is determined layer by layer through the second correspondence.

[0029] A standard color distribution can be understood as a reference for the actual color distribution of the material. When the material to be sorted and the reference material are of the same type, their color distributions have a high degree of similarity; when the material types are different, the color distributions show significant differences. Therefore, to determine the purity of the target component in the material to be sorted, a layer-by-layer comparison is used to determine the comparison results between the actual color distribution and the standard color distribution of each layer of the color image. Then, through a pre-determined second correspondence, the local purity of the target component in each layer is quickly determined. This not only improves the determination efficiency but also effectively reduces error interference and improves the accuracy of purity determination. The second correspondence is used to characterize the relationship between color difference and the purity of the target component.

[0030] Step S240: Based on the local content and corresponding local purity of each layer, determine the total content of the target component in the material to be sorted.

[0031] By analyzing the local content and purity of the target component in each layer, the distribution of the target component in the material to be sorted can be comprehensively analyzed from two dimensions: component proportion and purity level. This helps to ensure the reliability and accuracy of local component data.

[0032] Since the component data of each layer has high accuracy, the total content of the target component in the material to be sorted can be determined comprehensively, which can significantly improve the accuracy and precision of the determination of the total content.

[0033] Step S250: Based on the total content, sort the materials to be sorted.

[0034] The required total content varies depending on the specific operating conditions. Therefore, by clearly defining the total content of the target component in the material to be sorted, it can be determined whether it is the desired material, allowing for targeted sorting to meet the material sorting requirements.

[0035] For example, taking the current working condition where ore needs to be separated into concentrate, middlings, and waste ore, the criteria for determining whether ore is concentrate is that the total content of the target component must be greater than or equal to 80%; the criteria for determining whether ore is middlings is that the total content of the target component is between 50% and 80%; and the criteria for determining whether ore is waste ore is that the total content of the target component is less than 50%. If the total content of the target component in the material to be separated is 95%, then the material to be separated is determined to be concentrate. If the total content of the target component in the material to be separated is 70%, then the material to be separated is determined to be middlings. If the total content of the target component in the material to be separated is 26%, then the material to be separated is determined to be waste ore.

[0036] According to the material sorting method provided in this disclosure, by using a pre-determined first and second correspondence relationship, combined with the dual-modal features of the ray profile map and the color image, the difficulty of determining the local content and purity of the target component in different layers can be significantly reduced, data processing delay can be reduced, and the identification efficiency can be effectively improved and the sorting process can be accelerated. Furthermore, based on the local content and purity of each layer, the total content of the target component is determined together, which can fully utilize the feature information provided by the ray profile map and the corresponding color image, reduce the error interference of single-dimensional detection, and make the determination process more accurate and reliable, thereby effectively improving the efficiency and performance of material sorting.

[0037] In some embodiments, step S240 above may include the following steps: Step a1: Determine the local component content of the target component in each layer based on the local content and corresponding local purity.

[0038] To reduce errors from a single data source and ensure data reliability, for each layer of the material to be sorted, the local content of the target component in that layer is determined through data fusion, combining the local content and purity of the target component. The data fusion method can include, but is not limited to, weighted fusion, probabilistic fusion, and decision fusion, and the specific method can be determined according to requirements. For example, using weighted fusion to determine the local component content... The expression can be: , in, This represents the local content of the target component in the i-th layer. This represents the local purity of the target component in the i-th layer. This represents the weighting coefficient, which can be set to 0.6 to highlight the advantage of X-rays in detecting internal content.

[0039] Step a2: Determine the local volume of the material in each layer based on the multiple pixels in each layer and the fixed-distance geometric parameter range.

[0040] The local volume of each material layer is determined separately. Based on this local volume and the corresponding local content of the target component, the concentration of the target component in that layer can be determined. For example, taking a fixed-distance geometric parameter range as a specified height difference, the local volume of a single-layer material can be determined using the following formula. : ; in, This represents the area of ​​the ROI in the i-th layer (multiple pixels in the i-th layer × actual pixel area). This represents a fixed-distance geometric parameter range.

[0041] Step a3: Based on the local volume of the material in each layer and the corresponding local component content, determine the total content of the target component in the material to be sorted.

[0042] Based on the local volume of the material in each layer and the corresponding local component content, the distribution of the target component in different locations of the material to be sorted can be determined. Then, the contents of each local component are integrated, and the total content of the target component in the material to be sorted is determined by the comprehensive area. This can effectively reduce error interference and ensure the accuracy and stability of the total content.

[0043] For example, the total content of a target component can be determined by calculating the average. The relevant formula is as follows: ;in, This indicates the total content.

[0044] In other embodiments, step S250 above may include the following steps: Step b1: Determine the number of effective layers based on the local component content and target threshold of each layer; Step b2: Determine the percentage of effective layers based on the total number of layers and the number of effective layers of the material to be sorted; Step b3: Determine the category of the material to be sorted based on the total content and the percentage of effective layers; Step b4: Sort the materials to be sorted according to their categories.

[0045] Specifically, a minimum effective content value for the target component is predetermined, i.e., a target threshold is determined. The local component content of each layer is compared with the target threshold to determine whether the target component content in each layer is valid data. If the local component content is greater than or equal to the target threshold, the layer can be considered a valid layer.

[0046] The effectiveness of determining the local component content of the target component in each layer is verified layer by layer, and the number of effective layers is counted. By comparing the number of effective layers with the total number of layers actually divided in the material to be sorted, the proportion of effective layers can be determined. This allows us to identify the regions in the material to be sorted that contain the target component and the distribution of its total content. Furthermore, by combining the total content with the analysis of the content, the category of the material to be sorted can be determined, ensuring the accuracy of category screening and improving the performance of material sorting.

[0047] In some embodiments, step S220 above may include the following steps: Step c1: Based on the ray transmittance of each pixel in the color image, determine the range of target geometric parameters of the material to be sorted in the target geometric dimension.

[0048] The magnitude of X-ray transmittance can reflect the thickness of the corresponding material entity to be sorted. Therefore, the correspondence between the target geometric dimension and X-ray transmittance can be established by data fitting. Then, based on the gradient change of X-ray transmittance at each pixel, the range of target geometric parameters of the material to be sorted in the target geometric dimension can be determined.

[0049] For example, taking the target's geometric dimension as height, the relationship between height and ray transmittance can be expressed by the following formula: Where k represents the calibration coefficient, which can be obtained based on prior knowledge or sample testing. This represents the ray transmittance of each pixel.

[0050] Step c2: Determine the minimum and maximum geometric parameters for each layer based on the interval of the geometric parameters.

[0051] Based on the fixed-distance geometric parameter range, the local geometric parameter range corresponding to each layer can be clearly defined, and then the target geometric parameters of the material to be sorted can be divided into continuous intervals to determine the division boundary of each layer, and finally the minimum and maximum geometric parameters of each layer can be obtained.

[0052] For example, the target geometric parameters range from 0 to 10, and the distance geometric parameters range... If both units are the same, then the material to be sorted can be divided into 5 layers: First layer: minimum geometric parameter is 0, maximum geometric parameter is 2, and the interval range is [0,2]. Second layer: The minimum geometric parameter is 2, the maximum geometric parameter is 4, and the interval range is [2,4]. The third layer has a minimum geometric parameter of 4, a maximum geometric parameter of 6, and a range of [4,6]. Fourth layer: minimum geometric parameter is 6, maximum geometric parameter is 8, and the interval range is [6,8]. Fifth layer: The minimum geometric parameter is 8, the maximum geometric parameter is 10, and the interval range is [8,10].

[0053] Step c3: Based on the minimum and maximum geometric parameters of each layer, the ray transmittance of each pixel, and the first correspondence, determine the local content of the target component in the material to be sorted layer by layer.

[0054] Based on the minimum and maximum geometric parameters of the current layer, the ray transmittance corresponding to each pixel in the current layer can be determined. Then, combined with the preset first correspondence, the target component reference value corresponding to the geometric parameter range of the layer and the ray transmittance can be matched by data interpolation or data fitting, thereby obtaining the local content of the target component of the material to be sorted in the layer.

[0055] For example, when the geometric parameter range of a certain layer is [2,4], the ray transmittance of all pixels in this range is extracted. Through the transmittance-content mapping relationship corresponding to the geometric range [2,4] in the first correspondence, the content value of the intermediate non-standard data points can be determined, and thus the local content of the target component of the layer can be obtained.

[0056] In some examples, the geometric parameter is a height value, and step c3 above may include the following steps: Step c31: Determine multiple target pixels corresponding to the current layer based on the minimum and maximum height values ​​of the current layer; Step c32: Through the first correspondence, determine the first ray transmittance corresponding to the minimum height value, the second ray transmittance corresponding to the maximum height value, and the neighborhood content of the target component between the minimum height value and the maximum height value; Step c33: Based on the first ray transmittance, the second ray transmittance, the neighborhood content, and the ray transmittance and corresponding height value of each target pixel, determine the local content of the target component of the material to be sorted in the current layer.

[0057] Specifically, taking height as an example using geometric parameters, the minimum height value of the current layer can be determined based on the fixed-distance geometric parameter range. and maximum height value Based on the relationship between ray transmittance, the interval of the fixed-distance geometric parameters, and the content of the target component in the first correspondence, the minimum height value can be directly determined. Corresponding first ray transmittance Maximum height value Corresponding second ray transmittance And the neighborhood content of the target component corresponding to that layer. The neighborhood content may include, but is not limited to, the neighborhood content corresponding to the minimum height value. Neighborhood content between minimum and maximum height values Neighborhood content between maximum and minimum height values and the neighborhood content corresponding to the maximum height value .

[0058] For the content corresponding to the i-th pixel in the current layer This can be determined using bilinear data interpolation. For example, the determination process can be expressed by the following formula: , in, This represents the ray transmittance of the i-th pixel. This represents the height value corresponding to the i-th pixel.

[0059] Based on the content corresponding to pixels at various height positions in the current layer, the content of the target component corresponding to each pixel in the current layer can be clearly identified. Therefore, through comprehensive analysis, the local content of the target component in the material to be sorted within the current layer can be obtained. The comprehensive analysis algorithm may include, but is not limited to, weighted average method and post-denoising statistical method. For example, the weighted average method may include: using the height value of each pixel in the current layer as a weight, and then performing a weighted average of the content of each pixel to obtain the local content of the target component in the material to be sorted within the current layer. The post-denoising statistical method may include: removing noise content in the current layer through threshold screening, and then determining the local content of the target component in the current layer by statistical analysis of the retained content. The processing methods used in the statistical analysis may include, but are not limited to, the median method, the arithmetic mean method, and the confidence interval method with root mean square error assistance.

[0060] In some embodiments, step S230 above may include the following steps: Step d1: Based on the processing results of the current layer color image after color space conversion, determine the actual color distribution of the current layer.

[0061] Different target components exhibit varying color distributions during actual image acquisition, and the native color spaces of different image acquisition devices may differ in format or color gamut. Therefore, converting the current layer's color image to the target color space unifies the data format, reduces processing complexity, improves efficiency, and enhances the accuracy of color feature representation, ultimately yielding an actual color distribution that reflects the color characteristics of the current layer. For example, if the current layer's color image is RGB and the target color space is HSV, converting RGB data to HSV data simplifies the quantitative analysis of color features when determining actual color differences, based on a unified data structure. As another example, if the target color space is LAB, color conversion not only makes the final determined color distribution more closely match human visual perception but also helps eliminate differences in device color gamuts, ensuring the accuracy and precision of subsequent actual color difference determination.

[0062] In some examples, to reduce errors and improve the reliability of the actual color distribution, when performing color space conversion on the current layer color image, the pixel values ​​of its pixels can be normalized first (e.g., mapping the 0-255 pixel values ​​in the RGB space to the 0-1 range) to eliminate the influence of differences in pixel value ranges on conversion accuracy. Then, based on the coordinate mapping relationship between the target color space and the original color space, the color space conversion is completed by combining the normalized pixel values, thereby obtaining the actual color distribution result that conforms to the target color space, which can ensure the consistency of color feature transmission in different color spaces.

[0063] Step d2: Compare the actual color distribution with the standard color distribution to obtain the actual color difference.

[0064] A standard color distribution can be understood as a reference for the actual color distribution of materials. By comparing the actual color distribution of the materials to be sorted with the standard color distribution, the color difference between the two can be clearly identified based on the actual colors obtained.

[0065] By measuring the actual color difference, the purity difference of the target component between the material to be sorted and the reference material can be reflected from the perspective of color. This allows for full utilization of the color features in color images, making the process of judging local purity more convenient and efficient. For example, the smaller the actual color difference value, the smaller the purity difference of the target component between the material to be sorted and the reference material; conversely, the larger the actual color difference value, the greater the purity difference of the target component between the material to be sorted and the reference material.

[0066] Step d3: Based on the actual color difference, determine the local purity of the target component in the current layer of the material to be sorted through the second correspondence relationship.

[0067] To ensure the accuracy of local purity determination, the purity value corresponding to different color difference ranges is determined through the second correspondence relationship. Then, based on the obtained actual color difference, the local purity of the target component of the material to be sorted in the current layer is determined.

[0068] For example, the second correspondence can be expressed by the following formula: ;in, Indicates actual color difference. This indicates the corresponding purity value.

[0069] In some embodiments, the material sorting method provided in this disclosure may further include: updating the first correspondence relationship based on the sorting results of the material to be sorted. That is, to ensure material sorting performance, during the actual material sorting process, the first correspondence relationship can be corrected according to the actual sorting results of the material to be sorted, so that the correspondence between X-ray transmittance, the interval of the fixed-distance geometric parameter range, and the content of the target component is more accurate and reliable, ensuring the reliability of the determination of the local content of the stratified components. Furthermore, through the automatic updating of the first correspondence relationship, the material sorting method provided in this disclosure can be applied to different working conditions, ensuring long-term stable operation and improving profitability.

[0070] For example, regarding the relationship between height and ray transmittance: Where k represents the calibration coefficient. This represents the ray transmittance of each pixel. By comparing the difference between the actual sorting result determined by this correspondence and the theoretical sorting result, the first correspondence can be corrected, that is, k can be recalibrated. Subsequently, when determining the height of the material to be sorted based on the updated first correspondence, the determined result can be more in line with the actual physical height of the material to be sorted. This helps to ensure the accuracy of the layer height boundary division, improve the calculation accuracy of the local content of the layer, and ultimately ensure the accuracy and reliability of the determination of the total content of the target component, thereby improving the overall performance of material sorting.

[0071] In some optional application scenarios, such as Figure 3 As shown, the material identification process can be as follows: Obtain the grayscale image of the ray and the corresponding RGB acquired image.

[0072] Preprocessing is performed on the ray grayscale image and the material color image. For example, for the ray grayscale image, outliers are removed, and contour recognition is performed through noise filtering and threshold segmentation to obtain a reliable ray contour image of the material to be sorted. For the RGB acquired image, invalid value filtering, white balance calibration, noise reduction, and normalization are performed to improve the quality and data reliability of the material color image.

[0073] Based on the mapping relationship between the grayscale ray image and the color image of the material, the color image corresponding to the ray contour image is determined. That is, according to the location of the ray contour image, the image region in the material color image corresponding to the material to be sorted is determined, and then the color image of the material to be sorted is obtained through image segmentation.

[0074] Based on the ray transmittance of each pixel in the color image and the target geometric dimension, the local content of the target component in the material to be sorted is determined layer by layer through the first correspondence.

[0075] Based on the comparison between the actual color distribution and the standard color distribution of each layer of color image, the local purity of the target component in the material to be sorted is determined layer by layer through the second correspondence.

[0076] Based on the local content and corresponding local purity of each layer, the local component content of the target component in the material to be sorted is determined layer by layer, and the total content of the target component in the material to be sorted is obtained.

[0077] The materials to be sorted are sorted based on the total content of the target components in the materials to be sorted.

[0078] In other alternative application scenarios, the materials to be sorted are sorted based on the total content, including: determining the effective layer ratio based on the local component content of each layer and the target threshold, and combining the total content to determine the category of the materials to be sorted.

[0079] In some alternative application scenarios, the first correspondence can be updated based on the sorting results of the materials to be sorted. The automatic updating of the first correspondence enables the material sorting method provided in this disclosure to be applicable to different operating conditions, ensuring long-term stable operation and improving profitability.

[0080] Based on the same inventive concept, this disclosure also provides a material sorting device. For example... Figure 4 As shown, the material sorting device 300 may include: The acquisition module 310 is used to acquire the ray profile diagram and the corresponding color image of the material to be sorted; The first processing module 320 is used to determine the local content of the target component in the material to be sorted layer by layer based on the ray transmittance of each pixel in the color image and the target geometric dimension, through a preset first correspondence relationship. The first correspondence relationship is used to characterize the relationship between ray transmittance, the interval of the fixed-distance geometric parameter, and the content of the target component. The second processing module 330 is used to determine the local purity of the target component in the material to be sorted layer by layer based on the comparison results between the actual color distribution and the standard color distribution of each layer of color image and through the second correspondence relationship. The second correspondence relationship is used to characterize the relationship between color difference and the purity of the target component. The third processing module 340 is used to determine the total content of the target component in the material to be sorted based on the local content and corresponding local purity of each layer. The fourth processing module 350 is used to sort the materials to be sorted based on the total content.

[0081] In some embodiments, the third processing module 340 may include: a first determining unit, configured to determine the local component content of the target component in each layer based on the local content and the corresponding local purity of each layer; a second determining unit, configured to determine the local volume of the material in each layer based on multiple pixels in each layer and a fixed distance geometric parameter range; and a third determining unit, configured to determine the total content of the target component in the material to be sorted based on the local volume of the material in each layer and the corresponding local component content.

[0082] In some embodiments, the fourth processing module 350 may include: a fourth determining unit, configured to determine the number of effective layers based on the local component content of each layer and a target threshold; a fifth determining unit, configured to determine the percentage of effective layers based on the total number of layers and the number of effective layers of the material to be sorted; a category determining unit, configured to determine the category of the material to be sorted based on the total content and the percentage of effective layers; and a sorting unit, configured to sort the material to be sorted according to the category.

[0083] In some embodiments, the first processing module 320 may include: a sixth determining unit, configured to determine the range of target geometric parameters of the material to be sorted in the target geometric dimension based on the ray transmittance of each pixel in the color image; a seventh determining unit, configured to determine the minimum and maximum geometric parameters of each layer according to the interval of the geometric parameters; and an eighth determining unit, configured to determine the local content of the target component in the material to be sorted layer by layer according to the minimum and maximum geometric parameters of each layer, the ray transmittance of each pixel, and the first correspondence.

[0084] In some embodiments, the geometric parameter is a height value; the eighth determining unit includes: a first execution unit, configured to determine a plurality of target pixels corresponding to the current layer based on the minimum height value and the maximum height value of the current layer; a second execution unit, configured to determine the first ray transmittance corresponding to the minimum height value, the second ray transmittance corresponding to the maximum height value, and the neighborhood content of the target component between the minimum height value and the maximum height value through a first correspondence; and a third execution unit, configured to determine the local content of the target component of the material to be sorted in the current layer based on the first ray transmittance, the second ray transmittance, the neighborhood content, the ray transmittance of each target pixel, and the corresponding height value.

[0085] In some embodiments, the second processing module 330 may include: a ninth determining unit, used to determine the actual color distribution of the current layer based on the processing result after converting the color space of the current layer color image; a comparison unit, used to compare the actual color distribution with the standard color distribution to obtain the actual color difference; and a purity determining unit, used to determine the local purity of the target component of the material to be sorted in the current layer according to the actual color difference and through a second correspondence relationship.

[0086] In some embodiments, the material sorting device may further include: an update module, used to update the first correspondence based on the sorting result of the material to be sorted.

[0087] Regarding the material sorting device in the above embodiments, the specific way in which each module performs its operation has been described in detail in the embodiments related to the method, and will not be elaborated here.

[0088] Based on the same inventive concept, this disclosure also provides a material sorting device, including: a memory and a processor, the memory and the processor being communicatively connected to each other, the memory storing computer instructions, and the processor executing the computer instructions to perform the material sorting method of any of the foregoing embodiments.

[0089] Based on the same inventive concept, this disclosure also provides a computer-readable storage medium storing a program for performing the material sorting method of any of the foregoing embodiments.

[0090] This disclosure uses specific terms to describe embodiments of the present disclosure. Terms such as "an embodiment," "one embodiment," and / or "some embodiments" refer to a particular feature, structure, or characteristic associated with at least one embodiment of the present disclosure. Therefore, it should be emphasized and noted that references to "an embodiment," "one embodiment," or "an alternative embodiment" in different locations throughout this specification do not necessarily refer to the same embodiment. Furthermore, certain features, structures, or characteristics in one or more embodiments of the present disclosure can be appropriately combined.

[0091] In the context of this disclosure, unless the context clearly indicates otherwise, the words "a," "an," "an," and / or "the" do not specifically refer to the singular and may also include the plural. Generally speaking, the terms "comprising" and "including" only indicate the inclusion of expressly identified steps and elements, which do not constitute an exclusive list, and the method or apparatus may also include other steps or elements.

[0092] Similarly, it should be noted that, in order to simplify the description of this disclosure and thus aid in the understanding of one or more embodiments, the foregoing description of embodiments of this disclosure may sometimes combine multiple features into a single embodiment, drawing, or description thereof. However, this disclosure method does not imply that the subject matter of this disclosure requires more features than the features claimed. In fact, the embodiments contain fewer features than all the features of the single embodiments disclosed above.

[0093] The basic concepts have been described above. It is obvious that the above disclosure is merely illustrative and does not constitute a limitation of this disclosure. Although not explicitly stated herein, various modifications, improvements, and corrections may be made to this disclosure by those skilled in the art. Such modifications, improvements, and corrections are suggested in this disclosure and therefore remain within the spirit and scope of the embodiments of this disclosure.

Claims

1. A material sorting method, comprising: Obtain the ray profile diagram and corresponding color image of the material to be sorted; Based on the ray transmittance of each pixel in the color image and the target geometric dimension, the local content of the target component in the material to be sorted is determined layer by layer through a preset first correspondence relationship. The first correspondence relationship is used to characterize the relationship between ray transmittance, the interval of the fixed-distance geometric parameter, and the content of the target component. Based on the comparison results between the actual color distribution and the standard color distribution of the color image at each layer, the local purity of the target component in the material to be sorted is determined layer by layer through the second correspondence relationship, wherein the second correspondence relationship is used to characterize the relationship between color difference and the purity of the target component; Based on the local content and corresponding local purity of each layer, the total content of the target component in the material to be sorted is determined; Based on the total content, the materials to be sorted are sorted.

2. The method of claim 1, wherein, The determination of the total content of the target component in the material to be sorted, based on the local content and corresponding local purity of each layer, includes: Based on the local content and corresponding local purity of each layer, the local component content of the target component in each layer is determined; The local volume of the material in each layer is determined based on multiple pixels in each layer and the interval of the fixed-distance geometric parameters. Based on the local volume of the material in each layer and the corresponding local component content, the total content of the target component in the material to be sorted is determined.

3. The method of claim 2, wherein, The process of sorting the materials to be sorted based on the total content includes: The number of effective layers is determined based on the local component content and target threshold of each layer; The percentage of effective layers is determined based on the total number of layers of the material to be sorted and the number of effective layers. Based on the total content and the effective layer ratio, the category of the material to be sorted is determined; The materials to be sorted are sorted according to the categories.

4. The material sorting method according to claim 1, wherein, The step of determining the local content of the target component in the material to be sorted, layer by layer, based on the ray transmittance of each pixel in the color image and the target geometric dimension, through a preset first correspondence, includes: Based on the ray transmittance of each pixel in the ray profile map, the range of target geometric parameters of the material to be sorted in the target geometric dimension is determined. Based on the specified interval of geometric parameters, determine the minimum and maximum geometric parameters for each layer; Based on the minimum and maximum geometric parameters of each layer, the ray transmittance of each pixel, and the first correspondence, the local content of the target component in the material to be sorted is determined layer by layer.

5. The material sorting method according to claim 4, wherein, The geometric parameter is a height value; the step of determining the local content of the target component in the material to be sorted layer by layer based on the minimum and maximum geometric parameters of each layer, the ray transmittance of each pixel, and the first correspondence includes: Based on the minimum and maximum height values ​​of the current layer, determine multiple target pixels corresponding to the current layer; Based on the first correspondence, the first ray transmittance corresponding to the minimum height value, the second ray transmittance corresponding to the maximum height value, and the neighborhood content of the target component between the minimum height value and the maximum height value are determined. Based on the first ray transmittance, the second ray transmittance, the neighborhood content, and the ray transmittance and corresponding height value of each target pixel, the local content of the target component of the material to be sorted in the current layer is determined.

6. The material sorting method according to claim 1, wherein, The comparison between the actual color distribution and the standard color distribution of each layer of the color image, based on a second correspondence, determines the local purity of the target component in the material to be sorted layer by layer, including: Based on the processing result of the color image of the current layer after color space conversion, the actual color distribution of the current layer is determined; The actual color difference is obtained by comparing the actual color distribution with the standard color distribution; Based on the actual color difference, the local purity of the target component of the material to be sorted in the current layer is determined through the second correspondence.

7. The material sorting method according to claim 1, wherein, The method further includes: Based on the sorting results of the materials to be sorted, update the first correspondence.

8. A material sorting device, comprising: The acquisition module is used to acquire the ray profile diagram and corresponding color image of the material to be sorted; The first processing module is used to determine the local content of the target component in the material to be sorted by layer based on the ray transmittance of each pixel in the color image and the target geometric dimension, through a preset first correspondence relationship. The first correspondence relationship is used to characterize the relationship between ray transmittance, the interval of the fixed-distance geometric parameter, and the content of the target component. The second processing module is used to determine the local purity of the target component in the material to be sorted layer by layer based on the comparison results between the actual color distribution and the standard color distribution of the color image at each layer, through a second correspondence relationship. The second correspondence relationship is used to characterize the relationship between color difference and the purity of the target component. The third processing module is used to determine the total content of the target component in the material to be sorted based on the local content and the corresponding local purity of each layer. The fourth processing module is used to sort the materials to be sorted based on the total content.

9. A material sorting device, comprising: A memory and a processor are interconnected, the memory storing computer instructions, and the processor executing the computer instructions to perform the material sorting method according to any one of claims 1 to 8.

10. A computer-readable storage medium storing computer instructions for causing a computer to perform the material sorting method according to any one of claims 1 to 8.

Images