A method and system for three-dimensional reconstruction of a carload of ore x-ray image

Abstract

The application relates to the technical field of X-ray image processing, in particular to a carriage ore X-ray image three-dimensional reconstruction method and system, which specifically comprises the following steps: determining each connected domain by analyzing the distribution of pixel gray values in the X-ray image of coal gangue in the carriage; determining the threshold adjustment coefficient of each connected domain based on the internal and external gray scale distribution of each connected domain, the area in the connected domain and the image noise estimation result, so as to set the threshold adjustment coefficient of each pixel in the X-ray image, and then determine the contrast limit threshold of each image sub-block in the limited contrast adaptive histogram equalization algorithm, so as to perform image enhancement on the X-ray image; constructing a three-dimensional reconstruction model of coal gangue in the carriage through the enhanced images of all X-ray images; the profile edge information of the blocky coal gangue in the X-ray image is enhanced, the noise information and the profile edge information of the coal powder and ore particles in the X-ray image are suppressed, and the accuracy of the final three-dimensional reconstruction model of the blocky coal gangue in the carriage is improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of X-ray image processing, in particular to a car ore X-ray image three-dimensional reconstruction method and system. BACKGROUND

[0002] In coal mine production, accurately and quickly measuring the volume of in-out warehouse and out-of-warehouse ore piles plays an important role in cost accounting, benefit evaluation and management optimization in the production process of coal mines, and the three-dimensional reconstruction technology based on X-ray image can quickly and non-destructively construct a three-dimensional model of the ore in the car, so as to calculate the volume of the ore in the car.

[0003] In the three-dimensional reconstruction technology based on X-ray image, due to the problems of high noise, overexposure or underexposure of the imaging device in the X-ray imaging system, the detail information in the collected X-ray image will be submerged, so the existing method usually performs image enhancement processing on the collected X-ray image to enhance the detail information in the X-ray image, and the limited contrast adaptive histogram equalization algorithm is a commonly used X-ray image enhancement algorithm, however, the threshold value in the traditional limited contrast adaptive histogram equalization algorithm is usually a fixed value set by experience, such as a crack defect detection method for industrial weld X-ray image in a publication CN113436168B, which uses a pre-set limited contrast threshold to perform limited contrast adaptive histogram equalization processing on the original X-ray image of the weld to be detected. This fixed threshold set by experience is difficult to adapt to the needs of different X-ray images and application scenarios, affecting the image enhancement effect of the algorithm, and further reducing the precision of the three-dimensional model of the ore in the car finally constructed. SUMMARY

[0004] In order to solve the above technical problems, the purpose of the present application is to provide a car ore X-ray image three-dimensional reconstruction method and system, and the technical solution adopted is as follows:

[0005] In the first aspect, the present application provides a car ore X-ray image three-dimensional reconstruction method, which comprises the following steps:

[0006] Collecting X-ray images of coal gangue in the car from each direction;

[0007] For each X-ray image, the distribution of pixel gray value in the X-ray image is counted to divide each gray value interval of the X-ray image; the position distribution of the pixel points in each gray value interval in the X-ray image is determined to determine each connected domain in the X-ray image; and the coal gangue characteristic value of each connected domain is determined based on the area and gray value distribution characteristics of each connected domain.

[0008] The gray value characteristic value of each connected domain is determined based on the proximity of the gray values inside and outside the contour of each connected domain.

[0009] According to the noise estimation result of the image in each connected domain, the threshold adjustment coefficient of each connected domain is determined in combination with the coal gangue characteristic value and the gray characteristic value, so as to set the threshold adjustment coefficient of each pixel point in the X-ray image;

[0010] Each image sub-block is obtained by dividing the X-ray image, and the contrast limit threshold of each image sub-block in the limited contrast adaptive histogram equalization algorithm is determined based on the threshold adjustment coefficient of the pixel point in each image sub-block, so as to perform image enhancement on the X-ray image; and a three-dimensional reconstruction model of the coal gangue in the carriage is constructed through the enhanced images of all X-ray images.

[0011] In one embodiment, the process of obtaining each gray value interval is as follows:

[0012] A segmentation threshold of the gray value of the pixel point in the X-ray image is obtained, and a gray histogram is drawn for all pixel points with a gray value greater than the segmentation threshold in the X-ray image; the data in the gray histogram is curve-fitted, and each wave peak in the fitted curve is determined by the starting point and the ending point to determine each gray value interval.

[0013] In one embodiment, the process of obtaining the coal gangue characteristic value is as follows:

[0014] The mean value of the gray value of all pixel points in each connected domain in the X-ray image and the number of pixel points in each connected domain are obtained, and the coal gangue characteristic value of each connected domain is in a positive correlation with the mean value of the gray value and the number of pixel points.

[0015] In one embodiment, the coal gangue characteristic value is the mean value of the normalized value of the mean value of the gray value and the normalized value of the number of pixel points in the connected domain.

[0016] In one embodiment, the process of obtaining the gray characteristic value is as follows:

[0017] The circumscribed rectangle of any connected domain in the X-ray image is obtained, the difference between the mean value of the gray value of the pixel points outside the connected domain and the pixel points in the connected domain in the circumscribed rectangle is calculated, and the gray characteristic value of the any connected domain is in a negative correlation with the difference.

[0018] In one embodiment, the gray characteristic value is the reciprocal of the sum of the difference and a preset minimum positive number.

[0019] In one embodiment, the expression of the threshold adjustment coefficient of each connected domain is as follows:

[0020] In the formula, S1(d), S2(d) and S3(d) respectively represent the normalized value of the coal gangue characteristic value, the normalized value of the gray characteristic value and the normalized value of the noise estimation result of the connected domain d. is a preset minimum positive number.

[0021] In one embodiment, the threshold adjustment coefficient of each pixel point is obtained by:

[0022] The threshold adjustment coefficient of each connected domain in the X-ray image is set as the threshold adjustment coefficient of each pixel point in the connected domain, and the threshold adjustment coefficient of the pixel point outside all connected domains in the X-ray image is set as 0.

[0023] In one embodiment, the contrast limited threshold is obtained by:

[0024] The mean value of the threshold adjustment coefficients of all pixel points in each image sub-block of the X-ray image is obtained, and the contrast limited threshold of each image sub-block in the limited contrast adaptive histogram equalization algorithm is determined.

[0025] In a second aspect, the embodiments of the present application further provide a car ore X-ray image three-dimensional reconstruction system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, and the processor implements the steps of the method of any one of the above aspects when executing the computer program.

[0026] The embodiments of the present application have at least the following beneficial effects:

[0027] The present application can effectively distinguish the corresponding image regions of the blocky coal gangue, coal powder and ore particles in the target car in the collected X-ray image by constructing the coal gangue feature value through analyzing the image distribution characteristics of the blocky coal gangue, coal powder and ore particles in the target car in the collected X-ray image, and can enhance the local contrast of the blocky coal gangue contour edge in the X-ray image and reduce the local contrast of the coal powder and ore particle contour edge in the subsequent image enhancement process, thereby achieving the effect of suppressing the contour edge information of the coal powder and ore particles in the X-ray image.

[0028] The contrast of the gray scale value inside and outside the contour edge of the lump coal gangue in the target car in the collected X-ray image and the interference degree of noise in different regions of the image are analyzed to construct the gray scale characteristic value and the noise estimation value, and the threshold adjustment coefficient constructed based on the coal gangue characteristic value, the gray scale characteristic value and the noise estimation value is used to select the appropriate contrast limit threshold for each image sub-block in the collected X-ray image, which can effectively enhance the contour edge information of the lump coal gangue in the X-ray image and suppress the noise information and the contour edge information of the coal powder and the ore particles in the X-ray image, and then the coal powder and the ore particles in the car and the noise information in the image can be avoided to form false edges in the obtained edge image when the edge image of the lump coal gangue is extracted from the subsequent X-ray image, so that the accuracy of the final three-dimensional reconstruction model of the lump coal gangue in the car is improved. BRIEF DESCRIPTION OF DRAWINGS

[0029] In order to more clearly illustrate the technical solutions and advantages of the embodiments of the present application or the prior art, the drawings needed to be used in the embodiments or the prior art description will be briefly introduced. Obviously, the drawings in the following description are only some embodiments of the present application, and other drawings can be obtained by those skilled in the art without creative labor.

[0030] Figure 1 A step flow chart of a car ore X-ray image three-dimensional reconstruction method provided by an embodiment of the present application is shown in the figure.

[0031] Figure 2 An acquisition process diagram for each gray scale value interval is shown in the figure. DETAILED DESCRIPTION

[0032] In order to further illustrate the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the specific embodiments, structure, features and effects of the car ore X-ray image three-dimensional reconstruction method and system according to the present application are described in detail as follows. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0033] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which the present application belongs.

[0034] The specific scheme of the car ore X-ray image three-dimensional reconstruction method and system provided by the present application is described in detail below with reference to the drawings.

[0035] Please refer to Figure 1Fig. 1 shows a flow chart of a method for three-dimensional reconstruction of a wagon ore X-ray image according to an embodiment of the present application, which comprises the following steps:

[0036] Step S1, X-ray images of coal gangue in the wagon are collected from each direction.

[0037] In this application, taking the three-dimensional reconstruction of blocky coal gangue transported by the wagon in coal mine production as an example, the target wagon is placed on a rotating table, and the rotating table is rotated at an interval of 10° for one revolution, and an X-ray image set E1 composed of X-ray images collected at different rotation angles of the target wagon is obtained by using an industrial detection X-ray machine. The X-ray image set contains a total of 36 X-ray images.

[0038] Each X-ray image in the X-ray image set E1 is respectively subjected to denoising processing by using an image denoising algorithm, so as to reduce the influence of noise interference on the subsequent processing of each X-ray image in the image collection process, and an X-ray image set E2 after image denoising processing is obtained. The image denoising algorithm can be a Gaussian filter algorithm, a median filter algorithm, or a bilateral filter algorithm. In this application, the Gaussian filter algorithm is selected, wherein the Gaussian filter algorithm is a known technology, and the specific process is not described in detail.

[0039] Step S2, for each X-ray image, the distribution of pixel gray value in the X-ray image is counted to divide each gray value interval of the X-ray image; each connected domain in the X-ray image is determined by the position distribution of the pixel points in each gray value interval; and the coal gangue characteristic value of each connected domain is determined based on the area and gray value distribution characteristics of each connected domain.

[0040] (1) Since the three-dimensional reconstruction technology based on X-ray images mainly utilizes the contour edge information of the object to be reconstructed in the X-ray image to realize the three-dimensional reconstruction of the object to be reconstructed, in the process of coal mine production, coal dust and ore particles are usually carried in the wagon when the wagon is used to transport blocky coal gangue, and the coal dust and ore particles in the wagon will form pseudo-edges when the contour edge information of the blocky coal gangue in the collected X-ray image is extracted, thereby affecting the final three-dimensional reconstruction accuracy of the blocky coal gangue in the wagon. Therefore, in order to reduce the influence of the coal dust and ore particles in the wagon on the subsequent extraction of the contour edge information of the blocky coal gangue in the collected X-ray image, the following processing is performed.

[0041] Since the regions where the blocky coal gangue, coal dust, and ore particles in the wagon are located have stronger X-ray absorption capacity than the regions in the wagon that do not contain minerals, the regions where the blocky coal gangue, coal dust, and ore particles in the wagon are located have higher gray values in the collected X-ray image.

[0042] Therefore, taking any one of the X-ray images A in the X-ray image set E2 as an example, the gray values of all the pixel points in the X-ray image A are taken as the input of the maximum inter-class variance algorithm, and the output is a gray value threshold. A set composed of all the pixel points with a gray value greater than the gray value threshold is denoted as a first pixel point set B1 of the X-ray image A, which represents the region where the bulk coal gangue, coal powder and ore particles in the carriage are located in the set composed of all the pixel points corresponding to the X-ray image A. The maximum inter-class variance algorithm is a known technology, and the specific process will not be described again.

[0043] It should be noted that for the acquisition of the gray value threshold, the present application only provides a threshold segmentation method. There are many existing threshold segmentation methods, and the implementer can also use other threshold segmentation algorithms to acquire the gray value threshold. The present application does not make specific limitations.

[0044] A gray histogram C1 of all the gray values corresponding to all the pixel points in the first pixel point set B1 in the X-ray image A is extracted, which is used to evaluate the gray value distribution of the region where the bulk coal gangue, coal powder and ore particles in the carriage are located in all the pixel points corresponding to the X-ray image A. The extraction of the gray histogram is a known technology, and the specific process will not be described again.

[0045] (2) Since different substances have different X-ray attenuation abilities, different minerals in the carriage are concentrated in different gray value intervals in the collected X-ray images. Therefore, taking the gray values and their corresponding amplitudes in the gray histogram C1 as the horizontal and vertical coordinates, respectively, a curve fitting algorithm based on the least squares method is used to perform curve fitting on all the amplitudes in the gray histogram C1, to obtain a fitting curve L of the gray histogram C1. All the peak points in the fitting curve L are extracted, and the derivative detection method is used to obtain the starting point and the ending point of each peak in the fitting curve L, and the gray value interval formed by the starting point and the ending point of each peak point in the fitting curve L is obtained. A set composed of all the gray value intervals is denoted as a gray value interval set of the X-ray image A, which represents a set composed of all the gray value intervals where all the minerals in the carriage are concentrated in the collected X-ray images. The curve fitting algorithm based on the least squares method, the extraction of the curve peak and the derivative detection method are all known technologies, and the specific process will not be described again.

[0046] All the pixel points in the X-ray image A with a gray value in each of the gray value intervals in the gray value interval set are subjected to connected component analysis, respectively. A set composed of all the connected components is denoted as a connected component set D of the X-ray image A, which represents a set composed of all the connected components corresponding to the image regions where all the minerals in the carriage are located in the collected X-ray images. The connected component analysis of the image is a known technology, and the specific process will not be described again.

[0047] (3) Since the lumpy coal gangue and ore particles in the carriage usually contain metal elements, they have a higher density than coal powder, which is a carbonaceous material. As a result, the coal powder in the carriage has a smaller gray value in the acquired X-ray image compared to the lumpy coal gangue and ore particles. Furthermore, the lumpy coal gangue area in the acquired X-ray image usually has a larger area than the coal powder and ore particle area. Therefore, the larger the overall gray value of the connected region in X-ray image A and the larger the area of ​​the connected region, the more likely the connected region is to be the connected region corresponding to the image area where the lumpy coal gangue in the carriage is located in X-ray image A.

[0048] Based on the above analysis, taking any connected region d in the set of connected regions D as an example, the mean gray value of all pixels corresponding to the connected region d in the X-ray image A is calculated and the number of all pixels is counted. The mean and the number are respectively used as the average gray value and the area of ​​the connected region d.

[0049] (4) The mean gray value and area of ​​all connected components in the set of connected components D are normalized using the Min-Max normalization method. Taking connected component d as an example, the mean of the normalized results of the mean gray value and area of ​​connected component d is recorded as the coal gangue feature value of connected component d. This is used to evaluate whether connected component d is the connected component corresponding to the image region where the blocky coal gangue in the carriage is located in the X-ray image A. The larger the coal gangue feature value, the more likely connected component d is to be the connected component corresponding to the image region where the blocky coal gangue in the carriage is located in the X-ray image A. The Min-Max normalization method is a well-known technique, and the specific process will not be described in detail.

[0050] In the contrast-limited adaptive histogram equalization algorithm, a larger contrast limit threshold will improve the local contrast effect of the image, thereby enhancing the local contour edges, texture and other details of the image, but will amplify the noise in the image. On the other hand, a smaller contrast limit threshold will reduce the local contrast effect of the image, thereby suppressing the local contour edges, texture and other details of the image, but will suppress the noise in the image.

[0051] Therefore, when using the contrast-limited adaptive histogram equalization algorithm to enhance X-ray image A, in order to enhance the local contrast of the contour edges of the blocky coal gangue in the carriage in X-ray image A, and reduce the local contrast of the contour edges of the coal powder and ore particles in the carriage in X-ray image A, the smaller the coal gangue feature value of the connected region d, the smaller the contrast limit threshold corresponding to the image sub-block where the connected region d is located in X-ray image A should be in this algorithm, so as to achieve the effect of suppressing the contour edge information of the coal powder and ore particles in the carriage in the X-ray image after image enhancement.

[0052] Step S3: Determine the gray-level feature value of each connected component based on the similarity of gray-level values ​​inside and outside the contour of each connected component.

[0053] When the gray values ​​of the inner and outer edges of a certain block of coal gangue in the X-ray image are closer in the image region where it is located, the more difficult it is to detect the outline edge of the block of coal gangue in the X-ray image in subsequent edge extraction of the X-ray image. Therefore, in order to improve the accuracy of extracting the outline edge information of the block of coal gangue in the X-ray image, the contrast limit threshold corresponding to the image sub-block where the block of coal gangue is located in the X-ray image should be increased in the contrast-limited adaptive histogram equalization algorithm to enhance the local contrast of the outline edge of the block of coal gangue in the X-ray image.

[0054] Based on the above analysis, taking the connected component d as an example, we obtain the minimum bounding rectangle of the connected component d in the X-ray image A. We then calculate the mean of all gray values ​​of all pixels in the bounding rectangle that are not within the connected component d in the X-ray image A. Finally, we add a constant to the absolute value of the difference between the obtained mean and the average gray value of the connected component d. The reciprocal of is denoted as the gray-level feature value of the connected component d, used to evaluate the similarity of gray-level values ​​inside and outside the contour edge of the mineral corresponding to the connected component d in the X-ray image A within the image region. The larger the gray-level feature value, the greater the similarity, and the larger the contrast limit threshold corresponding to the image sub-block of the connected component d in the X-ray image A should be in the contrast-limited adaptive histogram equalization algorithm, so as to enhance the local contrast of the contour edge of the mineral corresponding to the connected component d in the X-ray image A within the image region. A constant is added here. This is to prevent the denominator from being 0, a constant. For positive numbers close to 0, in this application, the constant is... Set to 0.001. In other embodiments of this application, the implementer may set it according to the actual situation. The value of .

[0055] Step S4: Based on the noise estimation results of the image in each connected region, combined with the coal gangue feature value and the grayscale feature value, determine the threshold adjustment coefficient of each connected region, so as to set the threshold adjustment coefficient of each pixel in the X-ray image.

[0056] The noise estimation algorithm is used to estimate the noise of each connected component in the connected component set D, and the noise estimate value of each connected component is obtained. This is used to evaluate the degree of noise interference to the image information in each connected component. The larger the noise estimate value, the smaller the contrast limit threshold corresponding to the image sub-block of each connected component in the X-ray image A in the contrast-limited adaptive histogram equalization algorithm should be, so as to suppress the noise information carried by each connected component. The image noise estimation algorithm is a well-known technology, and the specific process will not be described in detail.

[0057] The Min-Max normalization method is used to normalize the coal gangue feature values, grayscale feature values, and noise estimates of all connected components in the set of connected components D, respectively. This yields the normalized results for the coal gangue feature values, grayscale feature values, and noise estimates of each connected component in the set of connected components D. The Min-Max normalization method is a well-known technique, and its specific process will not be elaborated upon here. Implementers may also use other normalization methods, and this application does not impose any specific restrictions.

[0058] Taking connected component d as an example, the threshold adjustment coefficient W(d) of connected component d is obtained. This coefficient is used to evaluate the magnitude of the contrast limitation threshold of the image sub-block containing connected component d in X-ray image A when performing image enhancement processing on X-ray image A using the contrast-limited adaptive histogram equalization algorithm. The calculation of the threshold adjustment coefficient W(d) is as follows:

[0059] In the formula, S1(d), S2(d), and S3(d) represent the normalized results of the coal gangue characteristic value, gray value, and noise estimate of the connected domain d, respectively. The constant is a preset minimum positive number used to prevent the denominator from being zero. In this application, the constant is... Set to 0.001; in other embodiments of this application, the implementer may set it according to the actual situation. value.

[0060] The larger the threshold adjustment coefficient W(d), the larger the contrast limit threshold of the image sub-block where the connected component d is located in the X-ray image A should be in the algorithm, where the threshold adjustment coefficient W(d) is between 0 and 1.

[0061] The threshold adjustment coefficient of each connected component in the connected component set D is used as the threshold adjustment coefficient of all pixels corresponding to each connected component in the X-ray image A, and the threshold adjustment coefficient of all pixels in the X-ray image A that do not belong to each connected component is assigned a value of 0.

[0062] Step S5: Divide the X-ray image into image sub-blocks, determine the contrast limit threshold of each image sub-block in the contrast-limited adaptive histogram equalization algorithm based on the threshold adjustment coefficient of the pixels in each image sub-block, so as to enhance the X-ray image; construct a three-dimensional reconstruction model of coal gangue in the carriage through the enhanced images of all X-ray images.

[0063] Taking X-ray image A as an example, X-ray image A is evenly divided into M image sub-blocks. In this application, M is set to 64. In other embodiments of this application, the implementer can set the value of M according to the actual situation. Since the contrast limit threshold in the contrast-limited adaptive histogram equalization algorithm is usually a floating-point number between 0 and 1, the rounded result of the average threshold adjustment coefficient of all pixels in each image sub-block of X-ray image A (retaining one decimal place in this application) is used as the contrast limit threshold value of each image sub-block in the contrast-limited adaptive histogram equalization algorithm. The contrast-limited adaptive histogram equalization algorithm is then used to perform subsequent processing on X-ray image A to complete the image enhancement processing of X-ray image A. The contrast-limited adaptive histogram equalization algorithm is a well-known technology and will not be described in detail here.

[0064] Using the same method as X-ray image A, each X-ray image in the X-ray image set E2 is subjected to image enhancement processing using the limited contrast adaptive histogram equalization algorithm, resulting in the image enhancement processing X-ray image set E3.

[0065] Edge images are extracted from each X-ray image in the X-ray image set E3 using the Canny edge detection algorithm. A three-dimensional reconstruction method based on spatial sculpting of multi-view X-ray two-dimensional images is then used to perform spatial sculpting processing on all the obtained edge images, resulting in a three-dimensional reconstruction model of the blocky coal gangue in the target carriage. The volume of the three-dimensional reconstruction model can be calculated using methods such as polygon mesh algorithms, boundary representation algorithms, and voxelization algorithms. This embodiment uses a polygon mesh algorithm to calculate the volume of the three-dimensional reconstruction model; however, other methods can be used, and this application does not impose specific limitations. This achieves the estimation of the volume of the blocky coal gangue in the target carriage. The Canny edge detection algorithm, the three-dimensional reconstruction method based on spatial sculpting of multi-view X-ray two-dimensional images, and the three-dimensional model volume calculation method are all well-known technologies and will not be elaborated further. The three-dimensional reconstruction of the blocky coal gangue in the target carriage using X-ray images is thus completed.

[0066] A schematic diagram illustrating the process of obtaining each grayscale value range is shown below. Figure 2 As shown.

[0067] Based on the same inventive concept as the above method, this application embodiment also provides a three-dimensional reconstruction system for X-ray images of ore in a train carriage, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of any one of the above-described methods for three-dimensional reconstruction of X-ray images of ore in a train carriage.

[0068] In summary, this application provides a method for three-dimensional reconstruction of X-ray images of ore in a train carriage. By analyzing the image distribution characteristics of blocky coal gangue, coal powder, and ore particles in the target carriage in the acquired X-ray images, coal gangue feature values ​​are constructed. This method can effectively distinguish the corresponding image regions of blocky coal gangue, coal powder, and ore particles in the target carriage in the acquired X-ray images. Consequently, in the subsequent image enhancement process, the local contrast of the contour edges of blocky coal gangue in the X-ray images can be enhanced while the local contrast of the contour edges of coal powder and ore particles can be reduced, thereby suppressing the contour edge information of coal powder and ore particles in the X-ray images.

[0069] This application analyzes the contrast of gray values ​​inside and outside the contour edges of the blocky coal gangue in the target carriage in the acquired X-ray image, as well as the degree of noise interference in different areas of the image, to construct gray-level feature values ​​and noise estimates respectively. Based on the coal gangue feature values, gray-level feature values, and noise estimates, a threshold adjustment coefficient is constructed to select an appropriate contrast limit threshold for each image sub-block in the acquired X-ray image. This effectively enhances the contour edge information of the blocky coal gangue in the X-ray image and suppresses noise information and contour edge information of coal powder and ore particles in the X-ray image. As a result, when extracting the edge image of the blocky coal gangue in subsequent X-ray images, it avoids the formation of false edges in the obtained edge image by coal powder and ore particles in the carriage and noise information in the image, thereby improving the accuracy of the final three-dimensional reconstruction model of the blocky coal gangue in the carriage.

[0070] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. Furthermore, the above description focuses on specific embodiments of this application. Additionally, the processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired results. In some implementations, multitasking and parallel processing are possible or may be advantageous.

[0071] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

[0072] The above description is only a preferred embodiment of this application and is not intended to limit this application. Any modifications, equivalent substitutions, improvements, etc., made within the principles of this application should be included within the protection scope of this application.

Claims

1. A method of three-dimensional reconstruction of a carload of ore X-ray images, characterized by, The method comprises the following steps: Collecting X-ray images of coal gangue in the carriage from all directions; For each X-ray image, the distribution of pixel gray value in the X-ray image is counted to divide each gray value interval of the X-ray image; each connected domain in the X-ray image is determined by the position distribution of the pixel in each gray value interval; the coal gangue characteristic value of each connected domain is determined based on the area and the gray value distribution characteristics of each connected domain; The gray characteristic value of each connected domain is determined based on the proximity of the gray values inside and outside the contour of each connected domain; According to the noise estimation result of the image in each connected domain, the coal gangue characteristic value and the gray characteristic value are combined to determine the threshold adjustment coefficient of each connected domain, so as to set the threshold adjustment coefficient of each pixel in the X-ray image; The X-ray image is divided to obtain each image sub-block, and the contrast limit threshold in the limited contrast adaptive histogram equalization algorithm of each image sub-block is determined based on the threshold adjustment coefficient of the pixel in each image sub-block, so as to perform image enhancement on the X-ray image; and a three-dimensional reconstruction model of the coal gangue in the carriage is constructed by using the enhanced images of all X-ray images. The expression of the threshold adjustment coefficient of each connected domain is: In the formula, S1(d), S2(d), and S3(d) respectively represent the coal gangue characteristic value normalized value, the gray characteristic value normalized value, and the noise estimation result normalized value of the connected domain d; is a preset minimum positive number; The process of obtaining the threshold adjustment coefficient of each pixel is: The threshold adjustment coefficient of each connected domain in the X-ray image is taken as the threshold adjustment coefficient of each pixel in each connected domain, and the threshold adjustment coefficient of all pixels outside each connected domain in the X-ray image is assigned as 0.

2. The method of claim 1, wherein the method further comprises: The process of obtaining each gray value interval is: A segmentation threshold of the pixel gray value in the X-ray image is obtained, and a gray histogram of all pixels with a gray value greater than the segmentation threshold in the X-ray image is drawn; the data in the gray histogram is curve-fitted, and the starting point and ending point of each wave peak in the fitted curve are used to determine each gray value interval.

3. The method of claim 1, wherein the method further comprises: The process of obtaining the coal gangue characteristic value is: The mean value of the gray value of all pixels in each connected domain in the X-ray image and the number of pixels in each connected domain are obtained, and the coal gangue characteristic value of each connected domain is positively correlated with the mean value of the gray value and the number of pixels.

4. The method of claim 3, wherein the method further comprises: The coal gangue characteristic value is the mean value of the normalized value of the mean value of the gray value and the normalized value of the number of pixels in the connected domain.

5. The method for three-dimensional reconstruction of X-ray images of ore in a train carriage as described in claim 1, characterized in that, The process of obtaining the gray characteristic value is: The circumscribed rectangle of any connected domain in the X-ray image is obtained, the difference between the mean values of the gray values of the pixels inside and outside the connected domain in the circumscribed rectangle is calculated, and the gray characteristic value of the any connected domain is negatively correlated with the difference.

6. The method of claim 5, wherein the method further comprises: The gray characteristic value is the reciprocal of the sum of the difference and a preset minimum positive number.

7. The method of claim 1, wherein the method further comprises: determining a three-dimensional model of the ore in the car based on the X-ray image. The process of obtaining the contrast limit threshold is: ​ The mean value of the threshold adjustment coefficients of all pixels in each image sub-block of the X-ray image is obtained, and the contrast limit threshold in the limited contrast adaptive histogram equalization algorithm of each image sub-block is determined.

8. A system for three-dimensional reconstruction of a truckload of ore from X-ray images, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, The processor executes the computer program to realize the steps of the method according to any one of claims 1-7. The processor executes the computer program to realize the steps of the method according to any one of claims 1-7.

Images