Intelligent ore dressing method and device based on machine vision technology

Abstract

The application relates to the technical field of ore processing, and discloses an intelligent ore dressing method and device based on machine vision technology. The method uses an image collector to collect an ore image in real time; the ore image is identified to determine an ore boundary line in the ore image; the position information of a plurality of ore separators is identified in the ore image, and an adjustment line is determined according to the position information of the ore separators; the intersection of the ore boundary line and the adjustment line is obtained; the adjustment mode of each ore separator is determined according to the intersection, the position information of each ore separator and a preset deviation distance corresponding to each ore separator; and each ore separator is controlled to be adjusted to a corresponding target position through the corresponding adjustment mode. The application realizes real-time identification of the position of target ore through machine vision technology, automatic adjustment of the position and direction of the ore separator, full-automatic ore screening, no human intervention when the type or taste of the ore changes, improved work efficiency and improved identification accuracy of the target ore.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of ore processing, in particular to an intelligent ore dressing method and device based on machine vision technology. BACKGROUND

[0002] Due to the resource shortage problem of mineral resources (iron ore, non-ferrous ore) in the industrial chain, the intensity of mining minerals increases, and high-grade mineral resources are becoming lower and lower, so the requirements for existing ore dressing technology are also increasing.

[0003] At present, intelligent ore dressing technology mainly includes solid ore screening technology based on vision and gas blowing, and solid ore pre-selection technology based on X-ray detection. At present, for the ore slurry formed by dissolving the ore after crushing and grinding in water, the traditional ore slurry screening technology is still used, such as magnetic separation, chute, shaking table, centrifugal ore dressing, vibrating screen, screening machine, blanket machine, and washing machine. In these ore screening processes and the final waste throwing process, equal division or manual adjustment or manual setting of the position of the ore separator is adopted. However, when using this equal division or manual adjustment or manual setting method, whenever the type or taste of the ore changes, manual intervention is required, and real-time online adjustment cannot be realized, and the production efficiency and ore quality are difficult to further improve. SUMMARY

[0004] The present application provides an intelligent ore dressing method and device based on machine vision technology, which can realize full-automatic ore screening, without human intervention when the type or taste of the ore changes, improving the work efficiency and thus improving the identification accuracy of the target ore.

[0005] In order to solve the above technical problems, the present application provides an intelligent ore dressing method based on machine vision technology, comprising:

[0006] using an image collector to collect a mineral image in real time;

[0007] identifying the mineral image and determining a mineral boundary line in the mineral image; wherein the mineral boundary line is used to distinguish target ore and non-target ore;

[0008] identifying position information of a plurality of ore separators in the mineral image, and determining an adjustment line according to the position information of each ore separator;

[0009] obtaining the intersection of the mineral boundary line and the adjustment line;

[0010] determining the adjustment mode of each ore separator according to the intersection, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator;

[0011] The ore separators are controlled to adjust to corresponding target positions through corresponding adjustment modes.

[0012] Further, the ore map is identified to determine an ore boundary line in the ore map, in particular:

[0013] The ore map is subjected to picture equalization to determine a range of target colors in the ore map, and an ore boundary line is determined in combination with a preset minimum color band width; wherein the target colors include a preset target ore color and a non-target ore color;

[0014] Alternatively, the ore map is subjected to edge detection based on an edge detection algorithm, and an ore boundary line is determined according to an edge detection result; wherein the edge detection result includes a target ore edge and a non-target ore edge;

[0015] Alternatively, the ore map is subjected to boundary line detection based on a first target detection algorithm, and an ore boundary line is determined according to a boundary line detection result.

[0016] Further, the ore separator is a movable ore separating groove, the position information is a position coordinate of the ore separator, the position information of the plurality of ore separators is identified in the ore map, and an adjustment line is determined according to the position information of each ore separator, in particular:

[0017] The ore separator is identified in the ore map based on a second target detection algorithm to obtain position coordinates of the plurality of ore separators;

[0018] The adjustment line is determined based on the position coordinates of each ore separator.

[0019] Further, the ore separator is a rotating arm, the position information is a rotating fulcrum coordinate of the ore separator, the position information of the plurality of ore separators is identified in the ore map, and an adjustment line is determined according to the position information of each ore separator, in particular:

[0020] The ore separator is identified in the ore map based on a second target detection algorithm to obtain rotating fulcrum coordinates of the plurality of ore separators;

[0021] The adjustment line is determined based on the rotating fulcrum coordinates of each ore separator.

[0022] Further, the intersection of the ore boundary line and the adjustment line is obtained, in particular:

[0023] A plurality of boundary line coordinates are obtained on the ore boundary line;

[0024] Straight line fitting is performed based on the plurality of boundary line coordinates, and a fitted straight line is determined as a boundary line fitting straight line;

[0025] The intersection point of the fitting straight line of the boundary line and the adjustment line is determined as the intersection point of the ore boundary line and the adjustment line.

[0026] Further, a plurality of boundary line coordinates are obtained on the ore boundary line, specifically:

[0027] The coordinates of all points on the identified ore boundary line are obtained in the ore map, and the coordinates of all points on the ore boundary line are determined as boundary line coordinates.

[0028] Alternatively, according to a preset end determination method, a boundary line end range is determined on the ore boundary line, a plurality of end coordinates are obtained in the boundary line end range, and the plurality of end coordinates are determined as boundary line coordinates.

[0029] Alternatively, inflection point recognition is performed on the ore boundary line, the end direction of the ore boundary line is determined according to a preset end determination method, the inflection point closest to the end direction of the ore boundary line is determined as an end inflection point, and the coordinates of all points recognized in the end direction of the ore boundary line are determined as boundary line coordinates.

[0030] Further, when the ore separator is a movable ore separation groove, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, specifically:

[0031] The intersection point coordinates are obtained on the ore map;

[0032] According to the intersection point coordinates, the target position coordinates corresponding to each ore separator are obtained in combination with a preset deviation distance corresponding to each ore separator;

[0033] The adjustment mode of each ore separator is determined as moving each ore separator to the target position coordinates along the adjustment line.

[0034] Further, when the ore separator is a rotating arm, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, specifically:

[0035] The intersection point coordinates, the rotation fulcrum coordinates and the effective arm length, the first angle and the second angle of each ore separator are obtained on the ore map; the first angle is the included angle between the ore boundary line and the adjustment line; and the second angle is the included angle between each ore separator and the adjustment line;

[0036] In combination with a preset deviation distance corresponding to each ore separator, an auxiliary line parallel to the ore boundary line corresponding to each ore separator is constructed;

[0037] An auxiliary circle of each of the ore sorters is constructed with the effective arm length of each of the ore sorters as a radius and a rotation pivot as a center;

[0038] A target position of a rotating arm end of each of the ore sorters is determined in two intersection points of the auxiliary line and the auxiliary circle;

[0039] A target angle of a target included angle between each of the ore sorters and the adjustment line is determined according to the intersection point coordinates, the rotation pivot coordinates of each of the ore sorters, the effective arm length, and a first angle;

[0040] A rotation angle of each of the ore sorters is calculated according to the second angle and the target angle;

[0041] The adjustment mode of each of the ore sorters is determined as rotating the ore sorters by the rotation angle, so that the rotating arm end of each of the ore sorters is at the target position of the rotating arm end.

[0042] Further, the target angle of the target included angle between each of the ore sorters and the adjustment line is determined according to the intersection point coordinates, the rotation pivot coordinates of each of the ore sorters, the effective arm length, and a first angle, specifically:

[0043] A first distance between each of the ore sorters and the intersection point is calculated according to the intersection point coordinates and the rotation pivot coordinates of each of the ore sorters;

[0044] An intersection point between the auxiliary line and the adjustment line is determined as an auxiliary intersection point of each of the ore sorters;

[0045] A second distance between each of the ore sorters and the auxiliary intersection point is calculated according to the first distance between each of the ore sorters and the intersection point and a preset deviation distance corresponding to each of the ore sorters;

[0046] A third angle between a target line segment and the auxiliary line is calculated according to the effective arm length of each of the ore sorters, the first angle, and the second distance by using the sine theorem, wherein the target line segment is a line segment formed by connecting the target position of the rotating arm end of each of the ore sorters and the rotation pivot;

[0047] The target angle of the target included angle between each of the ore sorters and the adjustment line is calculated according to the first angle and the third angle by using the theorem of interior angles of a triangle.

[0048] The application provides an intelligent ore dressing method based on machine vision technology, which uses an image collector to collect a mine stone image in real time; identifies the mine stone image, and determines a mine stone boundary line in the mine stone image; identifies position information of a plurality of ore separators in the mine stone image, and determines an adjustment line according to the position information of the ore separators; obtains intersection points of the mine stone boundary line and the adjustment line; determines an adjustment mode of each ore separator according to the intersection points, the position information of the ore separators, and a preset deviation distance corresponding to each ore separator; and controls each ore separator to be adjusted to a corresponding target position through the corresponding adjustment mode. The application realizes automatic ore screening by using machine vision technology to identify the position of target ore in real time, automatically adjusting the position and direction of the ore separator, and improving the work efficiency and the identification accuracy of the target ore when the type or taste of the ore changes without human intervention.

[0049] Correspondingly, the application provides an intelligent ore dressing device based on machine vision technology, which comprises an acquisition module, a boundary line determination module, an adjustment line determination module, an intersection point acquisition module, an adjustment mode determination module and a control module.

[0050] The acquisition module is used to collect a mine stone image in real time by using an image collector.

[0051] The boundary line determination module is used to identify the mine stone image, and determine a mine stone boundary line in the mine stone image; wherein the mine stone boundary line is used to distinguish target ore and non-target ore.

[0052] The adjustment line determination module is used to identify position information of a plurality of ore separators in the mine stone image, and determine an adjustment line according to the position information of the ore separators.

[0053] The intersection point acquisition module is used to obtain intersection points of the mine stone boundary line and the adjustment line.

[0054] The adjustment mode determination module is used to determine an adjustment mode of each ore separator according to the intersection points, the position information of the ore separators, and a preset deviation distance corresponding to each ore separator.

[0055] The control module is used to control each ore separator to be adjusted to a corresponding target position through the corresponding adjustment mode.

[0056] The application provides an intelligent ore dressing device based on a machine vision technology, which is based on organic combination among modules, uses an image collector to collect a mine image in real time, identifies the mine image, determines a mine boundary line in the mine image, identifies position information of a plurality of ore separators in the mine image, determines an adjustment line according to the position information of the ore separators, obtains intersection points of the mine boundary line and the adjustment line, determines an adjustment mode of each ore separator according to the intersection points, the position information of the ore separators and a preset deviation distance corresponding to each ore separator, and controls each ore separator to be adjusted to a corresponding target position through the corresponding adjustment mode. BRIEF DESCRIPTION OF DRAWINGS

[0057] Figure 1 A flowchart of an embodiment of the intelligent ore dressing method based on the machine vision technology provided by the application;

[0058] Figure 2 An embodiment of the boundary line coordinate acquisition method provided by the application;

[0059] Figure 3 Another embodiment of the boundary line coordinate acquisition method provided by the application;

[0060] Figure 4 Still another embodiment of the boundary line coordinate acquisition method provided by the application;

[0061] Figure 5 An embodiment of the rotary arm adjustment method provided by the application;

[0062] Figure 6 An embodiment of the structure of the intelligent ore dressing device based on the machine vision technology provided by the application. DETAILED DESCRIPTION

[0063] The technical solutions in the embodiments of the application will be clearly and completely described below with reference to the drawings in the embodiments of the application. Obviously, the described embodiments are only a part of the embodiments of the application, but not all the embodiments of the application. Based on the embodiments in the application, all the other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the application.

[0064] The flowchart shown in the drawings is only an example and does not necessarily include all the contents and operations / steps, nor does it have to be executed in the order described. For example, some operations / steps can be further divided, combined or partially merged, so the actual execution order can be changed according to the actual situation.

[0065] Some embodiments of the present application will be described in detail below with reference to the accompanying drawings. The following examples and features in the examples can be combined with each other without conflict.

[0066] Example 1

[0067] Referring to Figure 1 is a flowchart of an embodiment of the intelligent beneficiation method based on machine vision technology provided by the present application, which includes steps 101 to 106, and each step is as follows:

[0068] Step 101: Real-time acquisition of ore map by image collector.

[0069] In the first embodiment of the present application, a camera or other image collector is used to real-time acquisition of ore map on beneficiation equipment during screening of ore or discarding of waste, which contains concentrate and waste, and a plurality of ore classifiers for classifying ore.

[0070] Step 102: Identifying the ore map and determining the ore boundary line in the ore map; wherein the ore boundary line is used to distinguish target ore and non-target ore.

[0071] Further, in the first embodiment of the present application, the ore map is identified and the ore boundary line is determined in the ore map, specifically:

[0072] The ore map is subjected to picture equalization to determine the range of target color in the ore map, and the ore boundary line is determined in combination with the preset minimum color bandwidth; wherein the target color includes the preset target ore color and non-target ore color;

[0073] Alternatively, the ore map is subjected to edge detection based on an edge detection algorithm, and the ore boundary line is determined according to the edge detection result; wherein the edge detection result includes the target ore edge and the non-target ore edge;

[0074] Alternatively, the ore map is subjected to boundary line detection based on a first target detection algorithm, and the ore boundary line is determined according to the boundary line detection result.

[0075] In the first embodiment of the present application, the determination of the ore boundary in the ore map can be realized by different algorithms. For example, the ore boundary can be determined based on color difference or contrast, the ore map is subjected to image equalization, at this time, different colors are displayed in the ore map, the different colors in the ore map are identified and classified by pre-setting the colors of the target ore and the non-target ore, the boundary of different colors is obtained, and the width of the ore boundary is determined based on the pre-set minimum color bandwidth, so as to determine the ore boundary. For another example, the ore map is subjected to edge detection based on an edge detection algorithm, the edges of the target ore and the non-target ore are identified, and the ore boundary is determined in combination with the two edges. The edge detection algorithm includes a first-order detection operator and a second-order detection operator, and the specific operators include a horizontal difference operator, a vertical difference operator, a Roberts cross operator, a Prewitt edge detection operator, a Sobel edge detection operator, a Canny edge detection operator, a Laplacian operator, etc. For another example, the ore map is subjected to boundary target detection based on a target detection algorithm, and the detection result is determined as the ore boundary. The target detection algorithm includes an R-CNN algorithm, a Faster R-CNN algorithm, a series of YOLO algorithms, and an SSD algorithm.

[0076] Step 103: identifying position information of a plurality of ore distributors in the ore map, and determining an adjustment line according to the position information of each ore distributor.

[0077] Further, in the first embodiment of the present application, when the ore distributor is a movable ore distribution groove, the position information is the position coordinates of the ore distributor, and the identification of the position information of a plurality of ore distributors in the ore map and the determination of the adjustment line according to the position information of each ore distributor are specifically as follows:

[0078] performing ore distributor identification in the ore map based on the second target detection algorithm to obtain position coordinates of a plurality of ore distributors;

[0079] determining the adjustment line based on the position coordinates of each ore distributor.

[0080] Further, in the first embodiment of the present application, when the ore distributor is a rotating arm, the position information is the rotation fulcrum coordinates of the ore distributor, and the identification of the position information of a plurality of ore distributors in the ore map and the determination of the adjustment line according to the position information of each ore distributor are specifically as follows:

[0081] performing ore distributor identification in the ore map based on the second target detection algorithm to obtain rotation fulcrum coordinates of a plurality of ore distributors;

[0082] determining the adjustment line based on the rotation fulcrum coordinates of each ore distributor.

[0083] In the first embodiment of the present application, the ore separator includes two types, one is a movable ore separation groove, and the ore separation groove is identified based on the second target detection algorithm which is completed by pre-training of the ore separator, so that the position coordinates of the ore separation groove can be collected, and the position coordinates of each ore separation groove are connected to obtain the adjustment line. The other type of the ore separator is a rotating arm, and the rotating arm is identified based on the second target detection algorithm which is completed by pre-training of the ore separator, so that the rotating fulcrum coordinates of the rotating arm can be collected, and the rotating fulcrum coordinates of each ore separation groove are connected to obtain the adjustment line. The present application can effectively improve the accuracy of the data and the speed of data acquisition by identifying the relevant data of the ore separator in real time through computer algorithm.

[0084] Step 104: obtaining the intersection point of the ore boundary line and the adjustment line.

[0085] Further, in the first embodiment of the present application, the intersection point of the ore boundary line and the adjustment line is obtained, specifically:

[0086] a plurality of boundary line coordinates are obtained on the ore boundary line;

[0087] a straight line fitting is performed based on the boundary line coordinates, and the fitted straight line is determined as a boundary line fitting straight line;

[0088] the intersection point of the boundary line fitting straight line and the adjustment line is determined as the intersection point of the ore boundary line and the adjustment line.

[0089] In the first embodiment of the present application, the ore boundary line obtained by identification detection on the ore map is not necessarily a straight line segment, but may be an irregular curve. Therefore, before determining the intersection point of the ore boundary line and the adjustment line, the ore boundary line needs to be straight line fitted. A large number of coordinate points are included on the ore boundary line, and a part of the coordinate points can be collected for straight line fitting. After the straight line fitting is completed, the intersection point of the ore boundary line and the adjustment line can be obtained by mathematical processing method.

[0090] Further, in the first embodiment of the present application, a plurality of boundary line coordinates are obtained on the ore boundary line, specifically:

[0091] all the coordinates of the points on the ore boundary line identified in the ore map are obtained, and all the coordinates of the points on the ore boundary line are determined as boundary line coordinates;

[0092] Alternatively, according to a preset end determination method, an end range is determined on the ore boundary line, a plurality of end coordinates are obtained in the end range, and the plurality of end coordinates are determined as boundary line coordinates;

[0093] Alternatively, a turning point identification is performed on the ore boundary line, an end direction of the ore boundary line is determined according to a preset end determination method, an end turning point closest to the end direction of the ore boundary line is determined as an end turning point, and coordinates of all points identified along the end direction of the ore boundary line are determined as boundary line coordinates.

[0094] As an example of the first embodiment of the present application, refer to Figure 2 , a schematic diagram of the boundary line coordinate acquisition method is provided, coordinates of all points on the ore boundary line are determined as boundary line coordinates, and linear fitting is performed to obtain the intersection point of the ore boundary line and the adjustment line.

[0095] As an example of the first embodiment of the present application, refer to Figure 3 , another schematic diagram of the boundary line coordinate acquisition method is provided, distances between the two ends of the ore boundary line and the adjustment line are calculated respectively, and the end with shorter distance is determined as the end of the ore boundary line. A plurality of end coordinates are collected at the end of the boundary line as boundary line coordinates, and linear fitting is completed. The method of collecting a plurality of end coordinates at the end of the boundary line includes: 1. Pre-set the percentage X of end coordinates, assuming that the number of data points on the ore boundary line is N, then the coordinates of N*X data points close to the end of the ore boundary line are determined as end coordinates. 2. Pre-set the percentage Y of end length, assuming that the length of the ore boundary line is L, then the coordinates of the data points within L*Y length close to the end of the ore boundary line are determined as end coordinates.

[0096] As an example of the first embodiment of the present application, refer to Figure 4 , another schematic diagram of the boundary line coordinate acquisition method is provided, the coordinates of the data points after the end turning point of the boundary line are determined as the boundary line coordinates. Specifically, the distances between the two ends of the ore boundary line and the adjustment line are calculated respectively, and the end with shorter distance is determined as the end of the ore boundary line; turning point identification is performed on the ore boundary line to obtain all turning points on the ore boundary line; the turning point closest to the end of the ore boundary line is determined as the end turning point, and the coordinates of all data points between the end turning point and the end of the ore boundary line are determined as the boundary line coordinates, and linear fitting is completed. The method of turning point identification on the ore boundary line is to perform smoothing processing on the ore boundary line first, then calculate the maximum value points on the ore boundary line, and determine all maximum value points as the turning points of the ore boundary line.

[0097] Step 105: According to the intersection point, the position information of each ore separator, and the preset deviation distance corresponding to each ore separator, the adjustment mode of each ore separator is determined.

[0098] Further, in the first embodiment of the present application, when the ore separator is a movable ore separating groove, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and the preset deviation distance corresponding to each ore separator, and specifically:

[0099] The intersection point coordinates are obtained on the ore map;

[0100] The target position coordinates corresponding to each ore separator are obtained according to the intersection point coordinates and the preset deviation distance corresponding to each ore separator;

[0101] The adjustment mode of each ore separator is determined as moving each ore separator along the adjustment line to the target position coordinates.

[0102] In the first embodiment of the present application, different kinds of ore separators correspond to different adjustment modes. When the ore separator is a movable ore separating groove, the adjustment mode is to translate the ore separator. First, the intersection point coordinates are obtained on the ore map, and then the target position coordinates corresponding to each ore separator are calculated based on the deviation distance corresponding to each ore separator, and each ore separator is translated on the adjustment line so as to move each ore separator to the target position coordinates.

[0103] As an example of the first embodiment of the present application, when the deviation distance corresponding to the ore separator is positive, it is determined that the target position coordinates of the ore separator are between the current position coordinates of the ore separator and the intersection point coordinates. At this time, the actual distance between the ore separator and the intersection point is calculated according to the current position coordinates of the ore separator and the intersection point coordinates, the translation distance of the ore separator is obtained by subtracting the deviation distance from the actual distance, and the ore separator is translated on the adjustment line based on the translation distance so as to move the ore separator to the target position coordinates.

[0104] As an example of the first embodiment of the present application, when the deviation distance corresponding to the ore separator is negative, it is determined that the distance between the target position coordinates of the ore separator and the current position coordinates is greater than the distance between the current position coordinates of the ore separator and the intersection point coordinates. At this time, the actual distance between the ore separator and the intersection point is calculated according to the current position coordinates of the ore separator and the intersection point coordinates, the translation distance of the ore separator is obtained by adding the deviation distance to the actual distance, and the ore separator is translated on the adjustment line based on the translation distance so as to move the ore separator to the target position coordinates. Further, in the first embodiment of the present application, when the ore separator is a rotating arm, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and the preset deviation distance corresponding to each ore separator, and specifically:

[0105] Obtaining intersection point coordinates, rotation fulcrum coordinates and effective arm length, first angle and second angle of each of the ore separators on the ore map respectively; the first angle is the included angle between the ore boundary line and the adjustment line; the second angle is the included angle between each of the ore separators and the adjustment line;

[0106] Based on the preset deviation distance corresponding to each of the ore separators, an auxiliary line corresponding to each of the ore separators and parallel to the ore boundary line is constructed;

[0107] An auxiliary circle of each of the ore separators is constructed with the effective arm length of each of the ore separators as a radius and the rotation fulcrum as a center;

[0108] The target position of the rotating arm end of each of the ore separators is determined from two intersection points of the auxiliary line and the auxiliary circle;

[0109] The target angle of the target included angle between each of the ore separators and the adjustment line is determined according to the intersection point coordinates, the rotation fulcrum coordinates, the effective arm length and the first angle of each of the ore separators;

[0110] The rotation angle of each of the ore separators is calculated according to the second angle and the target angle;

[0111] The adjustment mode of each of the ore separators is determined as rotating the rotation angle of each of the ore separators, so that the target position of the rotating arm end of each of the ore separators is located at the target position of the rotating arm end.

[0112] In the first embodiment of the present application, for the ore separator with a rotating arm, the adjustment mode is to rotate the rotating arm. Each of the ore separators corresponds to a deviation distance. When the target position of the rotating arm end is determined, an auxiliary line parallel to the ore boundary line is first constructed based on the deviation distance. The target position of the rotating arm end is on the auxiliary line. Therefore, an auxiliary circle of each of the ore separators can be constructed with the effective arm length of each of the ore separators as a radius and the rotation fulcrum as a center. The point closer to the initial flow position of the ore to be screened from the two intersection points between the auxiliary line and the auxiliary circle is the target position of the rotating arm end. After the target position of the rotating arm end is determined, the rotation angle of each of the ore separators can be calculated through mathematical operation according to the intersection point coordinates, the rotation fulcrum coordinates and the effective arm length, the first angle and the second angle of each of the ore separators obtained on the ore map. Based on the rotation angle, each of the ore separators is controlled to rotate with the rotation fulcrum as a fixed point, so that the target position of the rotating arm end of each of the ore separators is located at the target position of the rotating arm end, to adjust the ore separator with a rotating arm. It should be noted that, according to the different deviation distances corresponding to each of the ore separators, the auxiliary line parallel to the ore boundary line can be located on the left side or the right side of the ore boundary line.

[0113] Further, in the first embodiment of the present application, a target angle of a target included angle between each of the ore separators and the adjustment line is determined according to the intersection point coordinate, the rotation fulcrum coordinate of each of the ore separators, the effective arm length, and the first angle, specifically:

[0114] A first distance between each of the ore separators and the intersection point is calculated according to the intersection point coordinate and the rotation fulcrum coordinate of each of the ore separators;

[0115] An intersection point between the auxiliary line and the adjustment line is determined as an auxiliary intersection point of each of the ore separators;

[0116] A second distance between each of the ore separators and the auxiliary intersection point is calculated according to the first distance between each of the ore separators and the intersection point and a preset deviation distance corresponding to each of the ore separators;

[0117] A third angle between a target line segment and the auxiliary line is calculated according to the effective arm length of each of the ore separators, the first angle, and the second distance by using the sine theorem, wherein the target line segment is a line segment formed by connecting the rotation arm end target position of each of the ore separators and the rotation fulcrum;

[0118] The target angle of the target included angle between each of the ore separators and the adjustment line is calculated according to the first angle and the third angle by using the theorem of interior angles of a triangle.

[0119] As an example of the first embodiment of the present application, refer to Figure 5, is a schematic diagram of the rotating arm adjusting method provided by the present application, the adjusting mode of the ore separator of the rotating arm is rotating the rotating arm, and the purpose of the adjustment is to place the end of the rotating arm on an auxiliary line parallel to the ore boundary line. The second angle between the ore separator and the adjustment line can be obtained on the ore map, so that only the target angle of the target included angle between the ore separator and the adjustment line is required when calculating the rotating angle of the ore separator, that is, the difference between the second angle and the target angle can be determined as the rotating angle. The target angle can be calculated by using a triangle formed by the intersection point between the rotating fulcrum, the adjustment line and the auxiliary line, and the target position of the end of the rotating arm. First, the first angle between the ore boundary line and the adjustment line can be obtained on the ore map, and since the auxiliary line and the ore boundary line are parallel, the included angle between the auxiliary line and the adjustment line is also the first angle; second, the first distance between the ore separator and the intersection point is calculated based on the coordinates of the intersection point and the coordinates of the rotating fulcrum obtained on the ore map, and then the second distance between the ore separator and the auxiliary line is calculated based on the deviation distance; the distance between the rotating fulcrum and the target position of the end of the rotating arm is equivalent to the effective arm length of the ore separator. Therefore, in the auxiliary triangle, under the premise that the effective arm length of the ore separator, the first angle and the second distance are known, the third angle corresponding to the top vertex of the target position of the end of the rotating arm can be calculated by using the sine theorem. After the third angle is calculated, the target angle of the target included angle between the ore separator and the adjustment line is obtained by subtracting the sum of the first angle and the third angle from 180° based on the theorem of interior angles of a triangle.

[0120] In the first embodiment of the present application, there are two intersection points between the auxiliary line and the auxiliary circle, and the target position of the end of the rotating arm can be determined according to the initial flow position of the ore to be screened into the ore dressing equipment. In the first embodiment of the present application, the point closer to the initial flow position of the ore to be screened among the two intersection points is the target position of the end of the rotating arm, so that the target ore and the non-target ore in the ore to be screened can be separated at the end of the rotating arm, thereby achieving the purpose of automatically screening the ore. Specifically, the distances between the two intersection points between the auxiliary line and the auxiliary circle and the initial flow position of the ore to be screened are calculated respectively, and the intersection point closer to the initial flow position of the ore to be screened is determined as the target position of the end of the rotating arm.

[0121] In the first embodiment of the present application, after the angle value of the target angle is calculated, the actual angle value between the rotating arm and the adjustment line is subtracted from the target angle value to obtain the rotating angle, and the direction of each rotating arm is adjusted based on the rotating angle. When the calculated rotating angle is positive, the ore separator is controlled to rotate in the direction towards the adjustment line; when the calculated rotating angle is negative, the ore separator is controlled to rotate in the direction away from the adjustment line.

[0122] Step 106: control each ore separator to be adjusted to the corresponding target position by the corresponding adjustment mode.

[0123] In the first embodiment of the present application, different adjustment modes are adopted to adjust different types of ore separators, so that the ore separators reach the target positions by translation or rotation, thereby achieving better ore separation effect and improving work efficiency.

[0124] To sum up, the first embodiment of the present application provides an intelligent ore dressing method based on machine vision technology, which uses an image collector to collect a mine stone image in real time, identifies the mine stone image, determines a mine stone boundary line in the mine stone image, identifies position information of a plurality of ore separators in the mine stone image, determines an adjustment line according to the position information of the ore separators, obtains an intersection point of the mine stone boundary line and the adjustment line, determines an adjustment mode of each ore separator according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, and controls each ore separator to adjust to a corresponding target position through the corresponding adjustment mode. The present application realizes real-time identification of the position of the target mine stone through machine vision technology, automatic adjustment of the position and direction of the ore separator, and fully automatic mine stone screening. When the type or taste of the mine stone changes, no human intervention is required, work efficiency is improved, and the identification accuracy of the target mine stone is improved.

[0125] Embodiment 2

[0126] Reference Figure 6 is a structural schematic diagram of an embodiment of an intelligent ore dressing device based on machine vision technology provided by the present application. The device comprises a collection module 201, a boundary line determination module 202, an adjustment line determination module 203, an intersection point acquisition module 204, an adjustment mode determination module 205, and a control module 206.

[0127] The collection module 201 is used to collect a mine stone image in real time by using an image collector.

[0128] The boundary line determination module 202 is used to identify the mine stone image and determine a mine stone boundary line in the mine stone image. The mine stone boundary line is used to distinguish target mine stones and non-target mine stones.

[0129] The adjustment line determination module 203 is used to identify position information of a plurality of ore separators in the mine stone image and determine an adjustment line according to the position information of each ore separator.

[0130] The intersection point acquisition module 204 is used to obtain an intersection point of the mine stone boundary line and the adjustment line.

[0131] The adjustment mode determination module 205 is used to determine an adjustment mode of each ore separator according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator.

[0132] The control module 206 is configured to control each of the ore sorters to adjust to a corresponding target position by a corresponding adjustment mode.

[0133] Further, in the second embodiment of the present application, the ore map is identified, and an ore boundary line is determined in the ore map, specifically:

[0134] The ore map is subjected to picture equalization, a range of target colors is determined in the ore map, and an ore boundary line is determined in combination with a preset minimum color band width; wherein the target colors include a preset target ore color and a non-target ore color;

[0135] Alternatively, the ore map is subjected to edge detection based on an edge detection algorithm, and an ore boundary line is determined according to an edge detection result; wherein the edge detection result includes a target ore edge and a non-target ore edge;

[0136] Alternatively, the ore map is subjected to boundary line detection based on a first target detection algorithm, and an ore boundary line is determined according to a boundary line detection result.

[0137] Further, in the second embodiment of the present application, the ore sorter is a movable ore sorting groove, the position information is a position coordinate of the ore sorter, the position information of the plurality of ore sorters is identified in the ore map, and an adjustment line is determined according to the position information of each of the ore sorters, specifically:

[0138] The ore sorters are identified in the ore map based on a second target detection algorithm, and the position coordinates of the plurality of ore sorters are obtained;

[0139] The adjustment line is determined based on the position coordinates of each of the ore sorters.

[0140] Further, in the second embodiment of the present application, the ore sorter is a rotating arm, the position information is a rotating fulcrum coordinate of the ore sorter, the position information of the plurality of ore sorters is identified in the ore map, and an adjustment line is determined according to the position information of each of the ore sorters, specifically:

[0141] The ore sorters are identified in the ore map based on a second target detection algorithm, and the rotating fulcrum coordinates of the plurality of ore sorters are obtained;

[0142] The adjustment line is determined based on the rotating fulcrum coordinates of each of the ore sorters.

[0143] Further, in the second embodiment of the present application, an intersection of the ore boundary line and the adjustment line is obtained, specifically:

[0144] A plurality of boundary line coordinates are obtained on the ore boundary line;

[0145] The straight line fitting is performed based on the coordinates of the demarcation lines, and the fitted straight line is determined as the demarcation line fitting straight line;

[0146] The intersection point of the demarcation line fitting straight line and the adjustment line is determined as the intersection point of the ore demarcation line and the adjustment line.

[0147] Further, in the second embodiment of the present application, a plurality of demarcation line coordinates are obtained on the ore demarcation line, specifically:

[0148] The coordinates of all points on the identified ore demarcation line in the ore map are obtained, and the coordinates of all points on the ore demarcation line are determined as demarcation line coordinates;

[0149] Alternatively, according to a preset end determination method, a demarcation line end range is determined on the ore demarcation line, a plurality of end coordinates are obtained in the demarcation line end range, and the plurality of end coordinates are determined as demarcation line coordinates;

[0150] Alternatively, inflection point recognition is performed on the ore demarcation line, the end direction of the ore demarcation line is determined according to a preset end determination method, the inflection point closest to the end direction of the ore demarcation line is determined as an end inflection point, and the coordinates of all points recognized along the end direction of the ore demarcation line are determined as demarcation line coordinates.

[0151] Further, in the second embodiment of the present application, when the ore separator is a movable ore separation groove, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, specifically:

[0152] The intersection point coordinates are obtained on the ore map;

[0153] The target position coordinates corresponding to each ore separator are obtained according to the intersection point coordinates and a preset deviation distance corresponding to each ore separator;

[0154] The adjustment mode of each ore separator is determined as moving each ore separator to the target position coordinates along the adjustment line.

[0155] Further, in the second embodiment of the present application, when the ore separator is a rotating arm, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, specifically:

[0156] Obtaining intersection point coordinates, rotation fulcrum coordinates and effective arm length, first angle and second angle of each ore separator on the ore map respectively; the first angle is the included angle between the ore boundary line and the adjustment line; the second angle is the included angle between each ore separator and the adjustment line;

[0157] Combined with the preset deviation distance corresponding to each ore separator, an auxiliary line corresponding to each ore separator and parallel to the ore boundary line is constructed;

[0158] An auxiliary circle of each ore separator is constructed with the effective arm length of each ore separator as the radius and the rotation fulcrum as the center;

[0159] The target position of the rotation arm end of each ore separator is determined among the two intersection points of the auxiliary line and the auxiliary circle;

[0160] According to the intersection point coordinates, rotation fulcrum coordinates, effective arm length and first angle of each ore separator, the target angle of the target included angle between each ore separator and the adjustment line is determined;

[0161] According to the second angle and the target angle, the rotation angle of each ore separator is calculated;

[0162] The adjustment mode of each ore separator is determined as rotating the rotation angle of each ore separator, so that the rotation arm end of each ore separator is located at the target position of the rotation arm end.

[0163] Further, in the second embodiment of the present application, according to the intersection point coordinates, rotation fulcrum coordinates, effective arm length and first angle of each ore separator, the target angle of the target included angle between each ore separator and the adjustment line is determined, specifically:

[0164] According to the intersection point coordinates and rotation fulcrum coordinates of each ore separator, the first distance between each ore separator and the intersection point is calculated;

[0165] The intersection point between the auxiliary line and the adjustment line is determined as the auxiliary intersection point of each ore separator;

[0166] According to the first distance between each ore separator and the intersection point, and the preset deviation distance corresponding to each ore separator, the second distance between each ore separator and the auxiliary intersection point is calculated;

[0167] According to the effective arm length of each ore separator, the first angle and the second distance, the third angle between the target line segment and the auxiliary line is calculated by using the sine theorem; wherein the target line segment is a line segment formed by connecting the target position of the rotation arm end of each ore separator and the rotation fulcrum;

[0168] The target angle of the target included angle between each of the ore dividers and the adjustment line is calculated according to the first angle and the third angle by using the triangle interior angle theorem.

[0169] In summary, the second embodiment of the present application provides an intelligent ore dressing device based on machine vision technology, which is based on the organic combination of modules, uses an image collector to collect ore images in real time, identifies the ore images, determines the ore boundary line in the ore images, identifies the position information of a plurality of ore dividers in the ore images, determines the adjustment line according to the position information of each ore divider, obtains the intersection point of the ore boundary line and the adjustment line, determines the adjustment mode of each ore divider according to the intersection point, the position information of each ore divider, and the preset deviation distance corresponding to each ore divider, and controls each ore divider to be adjusted to the corresponding target position through the corresponding adjustment mode. The present application realizes real-time identification of the position of the target ore through machine vision technology, automatic adjustment of the position and direction of the ore divider, and fully automatic ore screening. When the type or taste of the ore changes, no human intervention is required, the work efficiency is improved, and the identification accuracy of the target ore is improved.

[0170] The above specific embodiments further illustrate the purpose, technical solutions and advantages of the present application. It should be understood that the above description is only for specific embodiments of the present application and is not intended to limit the protection scope of the present application. It is particularly pointed out that any modification, equivalent replacement, improvement, etc. made by those skilled in the art within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. An intelligent beneficiation method based on machine vision technology, characterized in that, The method comprises the following steps: real-time acquisition of an ore image by an image collector; identification of the ore image, determination of an ore boundary line in the ore image, specifically: picture equalization of the ore image, determination of a range of target colors in the ore image, and determination of the ore boundary line in combination with a preset minimum color band width, wherein the target colors include a preset target ore color and a non-target ore color; alternatively, edge detection of the ore image based on an edge detection algorithm, and determination of the ore boundary line according to the edge detection result, wherein the edge detection result includes a target ore edge and a non-target ore edge; alternatively, boundary line detection of the ore image based on a first target detection algorithm, and determination of the ore boundary line according to the boundary line detection result, wherein the ore boundary line is used to distinguish between target ores and non-target ores; identification of position information of a plurality of ore separators in the ore image, and determination of an adjustment line according to the position information of the ore separators; determination of an intersection point of the ore boundary line and the adjustment line, specifically: acquisition of a plurality of boundary line coordinates on the ore boundary line; straight line fitting based on the boundary line coordinates, and determination of a fitted straight line as a boundary line fitting straight line; determination of an intersection point of the boundary line fitting straight line and the adjustment line as the intersection point of the ore boundary line and the adjustment line; determination of an adjustment mode of each ore separator according to the intersection point, the position information of the ore separators, and a preset deviation distance corresponding to each ore separator; control of each ore separator to be adjusted to a corresponding target position through the corresponding adjustment mode; wherein the acquisition of a plurality of boundary line coordinates on the ore boundary line specifically comprises: acquisition of coordinates of all points on the identified ore boundary line in the ore image, and determination of the coordinates of all points on the ore boundary line as boundary line coordinates; alternatively, determination of a boundary line end range on the ore boundary line according to a preset end determination method, acquisition of a plurality of end coordinates in the boundary line end range, and determination of the plurality of end coordinates as boundary line coordinates; alternatively, inflection point identification on the ore boundary line, determination of an end direction of the ore boundary line according to a preset end determination method, determination of an inflection point closest to the end direction of the ore boundary line as an end inflection point, and determination of coordinates of all points identified in the end direction of the ore boundary line as boundary line coordinates.

2. The machine vision technology based intelligent beneficiation method as claimed in claim 1, wherein, The ore separator is a movable ore separation groove, the position information is a position coordinate of the ore separator, and the identification of the position information of a plurality of ore separators in the ore image and the determination of an adjustment line according to the position information of the ore separators specifically comprise: ore separator identification in the ore image based on a second target detection algorithm, and acquisition of position coordinates of a plurality of ore separators; determination of an adjustment line based on the position coordinates of the ore separators.

3. The machine vision technology based intelligent beneficiation method as claimed in claim 1, wherein, The ore separator is a rotating arm, the position information is a rotating fulcrum coordinate of the ore separator, and the identification of the position information of a plurality of ore separators in the ore image and the determination of an adjustment line according to the position information of the ore separators specifically comprise: The rotating fulcrum coordinates of a plurality of ore separators are obtained based on the second target detection algorithm in the ore map; The adjustment line is determined based on the rotating fulcrum coordinates of each ore separator.

4. The machine vision technology based intelligent beneficiation method as claimed in claim 2, wherein, When the ore separator is a movable ore separating groove, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, and specifically: The intersection point coordinates are obtained on the ore map; The target position coordinates corresponding to each ore separator are obtained according to the intersection point coordinates and a preset deviation distance corresponding to each ore separator; The adjustment mode of each ore separator is determined as moving each ore separator along the adjustment line to the target position coordinates.

5. The machine vision technology based intelligent beneficiation method as claimed in claim 3, wherein, When the ore separator is a rotating arm, the adjustment mode of each ore separator is determined according to the intersection point, the position information of each ore separator, and a preset deviation distance corresponding to each ore separator, and specifically: The intersection point coordinates, the rotating fulcrum coordinates and the effective arm length, the first angle and the second angle of each ore separator are obtained on the ore map; the first angle is the included angle between the ore boundary line and the adjustment line; and the second angle is the included angle between each ore separator and the adjustment line; An auxiliary line corresponding to each ore separator and parallel to the ore boundary line is constructed according to a preset deviation distance corresponding to each ore separator; An auxiliary circle of each ore separator is constructed with the effective arm length of each ore separator as the radius and the rotating fulcrum as the center; The target position of the end of the rotating arm of each ore separator is determined at the intersection of the auxiliary line and the auxiliary circle; The target angle of the target included angle between each ore separator and the adjustment line is determined according to the intersection point coordinates, the rotating fulcrum coordinates, the effective arm length, and the first angle of each ore separator; The rotating angle of each ore separator is calculated according to the second angle and the target angle; The adjustment mode of each ore separator is determined as rotating each ore separator by the rotating angle, so that the end of the rotating arm of each ore separator is located at the target position of the end of the rotating arm.

6. The machine vision technology based intelligent beneficiation method as claimed in claim 5, wherein, The target angle of the target included angle between each ore separator and the adjustment line is determined according to the intersection point coordinates, the rotating fulcrum coordinates, the effective arm length, and the first angle of each ore separator, and specifically: The first distance between each ore separator and the intersection point is calculated according to the intersection point coordinates and the rotating fulcrum coordinates of each ore separator; The intersection point between the auxiliary line and the adjustment line is determined as the auxiliary intersection point of each ore separator; The second distance between each ore separator and the auxiliary intersection point is calculated according to the first distance between each ore separator and the intersection point and a preset deviation distance corresponding to each ore separator; The third angle between the target line segment and the auxiliary line is calculated according to the effective arm length of each ore separator, the first angle and the second distance by using the sine theorem; wherein the target line segment is a line segment formed by connecting the target position of the end of the rotating arm of each ore separator and the rotating fulcrum. A target angle of a target included angle between each of the ore dividers and the adjustment line is calculated according to the first angle and the third angle by using the triangle interior angle theorem.

7. An intelligent beneficiation device based on machine vision technology, characterized in that, The intelligent ore dressing method based on the machine vision technology comprises a collection module, a boundary line determination module, an adjustment line determination module, an intersection point acquisition module, an adjustment mode determination module and a control module. The collection module is configured to collect a mine ore image in real time by using an image collector. The boundary line determination module is configured to identify the mine ore image and determine a mine ore boundary line in the mine ore image, wherein the mine ore boundary line is configured to distinguish target mine ore and non-target mine ore. The adjustment line determination module is configured to identify position information of a plurality of ore dividers in the mine ore image and determine an adjustment line according to the position information of the ore dividers. The intersection point acquisition module is configured to acquire intersection points of the mine ore boundary line and the adjustment line. The adjustment mode determination module is configured to determine an adjustment mode of each of the ore dividers according to the intersection points, the position information of the ore dividers and a preset deviation distance corresponding to each of the ore dividers. The control module is configured to control each of the ore dividers to be adjusted to a corresponding target position by using the corresponding adjustment mode.

Images