Method for monitoring cut tobacco allowance based on point cloud imaging

Abstract

Provided is a cut tobacco allowance monitoring method based on point cloud imaging. The method relies on a non-contact measurement means of stereo vision, uses binocular vision difference theory to establish a binocular imaging mechanism, uses object shadow degree and texture information to perform stereo combination and estimate surface information through a triangulation process, uses imaging characteristics of speckle structured light to convert static calibration into multi-directional unlimited area calibration, and finally obtains depth information of an image surface, thereby providing conditions for calculation of a change amount of a cut tobacco volume, achieving monitoring of cut tobacco separation effect of a system by using the change amount of the cut tobacco volume, and achieving high efficiency in measurement time, thereby providing efficient monitoring conditions for data acquisition and simulation of cut tobacco allowance in a falling material vibrating trough.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of cigarette production, and particularly relates to a stem filament residual amount monitoring method based on point cloud imaging. BACKGROUND

[0002] Whether the stem filament separation system of a cigarette machine can completely separate stem filaments will affect the finished product quality of cigarettes and the use rate of raw materials. If filaments contain stems during cigarette production, it will cause major quality accidents such as cigarette creases, cigarette burst, cigarette puncture, and the like, and will also affect the taste of the cigarettes for consumers and the favorability of the brand cigarettes. If stems contain filaments, it will cause a large amount of tobacco waste. Therefore, if the operation effect of the stem filament separation system can be monitored and the quality event can be responded in a timely manner, the finished product rate of cigarettes can be effectively improved, and a large amount of raw material waste can be avoided.

[0003] At present, the existing monitoring means mainly relies on manual observation. With the development of three-dimensional reconstruction technology, the volume change under three-dimensional reconstruction is used in multiple fields to solve some monitoring problems. The current three-dimensional profile data acquisition technology has a static scanning mode based on coded structured light. This mode needs to calibrate the target in multiple points and multiple directions, and is not suitable for objects with unstable surface state in a short time. The second is a scanning mode based on laser radar. After the special light is irradiated to the surface of the object, the phase difference of the light is obtained by a conversion device. If high-efficiency and accurate measurement is required, a 12-line laser radar needs to be used, which is too high in cost. Moreover, this method is not suitable for the three-dimensional reconstruction of the surface of an object with longitudinal data change. The data splicing mainly uses a curvature fitting method with unidirectional change and an edge segmentation algorithm. This method has low three-dimensional simulation accuracy for the surface of the object in the data set, and is not conducive to the establishment of a volume model. Therefore, the falling vibration chute is a fixed basic surface based on monitoring, the stem filament surface state is a data stream with longitudinal movement, and the required measurement time is relatively short. According to the surface state of the object, a high-efficiency algorithm suitable for stem filament residual amount data acquisition and simulation in the falling vibration chute needs to be designed. SUMMARY

[0004] The technical problem solved by the application is to provide a stem filament residual amount monitoring method based on point cloud imaging. The method relies on a non-contact measurement means of stereovision, uses a binocular imaging mechanism established based on the binocular vision difference theory, performs stereoscopic combination based on the shadow degree and texture information of an object, and estimates surface information through a triangulation process. The imaging characteristics of a speckle structured light are used to convert static calibration into multi-directional unlimited area calibration. Finally, the depth information of the image surface is obtained, which provides conditions for calculating the change amount of the stem filament volume, so as to monitor the stem filament separation effect of the system by using the change amount of the stem filament volume, measure in a short time, have high efficiency, and provide high-efficiency monitoring conditions for stem filament residual amount data acquisition and simulation in the falling vibration chute.

[0005] The technical scheme adopted by the present application is: a cut tobacco allowance monitoring method based on point cloud imaging, comprising the following steps:

[0006] 1) obtaining point cloud data of the cut tobacco surface: using a light emitter with a speckle structure to project light on the surface of the cut tobacco to be measured, and collecting the reflected speckle structure light by an infrared camera to obtain point cloud coordinate information of the cut tobacco in a three-dimensional image coordinate system;

[0007] 2) converting the discrete point cloud coordinate information of the cut tobacco in the three-dimensional image coordinate system into discrete point cloud coordinate information in the infrared camera coordinate system;

[0008] 3) using a gradient segmentation interpolation algorithm to sequentially perform X-axis direction interpolation fitting and Y-axis direction interpolation fitting on the obtained discrete point cloud data set M in the infrared camera coordinate system, and then fitting the direction data obtained by the X-axis direction interpolation fitting and the direction data obtained by the Y-axis direction interpolation fitting to form an irregular prism data set with a square bottom surface and close to the state of the cut tobacco entity surface;

[0009] 4) using a sobel operator to perform edge detection on the point cloud data image formed by the obtained irregular prism set, remove the boundary points in the point cloud data image, and obtain a cut tobacco entity graph including the cut tobacco surface point cloud data image set;

[0010] 5) calculating the volume V of the cut tobacco entity graph, and using the change amount of the cut tobacco volume to monitor the cut tobacco separation effect of the system.

[0011] In the above step 2), the infrared cameras are symmetrically arranged in two, and the specific process of converting the discrete point cloud coordinates of the cut tobacco in the three-dimensional image coordinate system into the discrete point cloud coordinates in the infrared camera coordinate system is as follows:

[0012] The optical axis directions of the left and right infrared cameras are set as axis, the optical center position of the light under the optical system of the infrared camera is set as the origin , the horizontal axis and the vertical axis in the infrared camera coordinate system are represented by and respectively, and and are parallel to the X-axis and the Y-axis of the three-dimensional image coordinate system respectively, point B is the point coordinate of the object in the infrared camera coordinate system, , are the imaging point coordinates of point P in the three-dimensional image coordinate system collected by the left and right infrared cameras respectively, and the optical center is at a distance of focal length from the three-dimensional image coordinate origin , and the distance between the two infrared cameras installed on the same plane is b, so that point In the three-dimensional image coordinate system and the infrared camera coordinate system, there is the same Y-axis coordinate, that is According to the triangle principle, the following mapping relationship can be obtained:

[0013] (1)

[0014] (2)

[0015] (3)

[0016] (4)

[0017] wherein represents the coordinate parallax value of the point in the corresponding point in the left and right infrared cameras, and from equation (4), the corresponding coordinates of the point in the infrared camera coordinate system in space are obtained as .

[0018] In the above step 3), the specific process of the discrete point cloud data in the infrared camera coordinate system in the X-axis direction interpolation fitting is as follows:

[0019] The discrete point cloud data on the cut tobacco surface is a set of points without continuity in space When the discrete point cloud data is interpolated and fitted in the X-axis direction, the development of the data in the Y-axis direction is not considered, and only the changes of the data in the X-axis and axis are concerned, that is, the starting point in the X-axis direction is selected, and the fitting of the point needs to meet , that is, the next fitting point in the X-axis is the point with the same Y value as the previous point in the direction parallel to the X-axis, the point closest to in the X-axis direction is taken for fitting, and is satisfied. When the data is not interpolated and supplemented, the smaller the value is, the closer the fitting result is to the original surface. At this time, the gradient value of the point to the point is , and the fitting line equation of the adjacent two points in the X-axis direction is:

[0020] (5)

[0021] The middle of and is divided into data points for data compensation, and it is assumed that ​​Then the X-axis direction of any point between the two data is known Coordinates are:

[0022] (6)

[0023] In the above formula, represents The first data in the data point belongs to N+;

[0024] Similarly, the minimum offset of the known point and in the X direction is calculated In order to meet the uniform distribution of data and the equal step splicing of later data, the number of interpolation points between point and should be That is, the number of interpolation points between the jth adjacent two points when Y value is equal should be From which the coordinates of any point in the data stream parallel to the X-axis direction when Y value is equal can be obtained

[0025] (7)

[0026] Where , , , i = 1, 2, 3… , i, j belong to N+;

[0027] As described above, the discrete point cloud data in the infrared camera coordinate system is fitted by Y-axis direction interpolation, and the data stream parallel to the Y-axis is supplemented, then the number of interpolation between any two adjacent points is Then the coordinates of any point are:

[0028] (8)

[0029] Where , , = 1, 2, 3… n+1, , j, All belong to N+.

[0030] In the above step 4), the specific process of the sobel operator for image boundary detection processing is as follows:

[0031] Let M represent the two-dimensional point cloud data information on the XY plane, and the data in M is subjected to horizontal edge detection and vertical edge detection, respectively, and the specific processing is as follows:

[0032] (9)

[0033] (10)

[0034] (11)

[0035] In the point cloud data, the gradient size G is obtained by using formula (11) in combination with the lateral gradient value of each point , the longitudinal gradient value When the gradient value G is greater than a set threshold, it is determined that the edge is located at the critical surface edge, and the point cloud data of the boundary point is processed to 0, so that the cut tobacco surface point cloud splicing entity is obtained.

[0036] In the above step 5), the cut tobacco entity surface data set is divided into multiple rows along the Y axis direction in turn, and each row includes a plurality of prisms with different heights, the volume sum of the plurality of prisms in each row is calculated, and the volume sum of the plurality of prisms in each row is accumulated to obtain the volume of the cut tobacco entity graph, and the specific calculation process is as follows:

[0037] Suppose the volume of a single prism in each row after segmentation is , The specific calculation formula of the volume sum of the plurality of prisms in each row is as follows:

[0038] (12)

[0039] In the above formula, represents the height of the four edges of a single prism, and then the volume sum of the prisms in the first row after segmentation is The specific calculation formula of the volume sum of the plurality of prisms in each row is as follows:

[0040] (13)

[0041] As shown above, the volume of the cut tobacco allowance is represented as the cumulative sum of the volume of each row in the Y direction, that is, the volume V is:

[0042] (14).

[0043] The advantages of the present application compared with the prior art are:

[0044] 1. The technical solution relies on a non-contact measurement means of stereovision, establishes a binocular imaging mechanism by using the binocular vision difference theory, and converts the coordinate system, so that the position relationship between the measured object and the infrared camera can be quickly and accurately calibrated, so that the infrared camera can accurately identify the coordinate information of the reflected light, and provide conditions for the accuracy of the collected point cloud data.

[0045] 2, The technical scheme is based on gradient segmentation interpolation algorithm, that is, on the basis of a large number of discrete point data on the surface of the stem, the missing part is interpolated and subdivided, which finds the corresponding source for subsequent data splicing, realizes the closest fitting of the surface state of the stem, and ensures the calculation accuracy of the volume change of the stem;

[0046] 3, The technical scheme utilizes the imaging characteristics of speckle structured light, converts static calibration into multi-directional unlimited area calibration, finally obtains the depth information of the image surface, provides conditions for the calculation of the volume change of the stem, and achieves the monitoring effect of stem separation by using the volume change of the stem; The measurement time is efficient and low in cost, which provides efficient monitoring conditions for the data acquisition and simulation of the remaining amount of stem in the falling material vibration tank. BRIEF DESCRIPTION OF DRAWINGS

[0047] Figure 1 The figure is a conversion principle diagram of the three-dimensional image coordinate system and the infrared camera coordinate system of the present application;

[0048] Figure 2 The figure is a principle diagram of interpolation fitting of discrete point cloud data in the X-axis direction of the present application;

[0049] Figure 3 The figure is a principle diagram of interpolation fitting of discrete point cloud data in the Y-axis direction of the present application;

[0050] Figure 4 The figure is a stem surface point cloud splicing diagram of the present application;

[0051] Figure 5 The figure is a stem surface point cloud splicing entity diagram of the present application;

[0052] Figure 6 The figure is a structure diagram of one row of the stem entity surface data set divided along the Y-axis direction of the present application;

[0053] Figure 7 The figure is a flow chart of the present application. DETAILED DESCRIPTION

[0054] The technical solutions in the embodiments of the present application will be described below. Figures 1-7 The technical solutions in the embodiments of the present application will be described below. Obviously, the described embodiments are only a part of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0055] It is to be understood that the terms "including", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can also include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the statement "comprising a... " does not exclude the presence of additional identical elements in the process, method, article, or apparatus that includes the element.

[0056] The cut tobacco residual amount monitoring method based on point cloud imaging comprises the following steps:

[0057] 1) Obtain point cloud data of the cut tobacco surface: use a speckle structure light emitter to project light on the surface of the cut tobacco to be measured, and collect the reflected speckle structure light by an infrared camera to obtain point cloud coordinate information of the cut tobacco in a three-dimensional image coordinate system; wherein the speckle structure light emitter is composed of a computer and a binocular camera to realize light projection on the surface of the measured cut tobacco, and the reflected speckle structure light is collected by the infrared camera to obtain point cloud coordinate information of the cut tobacco in a three-dimensional space;

[0058] 2) Convert the discrete point cloud coordinate information of the cut tobacco in the three-dimensional image coordinate system into discrete point cloud coordinate information in the infrared camera coordinate system; specifically, the infrared camera is symmetrically provided with two, and the specific process of converting the discrete point cloud coordinate of the cut tobacco in the three-dimensional image coordinate system into the discrete point cloud coordinate in the infrared camera coordinate system is as follows:

[0059] As shown in Figure 1 , the optical axis directions of the left and right infrared cameras are defined as the axis, the optical center position of the optical system under the infrared camera is defined as the origin , the horizontal axis and the vertical axis in the infrared camera coordinate system are represented by and respectively, and and are parallel to the X axis and the Y axis of the three-dimensional image coordinate system respectively, point B is the point coordinate of the object in the infrared camera coordinate system, , are the imaging point coordinates of point P in the three-dimensional image coordinate system collected by the left and right infrared cameras respectively, the optical center is at a distance of focal length from the three-dimensional image coordinate origin , the distance between the two infrared cameras installed on the same plane is b, so that point has the same Y axis coordinate in the three-dimensional image coordinate system and the infrared camera coordinate system, that is According to the triangle principle, the following mapping relationship can be obtained:

[0060] , (1)

[0061] (2)

[0062] (3)

[0063] (4)

[0064] wherein represents the coordinate parallax value of the point in the corresponding point in the left and right infrared cameras, and the corresponding coordinates of the point in the infrared camera coordinate system in the space can be obtained from equation (4) ;

[0065] 3) The gradient segmentation interpolation algorithm is adopted to sequentially perform X-axis direction interpolation fitting and Y-axis direction interpolation fitting on the obtained discrete point cloud data set M in the infrared camera coordinate system, and then the direction data obtained by the X-axis direction interpolation fitting and the direction data obtained by the Y-axis direction interpolation fitting are fitted to form an irregular prism data set with a square bottom and close to the surface state of the cut tobacco entity; specifically, the specific process of the X-axis direction interpolation fitting of the discrete point cloud data in the infrared camera coordinate system is as follows:

[0066] The discrete point cloud data of the cut tobacco surface is a set of points without continuity in space When the discrete point cloud data is fitted and spliced in the X-axis direction, the development of the Y-axis direction data is not considered, and only the changes of the data on the X-axis and the Y-axis are focused on, as shown in FIG. 8, the starting point on the X-axis direction is selected, then the fitting of the point needs to meet Figure 2 , that is, the next fitting point on the X-axis is the point with the same Y value as the previous point in the direction parallel to the X-axis, and the point closest to in the upward direction of the X-axis is taken for fitting, which meets , when the data is not interpolated and supplemented, the smaller the value of is, the closer the fitting result is to the original curved surface, at this time, the gradient value of the point to the point is , and the fitting line equation of the adjacent two points in the X-axis direction is:

[0067] ​​​(5)

[0068] will be divided into , intermediate data points to compensate for data, set , then the X-axis direction between the two data of any point , the coordinates are:

[0069] (6)

[0070] In the formula, represent data in the first data points, belongs to N+;

[0071] As shown in Figure 3 , similarly, the minimum offset between the known points in the X direction is calculated , in order to meet the uniform distribution of data and the equal step splicing of later data, then the number of interpolation points between points and should be , that is, the number of interpolation points between the jth adjacent two points when Y value is equal should be , so the coordinates of any point in the data stream parallel to the X-axis when Y value is equal are

[0072] (7)

[0073] where , , , i=1,2,3... , i, j belong to N+;

[0074] As described above, the discrete point cloud data in the infrared camera coordinate system is interpolated and fitted in the Y-axis direction, and the data stream parallel to the Y-axis is supplemented, then the number of interpolation points between any two adjacent points is , then the coordinates of any point are:

[0075] (8)

[0076] where , , =1,2,3....n+1, , j, all belong to N+; Figure 4 is a point cloud splicing diagram of cut tobacco surface; ​

[0077] The gradient-based segmentation interpolation algorithm is a method for interpolating and subdividing the missing part on the basis of a large number of discrete point data on the surface of the cut tobacco, and the purpose is to find the corresponding source for subsequent data splicing to realize the closest fitting of the surface state of the cut tobacco; the discrete point cloud data on the surface of the cut tobacco are points without continuity in space, and when the data are spliced, the jumping points are generated, resulting in a large fitting result error; if the minimum distance point in a radius r region of a point is taken as the splicing point, the surface fitting error in the region will be large under the condition of data missing, and the directionality of the data fitting is not easy to judge; the present application performs data transmission and splicing on the discrete point cloud data in the X-axis and Y-axis directions shown in the figures, so that the data change trends do not affect each other; when the selected value is large enough, the smaller the grid is, the more full the fitting result is, and the closer to the surface state of the entity Figure 2 、 3 When the selected value is large enough, the smaller the grid is, the more full the fitting result is, and the closer to the surface state of the entity When the selected value is large enough, the smaller the grid is, the more full the fitting result is, and the closer to the surface state of the entity

[0078] 4) The point cloud data image formed by the obtained irregular cylinder set is subjected to edge detection by using a sobel operator, and the boundary points in the point cloud data image are removed to obtain a cut tobacco entity graph including the cut tobacco surface point cloud data image set; the specific process of the sobel operator for image boundary detection processing is as follows:

[0079] Let M represent the two-dimensional point cloud data information on the XY plane, and the data in M are subjected to horizontal edge detection and vertical edge detection, and the specific processing is as follows:

[0080] (9)

[0081] (10)

[0082] (11)

[0083] The gradient size G is obtained by using the above formula (11) in combination with the horizontal gradient value and the vertical gradient value of each point in the point cloud data; when the gradient value G is greater than a set threshold value, it is determined that this is an edge, that is, the critical surface edge position, and the point cloud data of the determined boundary points are subjected to 0 processing, so that the cut tobacco surface point cloud splicing entity is obtained, as shown in Figure 5

[0084] 5) The volume V of the cut tobacco entity graph is calculated, and the change amount of the volume of the cut tobacco is used to monitor the cut tobacco separation effect of the system; specifically, as shown in Figure 6 ​As shown, the cut tobacco entity surface data set is sequentially divided into multiple rows along the Y-axis direction, and each row includes several prisms with different heights. The volume sum of the several prisms in each row is calculated, and the volume sum of the several prisms in each row is accumulated to obtain the volume of the cut tobacco entity graph. The specific calculation process is as follows:

[0085] The volume of a single prism in each row after segmentation is , The specific calculation formula of is as follows:

[0086] (12)

[0087] In the above formula, represents the height of the four edges of a single prism. Then, the volume sum of the prisms in the first row after segmentation is The specific calculation formula of is as follows:

[0088] (13)

[0089] As shown above, the volume of the cut tobacco excess is represented as the cumulative sum of the volume of each row in the Y direction, i.e., the volume V is:

[0090] (14).

[0091] The change in the volume of the cut tobacco is used to monitor the cut tobacco separation effect of the system. After obtaining the point cloud data, it is transmitted to the industrial computer, the above three-dimensional reconstruction algorithm is stored in the computer, and multiple training is performed to realize data update fitting. The three-dimensional reconstruction image and the volume calculation result are displayed on the display screen. The update frequency of the volume is based on the image acquisition time designed by us, and the normal range volume threshold value under the set experience. The cut tobacco separation effect can be judged according to the total volume mutation, or according to the volume mutation under the unit step distance. When the total volume is greater than the threshold value set by us, it is judged that the cut tobacco separation effect is not complete, and a warning prompt is given.

[0092] When the change in the volume of the cut tobacco is used to monitor the cut tobacco separation effect of the system, the judgment effect can be designed as follows: a warning button is used to prompt the equipment management personnel to timely investigate the quality accident. When the computer recognition result shows correct, the indicator light will display green light; when the computer recognition result shows that the cut tobacco flow is larger but less than the warning value, the indicator light will display yellow light; when the computer recognition result shows that the cut tobacco flow exceeds the warning value, the indicator light will display red light; when the volume mutation under the unit step distance reaches the warning value, the indicator light will display blue light; if the equipment needs to be shut down due to some special situation, the warning prompt can be manually turned off.

[0093] The technical scheme relies on a non-contact measurement means of stereovision, establishes a binocular imaging mechanism by using a binocular vision difference theory, and performs coordinate system conversion, so that the position relationship between a measured object and an infrared camera can be quickly and accurately calibrated, so that the infrared camera can accurately identify the coordinate information of reflected light, and the accuracy of collected point cloud data is provided; based on a gradient segmentation interpolation algorithm, that is, on the basis of a large number of discrete point data on the surface of the cut tobacco, interpolation and subdivision are performed on the missing part, a corresponding source is found for subsequent data splicing, closest fitting of the surface state of the cut tobacco is realized, and the calculation accuracy of the change amount of the cut tobacco volume is ensured; by using the imaging characteristics of the speckle structured light, static calibration is converted into multi-directional unlimited area calibration, finally the depth information of the image surface is obtained, the change amount of the cut tobacco volume is provided, the cut tobacco separation effect of the system is monitored by using the change amount of the cut tobacco volume; the measurement time is efficient, and efficient monitoring conditions are provided for cut tobacco remaining amount data collection and simulation in the falling material vibration tank.

[0094] It is apparent for those skilled in the art that the present application is not limited to the details of the above-described exemplary embodiments, but can be implemented in other concrete forms without departing from the spirit or essential characteristics of the present application. Therefore, the embodiments should be considered in all aspects as illustrative and not restrictive, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended to encompass all changes falling within the meaning and range of equivalents of the claims. Any reference signs in the claims should not be considered as limiting the claims involved.

[0095] In addition, it should be understood that although the present specification is described in terms of embodiments, not every embodiment contains only one independent technical solution, and the description of the specification is only for the sake of clarity, and those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can be combined appropriately to form other embodiments that those skilled in the art can understand.

Claims

1. A cut filler allowance monitoring method based on point cloud imaging, characterized in that The method comprises the following steps: 1) obtaining point cloud data of the surface of the cut tobacco: using a light emitter with speckle structure to project light on the surface of the cut tobacco to be measured, and collecting the reflected speckle structure light by an infrared camera to obtain point cloud coordinate information of the cut tobacco in a three-dimensional image coordinate system; 2) converting the discrete point cloud coordinate information of the cut tobacco in the three-dimensional image coordinate system into discrete point cloud coordinate information in the infrared camera coordinate system; The infrared camera is symmetrically provided with two infrared cameras, and the specific process of converting the discrete point cloud coordinate of the cut tobacco in the three-dimensional image coordinate system into the discrete point cloud coordinate in the infrared camera coordinate system is as follows: The optical axis direction of the left and right infrared cameras is defined as The optical center position of the light under the optical system of the infrared camera is defined as the origin of the axis The horizontal and vertical axes in the infrared camera coordinate system are represented by and respectively, and and are parallel to the X and Y axes of the three-dimensional image coordinate system respectively, point B is the point coordinate of the object in the infrared camera coordinate system, , are the imaging point coordinates of point P in the three-dimensional image coordinate system collected by the left and right infrared cameras respectively, the distance between the optical center and the three-dimensional image coordinate origin is the focal length , the distance between the two infrared cameras mounted on the same plane is b, so that point has the same Y axis coordinate in the three-dimensional image coordinate system and the infrared camera coordinate system, i.e. According to the triangle principle, the following mapping relationship can be obtained: ， （1） （2） （3） （4） wherein representing points The coordinate parallax value of the corresponding points in the left and right infrared cameras can be obtained from equation (4) as The corresponding coordinates in the infrared camera coordinate system are ; 3) using a gradient segmentation interpolation algorithm to sequentially perform X-axis direction interpolation fitting and Y-axis direction interpolation fitting on the obtained discrete point cloud data set M in the infrared camera coordinate system, and then fitting the direction data obtained by the X-axis direction interpolation fitting and the direction data obtained by the Y-axis direction interpolation fitting to form an irregular prism data set with a square bottom surface and close to the state of the surface of the cut tobacco entity; 4) using a sobel operator to perform edge detection on the point cloud data image formed by the obtained irregular prism set, remove the boundary points in the point cloud data image, and obtain a cut tobacco entity graph including the cut tobacco surface point cloud data image set; 5) calculating the volume V of the cut tobacco entity graph, and using the change amount of the volume of the cut tobacco to monitor the cut tobacco separation effect of the system; The cut tobacco entity surface data set is sequentially divided into multiple rows along the Y-axis direction, and each row includes a plurality of prisms with different heights, the volume sum of the plurality of prisms in each row is calculated, and the volume sum of the plurality of prisms in each row is accumulated to obtain the volume of the cut tobacco entity graph, and the specific calculation process is as follows: Let the volume of a single prism in each row after segmentation be , The specific calculation formula is as follows: （12） In the above formula, represents the height of four edges on a single prism, then the volume of each prism in the first row after the division is The specific calculation formula is as follows: （13） As shown above, the volume of the cut tobacco excess amount is represented as the cumulative sum of the volume of each row in the Y direction change, that is, the volume V is: （14）。 2. The point cloud imaging based shreds allowance monitoring method according to claim 1, characterized in that: In the above step 3), the specific process of performing X-axis direction interpolation fitting on the discrete point cloud data in the infrared camera coordinate system is as follows: The discrete point cloud data of the filament surface consists of spatially discontinuous points. When performing X-axis interpolation fitting and stitching on discrete point cloud data, the evolution of data in the Y-axis direction is not considered; only the data in the Y-axis direction is considered. shaft and Changes in data on the axis, i.e., selecting the starting point in the X-axis direction. Then point The fit must satisfy That is, the next fitting point on the X-axis is the point with the same Y value as the previous point in the direction parallel to the X-axis. Fit the nearest point to satisfy the condition. When the data is not interpolated and supplemented, The smaller the value, the closer the fitting result is to the original surface. At this time, the point... Time The gradient value is Two adjacent points along the X-axis ,point The equation of the fitted line is: （5） Will , Divided into the middle Data compensation is performed using data points, assuming... Given any point along the X-axis between two data points... The coordinates are: （6） In the above formula, represent the first data, belongs to N+; Similarly, the known points With The minimum offset in the X direction To meet the uniform distribution of data and the equal step splicing of later data, the number of interpolation points between points And Should be That is, the number of interpolation points between the jth adjacent two points when Y value is equal should be , so the coordinates of any point in the data stream parallel to the X axis when Y value is equal are （7） wherein , , , = 1, 2, 3,... i, j belong to N+; As the above steps for the infrared camera coordinate system in the discrete point cloud data Y axis direction interpolation fitting, parallel to the Y axis on the data flow is supplemented, then the interpolation number between any two adjacent points is , then the coordinates of any point ​ （8） wherein , , = 1, 2, 3... + 1, , j, all belong to N+.

3. The point cloud imaging based shreds allowance monitoring method according to claim 1, wherein: In the above step 4), the specific process of the sobel operator for image boundary detection processing is as follows: Let M represent two-dimensional point cloud data information on the XY plane, the data in M is subjected to horizontal edge detection and vertical edge detection, and the specific processing is as follows: （9） （10） （11） The gradient value G is obtained by using the above formula (11) in combination with the lateral gradient value and the longitudinal gradient value of each point When the gradient value G is greater than a set threshold value, it is determined that the position is an edge of the critical surface, and the point cloud data of the boundary point is set to 0, thereby obtaining the point cloud splicing entity of the cut tobacco surface.​

Images