Tobacco leaf feeding uniformity control method and system based on tobacco leaf drop point prediction

Abstract

The invention discloses a tobacco leaf feeding uniformity control method and system based on tobacco leaf drop point prediction. The method comprises the steps that continuous videos of tobacco leaves in the feeding process in a feeding roller are collected; according to the collected continuous videos, obtaining a historical motion track of the tobacco leaves; predicting the position of a leaf falling point at the next moment based on the historical motion trail of the tobacco leaves; and adjusting the aiming center of the nozzle of the tobacco leaf feeding system according to the predicted position of the leaf falling point at the next moment. According to the tobacco leaf feeding uniformity control method and system based on tobacco leaf drop point prediction, the drop point position of the tobacco leaf at the next moment is accurately predicted by observing and analyzing the dynamic behavior of the tobacco leaf in the feeding roller in real time in the blanking process, and the injection position of the feeding nozzle is guided by using the prediction result, so that the feeding uniformity of the tobacco leaf is improved. And it is ensured that the feed liquid is more accurately applied to the target tobacco leaves, so that intelligent and precise control over the feeding process is achieved, and the uniformity and stability of tobacco leaf feeding are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the technical field of tobacco processing, and more particularly to a method and system for controlling tobacco leaf feeding uniformity based on tobacco leaf landing point prediction. Background Art

[0002] During the tobacco leaf feeding process, the leaves enter the feeding drum through the feed port. Inside the drum, the leaves are subjected to centrifugal force and the mechanical action of the rake spikes, causing them to flip and roll. The leaf's movement, falling speed, and final landing point have a decisive influence on the uniformity of the subsequent atomization applied by the nozzle.

[0003] In the existing technology, if the posture and landing point of tobacco leaves during the dropping process are not accurately grasped and predicted, and the nozzle atomization control is not precise enough, the following problems may arise: 1. Uneven feeding: some tobacco leaves receive too much liquid, and some tobacco leaves receive too little, affecting the consistency of the sensory quality and chemical composition of the final product; 2. Adhesion to the drum wall: the liquid may be sprayed excessively on the inner wall of the drum due to improper injection timing or position, resulting in material waste and difficulty in cleaning the equipment; 3. Impact on subsequent processing: Uneven feeding may make it difficult to control the process parameters of subsequent baking, shredding and other processes, affecting the overall processing quality.

[0004] Therefore, there is an urgent need for a tobacco feeding uniformity control method and system based on tobacco leaf landing point prediction. Summary of the Invention

[0005] The purpose of the present invention is to provide a tobacco leaf feeding uniformity control method and system based on tobacco leaf landing point prediction to solve the problems in the above-mentioned prior art. It can accurately observe and predict the falling posture and landing point of tobacco leaves in the feeding drum, which has important practical significance for realizing intelligent control of nozzle injection position, improving feeding uniformity and stability, and reducing production problems.

[0006] The present invention provides a tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction, which includes:

[0007] Collect continuous video of tobacco leaves falling into the feeding drum;

[0008] Obtaining a historical movement trajectory of the tobacco leaf based on the collected continuous video;

[0009] Based on the historical movement trajectory of the tobacco leaves, predicting the location of the leaf fall point at the next moment;

[0010] According to the predicted position of the leaf falling point at the next moment, the aiming center of the nozzle of the tobacco feeding system is adjusted.

[0011] In the above-mentioned tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction, preferably, the continuous video of the tobacco leaf falling process in the feeding drum is collected, including:

[0012] Use a high-speed camera or industrial camera to capture continuous video of the tobacco leaves falling into the feeding drum;

[0013] Extracting the continuous video frame by frame into an image sequence;

[0014] Preprocessing is performed on each frame of the image sequence.

[0015] In the above-mentioned tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction, preferably, the preprocessing of each frame image in the image sequence includes:

[0016] Performing a binarization process on each frame of the image sequence: using a global threshold segmentation method to segment the image into a foreground and a background, wherein the foreground is the tobacco leaf, so as to extract the binary contour area of the tobacco leaf;

[0017] Mask operation: using the binary contour area as a mask and performing a pixel-by-pixel logical AND operation with the original image to generate a pure image containing only the tobacco leaf area and removing background interference.

[0018] In the above-mentioned tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction, preferably, obtaining the tobacco leaf historical motion trajectory based on the collected continuous video includes:

[0019] Determining the current tobacco leaf landing point based on each frame of the collected continuous video;

[0020] The current tobacco leaf landing points corresponding to each frame image in the collected continuous video are connected in chronological order to obtain the historical movement trajectory of the tobacco leaf.

[0021] In the above-mentioned tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction, preferably, determining the current tobacco leaf landing point based on each frame image in the collected continuous video includes:

[0022] The centroid of the tobacco leaf outline with the largest area in the current frame is taken as the current tobacco leaf landing point.

[0023] In the above-mentioned tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction, preferably, the predicting of the leaf landing point position at the next moment based on the tobacco leaf historical movement trajectory includes:

[0024] The second-order polynomial model was used to fit and predict the horizontal and vertical coordinates of the leaf-falling points respectively.

[0025] In the above-mentioned tobacco leaf charging uniformity control method based on tobacco leaf drop point prediction, preferably, the second-order polynomial model is used to fit and predict the horizontal coordinate and vertical coordinate of the leaf drop point, respectively, including:

[0026] According to the prediction error table under different numbers of leaf-falling points, determine the number of historical leaf-falling points used to predict the position of the leaf-falling point at the next moment;

[0027] The least squares method was used to fit the horizontal and vertical coordinate series of the number of historical leaf-falling points into second-order polynomial models.

[0028] In the above-mentioned tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction, preferably, the predicting of the leaf landing point position at the next moment based on the tobacco leaf historical movement trajectory includes:

[0029] The regression / filtering method based on the historical point sequence is used to fit and predict the horizontal and vertical coordinates of the leaf falling points respectively.

[0030] In the above-mentioned tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction, preferably, adjusting the aiming center of the nozzle of the tobacco leaf feeding system according to the predicted leaf landing point position at the next moment includes:

[0031] The nozzle control system drives the nozzle actuator according to the predicted leaf falling point position at the next moment, adjusts the nozzle aiming center, and makes the nozzle align with the leaf falling point position corresponding to the predicted leaf falling point at the next moment on the image plane;

[0032] The nozzle control system controls the nozzle to spray the liquid.

[0033] The present invention also provides a tobacco leaf feeding uniformity control system based on tobacco leaf landing point prediction using the above method, comprising:

[0034] Tobacco leaf video acquisition module, used to collect continuous video of the tobacco leaves falling into the feeding drum;

[0035] A tobacco leaf historical motion trajectory determination module, configured to obtain the tobacco leaf historical motion trajectory based on the collected continuous video;

[0036] A leaf-falling point position prediction module, configured to predict the leaf-falling point position at the next moment based on the historical movement trajectory of the tobacco leaves;

[0037] The feeding adjustment module is used to adjust the aiming center of the nozzle of the tobacco feeding system according to the predicted leaf falling point position at the next moment.

[0038] The present invention provides a tobacco leaf feeding uniformity control method and system based on tobacco leaf landing point prediction. By real-time observation and analysis of the dynamic behavior of tobacco leaves in the feeding drum during the dropping process, the landing point of the tobacco leaves at the next moment is accurately predicted, and the prediction result is used to guide the spray position of the feeding nozzle to ensure that the feed liquid is applied to the target tobacco leaves more accurately, thereby realizing intelligent and precise control of the feeding process, significantly improving the uniformity and stability of tobacco leaf feeding, ensuring the smooth operation of subsequent processing steps, and thus ensuring the constant quality of the final tobacco product. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described below with reference to the accompanying drawings, in which:

[0040] Figure 1 A flow chart of an embodiment of a method for controlling tobacco leaf feeding uniformity based on tobacco leaf landing point prediction provided by the present invention;

[0041] Figure 2 The following are three consecutive frames of original images and the effects after binarization and masking;

[0042] Figure 3 is the local centroid and comprehensive centroid of all tobacco leaf contours in three consecutive frames of images;

[0043] Figure 4 To identify the largest tobacco leaf outline and its centroid in three consecutive frames of images;

[0044] Figure 5 The motion trajectory drawn based on a series of maximum tobacco leaf contour centroids and the position of the next leaf drop point predicted based on the trajectory;

[0045] Figure 6 This is a structural block diagram of an embodiment of a tobacco leaf feeding uniformity control system based on tobacco leaf landing point prediction provided by the present invention. DETAILED DESCRIPTION

[0046] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is merely illustrative and is in no way intended to limit the present disclosure, its application, or use. The present disclosure can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to make the present disclosure thorough and complete and to fully convey the scope of the present disclosure to those skilled in the art. It should be noted that unless otherwise specifically stated, the relative arrangement of parts and steps, the composition of materials, numerical expressions, and numerical values set forth in these embodiments should be interpreted as being merely exemplary and not as limiting.

[0047] The terms "first," "second," and similar terms used in this disclosure do not indicate any order, quantity, or importance, but are simply used to distinguish different parts. Terms such as "include" or "comprising" mean that the elements preceding the term include the elements listed after the term, and do not exclude the possibility of also including other elements. Terms such as "upper," "lower," and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0048] In the present disclosure, when a specific component is described as being located between a first component and a second component, there may or may not be an intervening component between the specific component and the first component or the second component. When a specific component is described as being connected to another component, the specific component may be directly connected to the other component without an intervening component, or may not be directly connected to the other component but have an intervening component.

[0049] All terms (including technical or scientific terms) used in this disclosure have the same meaning as those understood by one of ordinary skill in the art to which this disclosure belongs, unless otherwise specifically defined. It should also be understood that terms defined in, for example, general dictionaries should be interpreted as having a meaning consistent with their meaning in the context of the relevant technology, and should not be interpreted in an idealized or highly formal sense, unless explicitly defined herein.

[0050] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0051] like Figure 1 As shown, the tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction provided in this embodiment includes the following steps in actual implementation:

[0052] Step S1: collecting continuous video of the tobacco leaves falling into the feeding drum.

[0053] In one embodiment of the tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction of the present invention, step S1 may specifically include:

[0054] Step S11: Use a high-speed camera or an industrial camera to capture continuous video of the tobacco leaves falling into the feeding drum.

[0055] In one embodiment of the present invention, on a tobacco leaf feeding production line, an industrial camera (for example, the frame rate is set to 37fps, corresponding to 224 frames / 6 seconds) is installed above the feeding roller to capture the process of tobacco leaves falling from the feed port and entering the roller in real time.

[0056] Step S12: extract the continuous video frame by frame into an image sequence.

[0057] For example, 6 seconds of video is captured to obtain 224 frames of images. In one embodiment of the present invention, the video stream is decomposed into image sequences by frame.

[0058] Step S13: pre-process each frame of the image sequence.

[0059] Through preprocessing, a clear and reliable data basis is provided for subsequent analysis. In one embodiment of the tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction of the present invention, the step S13 may specifically include:

[0060] Step S131 , performing binarization processing on each frame image in the image sequence: using a global threshold segmentation method to segment the image into a foreground and a background, wherein the foreground is the tobacco leaf, so as to extract the binary contour area of the tobacco leaf.

[0061] For example, a preset global grayscale threshold (set to 100 according to the lighting and background conditions) is applied to each frame of the image for binarization processing to obtain a black and white image, in which the white area represents the tobacco leaves.

[0062] Step S132, mask operation: using the binary contour area as a mask, and performing a pixel-by-pixel logical AND operation with the original image to generate a pure image containing only the tobacco leaf area and removing background interference.

[0063] Through logical AND operation, a tobacco leaf image with the background shielded can be obtained.

[0064] Step S2: Obtain the historical movement trajectory of the tobacco leaves based on the collected continuous video.

[0065] In one embodiment of the tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction of the present invention, step S2 may specifically include:

[0066] Step S21: Determine the current tobacco leaf landing point based on each frame image in the collected continuous video.

[0067] The centroid of a tobacco leaf is a key geometric feature that describes its spatial distribution. It represents the geometric center of all pixels within the outline. In the specific implementation, the centroid coordinates are calculated using image moments. For a binary image I(x, y), its (p, q)-order moment is defined as: pq = Σ x Σ y x p y q I(x, y).

[0068] For example, the formula for calculating the zero-order moment is: M00 = Σ x Σ y I(x, y) (area). The formula for calculating the first moment is: M 10 =Σ x Σ y x*I(x, y), M 01 =Σ x Σ y y*I(x,y). Center of mass coordinates (x c , y c ) is calculated as: c =M 10 / M 00 ,y c =M 01 / M 00 , (where M 00 >0).

[0069] In some embodiments of the present invention, the local centroids of all independent tobacco leaf contours are calculated for the preprocessed mask image, and all local centroids are averaged to obtain a comprehensive example, such as Figure 3 As shown, the green point is the local centroid and the red point is the comprehensive centroid. Figure 3 It can be found that small tobacco leaves will cause greater interference to the results and are not conducive to representing the landing point of the main material.

[0070] In order to accurately capture the movement trend of the main tobacco leaf group, the present invention selects the tobacco leaf contour with the largest area in the current frame and uses its centroid as the current tobacco leaf landing point (Landing Point) of the frame image. The largest contour usually represents the area with the densest distribution and the most concentrated mass of tobacco leaves. Figure 4 As shown in the figure, the blue box represents the largest tobacco leaf outline in the current frame, and the red dot represents the centroid of the blue box. In the specific implementation, the contour search function of the image processing library (such as OpenCV) is used to find all independent white areas (tobacco leaf outlines), and then the area of each outline is calculated to find the outline with the largest area. Finally, the image moment of the largest outline is calculated using the function example in OpenCV to obtain the centroid coordinates (cx, cy) of the largest outline, which is the tobacco leaf landing point in the current frame. Figure 4 As shown in FIG, in three consecutive frames of images, the centroid coordinates of the extracted maximum tobacco leaf contour are (555.00, 348.00), (534.00, 369.00), and (513.00, 212.00), respectively.

[0071] Step S22: Connect the current tobacco leaf landing points corresponding to each frame image in the collected continuous video in chronological order to obtain the historical movement trajectory of the tobacco leaf.

[0072] Step S3: predicting the location of the leaf falling point at the next moment based on the historical movement trajectory of the tobacco leaves.

[0073] In one embodiment of the present invention, observing Figure 4 The motion trajectory of the leaf falling point is found to be consistent with the parabola or curve motion characteristics. Therefore, the present invention uses a second-order polynomial model to fit and predict the x and y coordinates of the leaf falling point. The second-order polynomial model is: t+1 = a x t 2 + b x t + c x ;y t+1 = a y t 2 + b y t + c y , where t is the time or frame number.

[0074] In one embodiment of the tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction of the present invention, step S3 may specifically include:

[0075] Step S31: Determine the number of historical leaf-falling points used to predict the position of the leaf-falling point at the next moment according to the prediction error table under the grouping of different numbers of leaf-falling points.

[0076] In some embodiments of the present invention, different numbers (e.g., 5 to 50) of historical leaf points are selected as a group (GroupSize), and the second-order model is used to predict the next leaf point. The predicted value is compared with the actual value, and the root mean square error (RMSE) is calculated. By comparing the Total_RMSE under different GroupSizes (see Table 1), the optimal number of groups is determined, and thus the number of historical leaf point data used for fitting and prediction is determined to achieve the best results. The calculation formulas for the root mean square error in the X direction, the root mean square error in the Y direction, and the combined root mean square error are as follows:

[0077] X_RMSE = sqrt(Σ(x pred -x actual ) 2 / N)

[0078] Y_RMSE = sqrt(Σ(y pred -y actual ) 2 / N)

[0079] Total_RMSE = sqrt(X_RMSE 2 + Y_RMSE 2 ).

[0080] According to the experimental data, when GroupSize is 47, Total_RMSE is the smallest, which is 83.31, indicating that using the first 47 leaf-falling point data for second-order model fitting to predict the next leaf-falling point location is the best effect under the experimental conditions.

[0081] Step S32: Using the least squares method, respectively fit the horizontal coordinate sequence and the vertical coordinate sequence of the number of historical leaf-falling points into a second-order polynomial model.

[0082] like Figure 5 As shown in the figure, based on the fitting curve, the predicted position of the next leaf fall point is (922.60, 274.73).

[0083] In the implementation, the calculated leaf-falling point coordinates are stored in a queue or list. This list is used to retain the latest 47 leaf-falling point data (determined based on the optimization results). When the list reaches 47 points, the x- and y-coordinate sequences of these 47 points are extracted. A second-order polynomial model is then fitted to the x- and y-coordinate sequences using least squares or a similar regression method. Using this fitted model, the leaf-falling point coordinates (xpred, ypred) for the next frame (t+1) are predicted.

[0084] Table 1 Prediction error table under different number of groups

[0085]

[0086] In another embodiment of the present invention, a regression / filtering method based on a historical point sequence is used to fit and predict the horizontal and vertical coordinates of the leaf falling point, wherein the regression / filtering method based on a historical point sequence can be, for example, a Kalman filter.

[0087] Step S4: Adjust the aiming center of the nozzle of the tobacco feeding system according to the predicted leaf falling point position at the next moment.

[0088] In one embodiment of the tobacco leaf charging uniformity control method based on tobacco leaf landing point prediction of the present invention, step S4 may specifically include:

[0089] Step S41: The nozzle control system drives the nozzle actuator according to the predicted leaf falling point position at the next moment, adjusts the aiming center of the nozzle, and aligns the nozzle with the leaf falling point position on the image plane corresponding to the predicted leaf falling point position at the next moment.

[0090] In the specific implementation, the predicted coordinates of the falling leaf point (x pred , y pred) is sent to the nozzle control system. The nozzle control system then drives the nozzle actuator (such as a two-dimensional platform controlled by a stepper motor) to adjust the nozzle aiming center so that it aligns with the corresponding (x pred , y pred )'s physical location.

[0091] Step S42: The nozzle control system controls the nozzle to spray the liquid.

[0092] Among them, the nozzle control system controls the nozzle to spray the liquid at the appropriate time (which can be determined in combination with the tobacco leaf speed or the time to reach the predicted position), which can effectively apply the liquid to the area where the tobacco leaves are most concentrated, significantly improving the uniformity of the feeding.

[0093] By repeatedly executing steps S1-S4, continuous prediction of tobacco leaf landing points and dynamic tracking and control of the nozzle can be achieved.

[0094] The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction provided by the embodiment of the present invention accurately predicts the landing point of the tobacco leaf at the next moment by real-time observation and analysis of the dynamic behavior of the tobacco leaf during the dropping process in the feeding drum, and uses the prediction result to guide the spray position of the feeding nozzle to ensure that the feed liquid is applied to the target tobacco leaf more accurately, thereby realizing intelligent and precise control of the feeding process, significantly improving the uniformity and stability of tobacco leaf feeding, ensuring the smooth operation of subsequent processing steps, and thus ensuring the constant quality of the final tobacco product.

[0095] Accordingly, if Figure 6 As shown, the present invention also provides a tobacco leaf feeding uniformity control system based on tobacco leaf landing point prediction, comprising:

[0096] Tobacco leaf video acquisition module 101, used to collect continuous video of the tobacco leaf falling process in the feeding drum;

[0097] The tobacco leaf historical motion trajectory determination module 102 is configured to obtain the tobacco leaf historical motion trajectory based on the collected continuous video;

[0098] The leaf-falling point position prediction module 103 is used to predict the leaf-falling point position at the next moment based on the historical movement trajectory of the tobacco leaves;

[0099] The feeding adjustment module 104 is used to adjust the aiming center of the nozzle of the tobacco feeding system according to the predicted leaf falling point position at the next moment.

[0100] The tobacco leaf feeding uniformity control system based on tobacco leaf landing point prediction provided by the embodiment of the present invention accurately predicts the landing point of the tobacco leaf at the next moment by real-time observation and analysis of the dynamic behavior of the tobacco leaf during the dropping process in the feeding drum, and uses the prediction result to guide the spray position of the feeding nozzle to ensure that the feed liquid is applied to the target tobacco leaf more accurately, thereby realizing intelligent and precise control of the feeding process, significantly improving the uniformity and stability of tobacco leaf feeding, ensuring the smooth operation of subsequent processing steps, and thus ensuring the constant quality of the final tobacco product.

[0101] Thus far, various embodiments of the present disclosure have been described in detail. To avoid obscuring the concept of the present disclosure, some details known in the art have not been described. Based on the above description, those skilled in the art can fully understand how to implement the technical solutions disclosed herein.

[0102] Although some specific embodiments of the present disclosure have been described in detail through examples, those skilled in the art will understand that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. Those skilled in the art will understand that the above embodiments may be modified or some technical features may be replaced with equivalents without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction, characterized in that: include: Collect continuous video of tobacco leaves falling into the feeding drum; Obtaining a historical movement trajectory of the tobacco leaf based on the collected continuous video; Based on the historical movement trajectory of the tobacco leaves, predicting the location of the leaf fall point at the next moment; According to the predicted position of the leaf falling point at the next moment, the aiming center of the nozzle of the tobacco feeding system is adjusted.

2. The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction according to claim 1, characterized in that: The continuous video of the tobacco leaves falling into the feeding drum includes: Use a high-speed camera or industrial camera to capture continuous video of the tobacco leaves falling into the feeding drum; Extracting the continuous video frame by frame into an image sequence; Preprocessing is performed on each frame of the image sequence.

3. The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction according to claim 2, characterized in that: The preprocessing of each frame of the image sequence includes: Performing a binarization process on each frame of the image sequence: using a global threshold segmentation method to segment the image into a foreground and a background, wherein the foreground is the tobacco leaf, so as to extract the binary contour area of the tobacco leaf; Mask operation: using the binary contour area as a mask and performing a pixel-by-pixel logical AND operation with the original image to generate a pure image containing only the tobacco leaf area and removing background interference.

4. The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction according to claim 1, characterized in that: The method of obtaining the historical movement trajectory of tobacco leaves based on the collected continuous video includes: Determining the current tobacco leaf landing point based on each frame of the collected continuous video; The current tobacco leaf landing points corresponding to each frame image in the collected continuous video are connected in chronological order to obtain the historical movement trajectory of the tobacco leaf.

5. The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction according to claim 4, characterized in that: Determining the current tobacco leaf landing point based on each frame image in the collected continuous video includes: The centroid of the tobacco leaf outline with the largest area in the current frame is taken as the current tobacco leaf landing point.

6. The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction according to claim 1, characterized in that: The method of predicting the location of the leaf falling point at the next moment based on the historical movement trajectory of the tobacco leaf includes: The second-order polynomial model was used to fit and predict the horizontal and vertical coordinates of the leaf-falling points respectively.

7. The method for controlling tobacco leaf feeding uniformity based on tobacco leaf landing point prediction according to claim 6, characterized in that: The second-order polynomial model is used to fit and predict the horizontal coordinate and vertical coordinate of the leaf falling point, including: According to the prediction error table under different numbers of leaf-falling points, determine the number of historical leaf-falling points used to predict the position of the leaf-falling point at the next moment; The least squares method was used to fit the horizontal and vertical coordinate series of the number of historical leaf-falling points into second-order polynomial models.

8. The tobacco leaf feeding uniformity control method based on tobacco leaf landing point prediction according to claim 1, characterized in that: The method of predicting the location of the leaf falling point at the next moment based on the historical movement trajectory of the tobacco leaf includes: The regression / filtering method based on the historical point sequence is used to fit and predict the horizontal and vertical coordinates of the leaf falling points respectively.

9. The method for controlling tobacco leaf feeding uniformity based on tobacco leaf landing point prediction according to claim 1, characterized in that: The step of adjusting the aiming center of the nozzle of the tobacco feeding system according to the predicted leaf drop point position at the next moment includes: The nozzle control system drives the nozzle actuator according to the predicted leaf falling point position at the next moment, adjusts the nozzle aiming center, and makes the nozzle align with the leaf falling point position corresponding to the predicted leaf falling point at the next moment on the image plane; The nozzle control system controls the nozzle to spray the liquid.

10. A tobacco leaf feeding uniformity control system based on tobacco leaf landing point prediction using the method according to any one of claims 1 to 9, characterized in that: include: Tobacco leaf video acquisition module, used to collect continuous video of the tobacco leaves falling into the feeding drum; A tobacco leaf historical motion trajectory determination module, configured to obtain the tobacco leaf historical motion trajectory based on the collected continuous video; A leaf-falling point position prediction module, configured to predict the leaf-falling point position at the next moment based on the historical movement trajectory of the tobacco leaves; The feeding adjustment module is used to adjust the aiming center of the nozzle of the tobacco feeding system according to the predicted leaf falling point position at the next moment.

Images