A hay cutting device and method for gardening

Abstract

The present invention relates to the technical field of image region merging and simplification analysis, and particularly relates to a hay cutting device and method for gardens. The present invention first obtains a hay image and determines whether to adjust the device; further, when it is determined to adjust the device, according to the number distribution and edge intensity of pixel points of each gray level, the difference in the number of pixel points between each gray level and the adjacent gray level, and in combination with the gray difference between the pixel points of each gray level and the pixel points within a preset neighborhood, gray level merging is performed on the current hay image; further, the texture complexity is obtained based on the gray level co-occurrence matrix of the current merged hay image; and the hay cutting device is adjusted according to the texture complexity. By analyzing the number distribution, edge intensity and local gray difference of pixel points of gray levels, the present invention merges gray levels to reduce the calculation complexity, adjusts the hay cutting device in real time, reduces the energy and equipment loss costs, ensures the mowing effect, and reduces the time cost.

Description

Technical Field

[0001] The present invention relates to the technical field of image region merging and simplification analysis, and particularly relates to a hay cutting device and method for gardens. Background Art

[0002] In garden maintenance, cutting hay is usually to keep the landscape clean and beautiful, and at the same time, it helps the healthy growth of plants, maintains the ecological environment, and reduces the fire hazard.

[0003] The main working principle of the existing mowing devices is as follows: The mowing device moves forward through the crawler, and the cutter group rotates for cutting and crushing to process it into grass residue. However, the existing mowing devices often adopt a mowing method with a fixed power. For dry grasslands with different degrees of lushness, only one mowing mode is used, which will lead to waste of energy, increased equipment wear, or unsatisfactory mowing effect, and the time cost will increase due to re-mowing. Summary of the Invention

[0004] In order to solve the technical problem of the high cost of the existing mowing devices, the purpose of the present invention is to provide a hay cutting device and method for gardens, and the specific technical solutions adopted are as follows:

[0005] A hay cutting method for gardens, the method includes:

[0006] Obtain a hay image; determine whether to adjust the device according to the difference features between the current hay image and the historical adjacent hay images;

[0007] When it is determined to adjust the device, obtain the merging acceptance degree of each gray level according to the number distribution and edge intensity of the pixel points of each gray level; obtain the merging priority degree of each gray level according to the number difference of the pixel points between each gray level and the adjacent gray levels, and combine the gray difference between the pixel points of each gray level and the pixel points within the preset neighborhood; perform gray level merging on the current hay image according to the merging acceptance degree and the merging priority degree of each gray level;

[0008] Obtain the texture complexity based on the gray level co-occurrence matrix of the current merged hay image; adjust the hay cutting device according to the texture complexity.

[0009] Further, the method for determining whether to adjust the device includes:

[0010] Obtain the adjustment necessity according to the cosine similarity between the current hay image and the historical adjacent hay images; the adjustment necessity is negatively correlated with the cosine similarity;

[0011] When the adjustment necessity is greater than the preset adjustment threshold, it is determined that the device needs to be adjusted currently.

[0012] Further, the method for obtaining the merging acceptance includes:

[0013] Obtain the merging acceptance of the corresponding gray level according to the difference between the number of pixel points of each gray level and the average value of the number of pixel points of all gray levels, in combination with the ratio of the number of edge pixel points to non-edge pixel points within the corresponding gray level; the difference between the number of pixel points of each gray level and the average value of the number of pixel points of all gray levels is positively correlated with the merging acceptance; the ratio of the number of edge pixel points to non-edge pixel points is negatively correlated with the merging acceptance.

[0014] Further, the method for obtaining the merging priority includes:

[0015] Select any one of the gray levels as the target gray level;

[0016] Take the maximum ratio of the number of pixel points of adjacent gray levels to the number of pixel points of the target gray level as the numerator, take the average value of the squares of the differences between the gray values of the pixel points within the target gray level and the gray values of the pixel points within the preset neighborhood as the denominator, and take the fractional ratio as the merging priority of the target gray level.

[0017] Further, the method for merging the gray levels of the current hay image includes:

[0018] Obtain the merging possibility of each gray level according to the merging acceptance and the merging priority of each gray level; both the merging acceptance and the merging priority are positively correlated with the merging possibility;

[0019] When the merging possibility is greater than the preset merging threshold, determine that the corresponding gray level needs to be merged, and merge the corresponding gray level into the gray level with the most adjacent pixel points.

[0020] Further, the method for obtaining the texture complexity includes:

[0021] Obtain the texture complexity according to the entropy and contrast obtained from the gray-level co-occurrence matrix; both the entropy and the contrast are positively correlated with the texture complexity.

[0022] Further, the method for adjusting the hay cutting device according to the texture complexity includes:

[0023] Obtain the current moving speed according to the no-load speed of the hay cutting device and the texture complexity; the no-load speed is positively correlated with the current moving speed; the texture complexity is negatively correlated with the current moving speed;

[0024] Obtain a power increment based on the cutting power range difference of the hay cutting device and the texture complexity; both the cutting power range difference and the texture complexity are positively correlated with the power increment; use the sum of the minimum cutting power and the power increment as the current cutting power.

[0025] Further, the preset neighborhood is an eight-neighborhood.

[0026] Further, obtain the edge pixel points through the Canny operator.

[0027] The present invention also proposes a hay cutting device for gardening, the device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and when the processor executes the computer program, it implements the steps of any one of the hay cutting methods for gardening.

[0028] The present invention has the following beneficial effects:

[0029] The present invention first obtains a hay image to obtain an analysis basis; determines whether to adjust the device according to the difference features between the current hay image and the historical adjacent hay images, avoiding overly frequent adjustment of the device and increasing the working stability of the device; further when it is determined to adjust the device, taking advantage of the fact that the main part of the hay changes relatively smoothly and the edge contains texture features, according to the number distribution and edge intensity of the pixel points at each gray level, obtain the combined acceptance degree of each gray level, which represents the texture loss degree when the gray levels are combined, providing a basis for subsequent gray level combination; further based on the fact that the difference in the number of pixel points between a gray level and its adjacent gray levels represents the influence degree of the image structure during combination, and the gray difference between a pixel point and the pixel points within the preset neighborhood represents the local texture complexity of the pixel point, obtain the combination priority degree of each gray level, which represents the priority degree of the gray level being combined, providing more basis for subsequent gray level combination; further according to the combined acceptance degree and combination priority degree of each gray level, perform gray level combination on the current hay image, reducing the image texture loss while reducing the gray complexity of the image and saving computing resources; finally, obtain the texture complexity based on the gray level co-occurrence matrix of the current combined hay image; adjust the hay cutting device according to the texture complexity. The present invention analyzes the number distribution, edge intensity, and local gray difference of pixel points at each gray level, combines gray levels to reduce the computational complexity, adjusts the hay cutting device in real time, reduces the energy and equipment loss costs, ensures the mowing effect, and reduces the time cost. Description of the Drawings

[0030] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can be obtained based on these drawings.

[0031] Figure 1 Schematic diagram of the overall structure of a hay cutting device for gardens provided by an embodiment of the present invention;

[0032] Figure 2 Schematic diagram of the internal structure of a hay cutting device for gardens provided by an embodiment of the present invention;

[0033] Figure 3 Schematic diagram of the drum installation structure of a hay cutting device for gardens provided by an embodiment of the present invention;

[0034] Figure 4 Schematic diagram of the drum structure of a hay cutting device for gardens provided by an embodiment of the present invention;

[0035] Figure 5 Schematic diagram of the tool structure of a hay cutting device for gardens provided by an embodiment of the present invention;

[0036] Figure 6 Flowchart of a hay cutting method for gardens provided by an embodiment of the present invention.

[0037] The reference numerals in the figure are: 1, arc-shaped grass grabbing claw; 2, drum; 3, fully enclosed drum baffle; 4, tool; 6, negative pressure absorber; 8, iron box; 9, gland; 10, side drawer of the iron box; 11, first camera; 12, second camera; 13, housing; 14, power supply; 15, motor; 16, transmission shaft; 17, gear transmission device; 18, first wheel; 19, second wheel; 20, third wheel; 21, fourth wheel; 41, detection system; 42, adjustment and control system. Detailed implementation manners

[0038] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the intended invention purpose, the following, in combination with the drawings and preferred embodiments, details the specific implementation manners, structures, features and effects of a hay cutting device and method for gardens proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. In addition, the specific features, structures or characteristics in one or more embodiments can be combined in any suitable form.

[0039] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which this invention belongs.

[0040] The following specifically describes the specific solutions of a hay cutting device and method for garden use provided by the present invention in conjunction with the accompanying drawings.

[0041] The hay cutting device for garden use consists of the following modules: a cutting module, a conveying module, a collection module, an information acquisition module, and a transmission module.

[0042] Among them, the cutting module includes an arc-shaped hay-grabbing claw 1 for grabbing hay into the roller 2; two rollers 2, similar to the rollers of a noodle press, for flattening the hay and conveying it to the cutting knife group; a fully enclosed roller baffle 3 to ensure that the hay is conveyed forward during the transmission process and does not run out of both sides; two cutting tools 4, similar to the shape of the blades of an electric fan, for cutting the hay slices conveyed by the roller 2. The specific structure is as follows: The arc-shaped hay-grabbing claw 1 is in front of the two rollers 2. The two rollers 2 are vertically arranged side by side with a small interval in the middle. The fully enclosed roller baffle 3 is arranged around the arc-shaped hay-grabbing claw 1 and the roller 2 to form a passage. The two cutting tools 4 are arranged side by side directly behind the roller 2.

[0043] The conveying module includes a negative pressure absorber 6, similar to a vacuum cleaner, for sucking the hay residue pushed to the rear into the collection container. The specific structure is as follows: The negative pressure absorber 6 is placed at the end of the passage of the conveying module. The passage of the conveying module is closely connected to the passage of the cutting module.

[0044] The collection module includes an iron box 8 for placing the hay residue sucked by the negative pressure absorber 6; a pressing cover 9 for compacting the hay residue; an iron box side drawer 10 for conveniently taking out the collected hay residue. The specific structure is as follows: The pressing cover 9 is placed directly above the iron box 8. There are holes in the lower part on the side of the iron box 8 and small holes in the upper part on the opposite side. The iron box side drawer 10 is placed at the hole of the iron box 8. The holes in the lower part on the side of the iron box 8 and the small holes in the upper part on the opposite side are used for closely connecting with the passage of the conveying module.

[0045] The information acquisition module is divided into a detection system 41 and an adjustment and control system 42. The detection system 41 includes a first camera 11 and a second camera 12, and its main purpose is to collect and monitor the dense information of the hay image during the forward movement. The housing 13. The adjustment and control system 42 mainly includes an FPGA chip for calculating the data collected in the monitoring system and adjusting the running speed, cutting power, and storage power of the device. The specific structure is as follows: The first camera 11 and the second camera 12 are located directly above the arc-shaped hay-grabbing claw 1. The adjustment and control system 42 is located directly above the outside of the passage in the conveying module, and the entire device is covered with the housing 13.

[0046] The transmission module includes a power supply 14 that provides power for the entire device; a motor 15 that drives the operation of each module; a transmission shaft 16 that connects the power source and the main shaft of the transmission device to transmit power from the power source to the transmission device; a gear transmission device 17 that transmits and adjusts power; a first wheel 18, a second wheel 19, a third wheel 20, and a fourth wheel 21 that are used for the movement of the hay cutting device. The specific structure is as follows: The power supply 14 and the motor 15 are located within the housing 13. The transmission shaft 16 is located directly below the housing 13 and on the axis of symmetry. The gear transmission device 17 is connected to the transmission shaft 16, the first wheel 18, the second wheel 19, the third wheel 20, and the fourth wheel 21.

[0047] Please refer to Figure 6 , which shows a flowchart of a method for cutting hay in a garden provided by an embodiment of the present invention. Specifically, it includes:

[0048] Step S1: Obtain a hay image; determine whether to adjust the device according to the difference features between the current hay image and the historical adjacent hay images.

[0049] When using a hay cutting device for a garden to cut garden hay, the lushness of the garden hay is not the same in each area. However, during the operation of the hay cutting device for a garden, the lushness at adjacent times may be similar. Then, the same equipment parameters can be used to meet the cutting of the hay cutting device for a garden, and it is not necessary to adjust the equipment parameters at every moment. Therefore, first obtain a hay image to obtain an analysis basis; determine whether to adjust the device according to the difference features between the current hay image and the historical adjacent hay images, and avoid adjusting the device too frequently, thereby increasing the working stability of the device.

[0050] Preferably, in an embodiment of the present invention, considering that the greater the cosine similarity between two frames of images, the closer their directions are in the vector space, the higher the similarity, and the more similar the lushness of the hay, the less need to adjust the device; at the same time, the cosine similarity calculation is efficient. Therefore, obtain the necessity for adjustment according to the cosine similarity between the current hay image and the historical adjacent hay images; the necessity for adjustment is negatively correlated with the cosine similarity;

[0051] When the necessity for adjustment is greater than a preset adjustment threshold, it is determined that the device needs to be adjusted currently.

[0052] As an example, the image acquisition interval is 5 seconds. After normalizing the pixel values of the hay image, it is expanded into a one-dimensional vector row by row. First, arrange all pixel values in the first row, then the second row, and so on; then calculate the cosine similarity between the one-dimensional vector of the current hay image and the one-dimensional vector of the historical adjacent hay image. The calculation formula for the necessity for adjustment includes:

[0053] ;

[0054] where \(t\) is the current time sequence number; represents the necessity of adjustment at the \(t\)-th moment; represents the exponential function with the natural constant \(e\) as the base; represents the cosine similarity function; represents the one-dimensional vector at the \((t - 1)\)-th moment; represents the one-dimensional vector at the \(t\)-th moment.

[0055] In the calculation formula of the necessity of adjustment, through (\(x\) represents the independent variable), a negative correlation mapping is performed on the cosine similarity to adjust the logical relationship and obtain the necessity of adjustment. The difference characteristics between the current hay image and the adjacent historical hay images are represented by the cosine similarity. The smaller the cosine similarity, the more obvious the difference between the images, and the greater the necessity of adjustment.

[0056] As an example, the preset adjustment threshold is taken as 0.8.

[0057] It should be noted that the hay images output after image stitching collected by the first camera 11 and the second camera 12 are of fixed size and grayscale. The image stitching technology of the binocular camera is already an existing technology and will not be elaborated here.

[0058] In another embodiment of the present invention, considering that the cosine similarity is sensitive to the vector direction, while the Euclidean distance is sensitive to the vector length, the difference characteristics between the hay image and the adjacent historical hay images are also represented by the Euclidean distance between the images. Specifically: calculate the Euclidean distance between the one-dimensional vector of the current hay image and the one-dimensional vector of the adjacent historical hay image, and divide it by the square root of the number of pixel points in the hay image to normalize the Euclidean distance. The average value of the normalized Euclidean distance and is used as the necessity of adjustment.

[0059] Step S2: When determining the adjustment device, obtain the merging acceptance of each gray level according to the number distribution and edge intensity of the pixel points of each gray level; obtain the merging priority of each gray level according to the difference in the number of pixel points between each gray level and the adjacent gray levels, combined with the gray difference between the pixel points of each gray level and the pixel points in the preset neighborhood; perform gray level merging on the current hay image according to the merging acceptance and merging priority of each gray level.

[0060] When determining the adjustment device, it is necessary to analyze the texture complexity in the image to analyze the lushness of the hay during current cutting, so as to adjust the hay cutting device for gardening, save energy costs and time costs, and slow down equipment wear. Considering directly using the gray-level co-occurrence matrix of the original hay image to detect texture complexity, the gray-level co-occurrence matrix needs to expand the original hay image at all gray levels, and the overall computational complexity is large, which is not suitable for real-time mowing behavior. Moreover, not all gray levels can provide texture information. Therefore, the computational complexity can be reduced by merging gray levels and computational resources can be saved.

[0061] Considering that the main parts of the hay, such as the stems and leaves, occupy most of the image area and change smoothly; the relatively few stripes, folds or edges on the hay surface mainly contain the texture features of the hay. Therefore, according to the pixel point number distribution and edge intensity of each gray level, the merging acceptance degree of each gray level can be obtained, which characterizes the texture loss degree when the gray level is merged, providing a basis for subsequent gray level merging.

[0062] Preferably, in an embodiment of the present invention, considering that in the natural image of the hay, most areas are relatively smooth, the more the number of pixel points of a gray level, the more likely it corresponds to the areas inside the stems and leaves that lack obvious texture details, and the less texture information. The smaller the texture loss degree during merging; in the hay texture, edge pixel points often appear at the edges of the stems, the folds or stripes of the leaves, and these areas are exactly where the texture features of the hay are located. The smaller the proportion of edge pixel points, the less texture information and the smaller the texture loss degree during merging;

[0063] Based on this, according to the difference between the number of pixel points of each gray level and the average value of the number of pixel points of all gray levels, combined with the ratio of the number of edge pixel points to non-edge pixel points within the corresponding gray level, the merging acceptance degree of the corresponding gray level is obtained; the difference between the number of pixel points of each gray level and the average value of the number of pixel points of all gray levels is positively correlated with the merging acceptance degree; the ratio of the number of edge pixel points to non-edge pixel points is negatively correlated with the merging acceptance degree.

[0064] As an example, the calculation formula of the merging acceptance degree includes:

[0065] ;

[0066] where g is the serial number of the gray level; represents the merging acceptance degree of the g-th gray level; e is the natural constant; represents the number of pixel points of the g-th gray level; represents the average value of the number of pixel points of the gray levels; represents the number of edge pixel points of the g-th gray level; represents the number of non-edge pixel points at the g-th gray level; C0 represents a preset positive parameter for division by zero; C1 represents a preset positive parameter for division by zero. In this example, C0 = 1 and C1 = 0.01.

[0067] In the calculation formula of the merging acceptance, through represents the distribution of the number of pixel points. The larger it is, it indicates that the number of pixel points within the current gray level is more compared to the overall number of pixel points of all gray levels, the less texture information, and the smaller the texture loss degree during merging, and the larger the merging acceptance; the edge strength is represented by the ratio of the number of edge pixel points to the number of non-edge pixel points. The larger it is, it indicates that the proportion of edge pixel points is more, the stronger the edge strength, the more texture information, the larger the texture loss degree during merging, and the smaller the merging acceptance.

[0068] In another embodiment of the present invention, considering that the gradient magnitude also reflects the edge strength of pixel points, the larger the average gradient magnitude of the pixel points at a gray level, it indicates a high edge degree and the pixel points contain richer texture information. Therefore, the edge strength can also be represented by the average gradient magnitude of the pixel points at a gray level: ; where represents the average gradient magnitude of the pixel points at the g-th gray level.

[0069] In other embodiments of the present invention, it is also possible to replace by the ratio of the number of pixel points within the g-th gray level to the total number of pixel points of all gray levels to represent the distribution of the number of pixel points.

[0070] It should be noted that in one embodiment of the present invention, the edge pixel points are obtained through the Canny operator to distinguish the edge pixel points and non-edge pixel points in each gray level; in other embodiments of the present invention, the implementer can also use existing edge detection methods such as the Sobel operator, which will not be elaborated here.

[0071] Considering that when performing gray level merging, it is also necessary to protect the continuity of the image. Considering that merging pixel points with close gray values will not cause obvious visual jumps or texture breaks, the difference in the number of pixel points between a gray level and its adjacent gray levels represents the degree of influence on the image structure during merging, and the gray difference between a pixel point and the pixel points within a preset neighborhood represents the local texture complexity of the pixel point. Therefore, according to the difference in the number of pixel points between each gray level and its adjacent gray levels, combined with the gray difference between the pixel points of each gray level and the pixel points within a preset neighborhood, the merging priority of each gray level is obtained, which characterizes the priority degree of the gray level being merged, providing more basis for subsequent gray level merging.

[0072] Preferably, in one embodiment of the present invention, any gray level is selected as the target gray level for convenient analysis one by one;

[0073] Considering that merging gray levels with fewer pixel points into adjacent gray levels with more pixel points has less impact on the overall image structure and preserves the main gray distribution characteristics; at the same time, considering that the smaller the square of the gray difference between a pixel point and the pixel points in its neighborhood, the smaller the gray difference, the stronger the continuity of the gray value distribution, and the less texture information of the pixel points corresponding to the gray level, and there will be no obvious pseudo-texture phenomenon during merging;

[0074] Based on this, the maximum ratio of the number of pixel points in an adjacent gray level to the number of pixel points in the target gray level is used as the numerator to represent the difference in the number of pixel points between the gray level and the adjacent gray level; the average value of the squares of the differences between the gray values of the pixel points in the target gray level and the gray values of the pixel points in the preset neighborhood is used as the denominator to represent the gray difference between the pixel points and the pixel points in the preset neighborhood, and the fractional ratio is used as the merging priority of the target gray level.

[0075] As an example, the preset neighborhood is an eight-neighborhood.

[0076] It should be noted that generally, for the target gray level of a natural hay image, the average value of the squares of the differences between all pixel points and the gray values of the neighboring pixel points in the preset neighborhood is not zero. When a special case of zero occurs, a denominator of a very small positive parameter such as 0.01 is given; for the case where the number of pixel points in the target gray level is 0, it is directly skipped.

[0077] In other embodiments of the present invention, the implementer can also set other preset neighborhoods, such as a four-neighborhood in a cross shape, and use the absolute value of the difference to represent the gray difference, such as using the average value of the absolute values of the differences between the gray values of the pixel points in the target gray level and the gray values of the pixel points in the preset neighborhood as the denominator.

[0078] The merging acceptance and the merging priority provide the merging basis from the perspective of texture loss and image continuity respectively. Therefore, further according to the merging acceptance and the merging priority of each gray level, the gray level of the current hay image is merged, reducing the texture loss of the image while reducing the gray complexity of the image and saving computing resources.

[0079] In a preferred embodiment of the present invention, considering that the greater the merging acceptance, the smaller the texture loss degree during merging and the higher the merging possibility; the greater the merging priority, the smaller the impact on the overall image structure and the higher the merging possibility. Therefore, according to the merging acceptance and the merging priority of each gray level, the merging possibility of each gray level is obtained; both the merging acceptance and the merging priority are positively correlated with the merging possibility;

[0080] When the merging possibility is greater than the preset merging threshold, it is determined that the corresponding gray level needs to be merged, and the corresponding gray level is merged into the gray level with the most adjacent pixel points.

[0081] As an example, after linearly normalizing the product of the merging acceptance and the merging priority, it is used as the merging possibility corresponding to the gray level; the preset merging threshold is 0.85.

[0082] For example, when g = 200, the merging possibility is 0.9, and it is determined that merging is required; when g = 199 contains 150 pixels, and g = 201 contains 300 pixels, then the gray value of the pixels with g = 200 is changed to 201, so as to merge the gray level of g = 200 into the gray level of 201.

[0083] It should be noted that in the embodiment of the present invention, all gray levels are first determined and then merged; considering that there may be continuous merging situations, such as in the determination stage, the gray level of 200 is merged into 201, the gray level of 201 is merged into the gray level of 202, and the gray level of 202 is merged into the gray level of 203. At this time, the gray values of the pixels with gray levels of 200, 201, and 202 are all changed to 203; to avoid serious loss of image texture caused by excessive continuous merging, the maximum number of continuous merges is limited to 3 times. In the above example, no matter whether the gray level of 203 is determined to be merged into the gray level of 204, no merging is performed.

[0084] It should be noted that for the special situation where two adjacent gray levels are determined to merge into each other, such as the gray level of 200 is merged into 201, and the gray level of 201 is merged into 200. At this time, the gray level with more pixels is retained. For example, the gray level of 200 has 150 pixels, and the gray level of 201 has 180 pixels, and the gray level of 201 is retained.

[0085] Step S3: Obtain the texture complexity based on the gray-level co-occurrence matrix of the currently merged hay image; adjust the hay cutting device according to the texture complexity.

[0086] After simplifying the hay image through step S2, the computational complexity of the gray-level co-occurrence matrix is saved. Furthermore, the texture complexity is obtained based on the gray-level co-occurrence matrix of the currently merged hay image, which characterizes the lushness of the hay; finally, the hay cutting device is adjusted according to the texture complexity, so that the hay cutting device adapts to different lushness situations of the hay, reduces energy waste, and ensures the mowing effect.

[0087] In a preferred embodiment of the present invention, considering that the greater the entropy obtained from the gray-level co-occurrence matrix, the more complex the image texture, and the greater the contrast, the more complex the texture. Therefore, the texture complexity is obtained based on the entropy and contrast obtained from the gray-level co-occurrence matrix; both the entropy and the contrast are positively correlated with the texture complexity.

[0088] As an example, the entropy and contrast are fused by multiplication, and the product of the entropy and contrast obtained from the gray-level co-occurrence matrix is used as the texture complexity.

[0089] As another example, the entropy and contrast are fused by weighted summation, and the average value of the entropy and contrast obtained from the gray-level co-occurrence matrix is used as the texture complexity.

[0090] It should be noted that the gray-level co-occurrence matrix and the entropy and contrast obtained based on the gray-level co-occurrence matrix are already existing technologies and will not be elaborated here.

[0091] Preferably, in an embodiment of the present invention, considering that when the hay cutting device moves without load, the speed is the highest and the cutting power is the lowest, and when cutting, the higher the texture complexity, the denser the hay, and the more the cutting power needs to be increased and the movement speed needs to be reduced;

[0092] Based on this, the current movement speed is obtained according to the no-load speed and texture complexity of the hay cutting device; the no-load speed is positively correlated with the current movement speed; the texture complexity is negatively correlated with the current movement speed;

[0093] The power increment is obtained according to the cutting power range and texture complexity of the hay cutting device; both the cutting power range and the texture complexity are positively correlated with the power increment; the sum of the minimum cutting power and the power increment is used as the current cutting power.

[0094] As an example, the calculation formula for the movement speed of the current hay cutting device includes:

[0095] ;

[0096] Wherein, represents the movement speed of the hay cutting device at the current moment t; represents the texture complexity at the current moment t; represents the no-load speed.

[0097] In the calculation formula of the movement speed, the greater the texture complexity, the denser the hay, the smaller, the smaller the movement speed of the current hay cutting device.

[0098] The calculation formula for the cutting power of the current hay cutting device includes:

[0099] ;

[0100] Wherein, represents the cutting power of the hay cutting device at the current moment t; represents the cutting power range of the hay cutting device; represents the minimum cutting power of the hay cutting device; represents the power increment of the hay cutting device at the current moment t.

[0101] In the calculation formula of the cutting power, the adjustment range of the cutting power is represented by the range of the cutting power. Based on the minimum cutting power, the cutting power is increased. The more complex the texture and the denser the hay, the greater the power increment, and the greater the cutting power of the current hay cutting device.

[0102] It should be noted that the no-load speed, minimum cutting power, and maximum cutting power of the hay cutting device during cutting can be obtained from the device instruction manual or the product nameplate, and will not be limited here.

[0103] An embodiment of the present invention also provides a hay cutting device for gardens, which includes a memory, a processor, and a computer program. The memory is used to store the corresponding computer program, the processor is used to run the corresponding computer program, and when the computer program runs in the processor, it can implement a hay cutting method for gardens described in steps S1 - S3.

[0104] The working principle of the hay cutting device for gardens is briefly described as follows:

[0105] First, turn on the power switch. Before mowing, it is in the no-load mode. When the device travels to the area to be mowed, the first camera 11 and the second camera 12 in the detection system 41 are used to collect images and transmit the information to the FPGA chip in the adjustment control system 42;

[0106] Then, the FPGA chip calculates and analyzes the hay images at each moment to determine whether to adjust the device; when it is determined to adjust the device, the hay cutting device is adjusted;

[0107] When the hay contacts the device, first, the arc-shaped hay-grabbing claws 1 grab the hay and send it into the roller 2. The roller 2 rotates to squeeze the hay into a cake shape and send it to the cutter 4. At this time, the cutter 4 cuts the fed hay cake into fragments;

[0108] The fragments cut by the cutter 4 move backward under the suction of the negative pressure absorber 6 until they fall into the iron box 8 for collection;

[0109] The pressing cover 9 presses down once every certain period of time (such as 30 seconds) under the control of the FPGA chip.

[0110] In summary, in view of the technical problem of the high cost of existing mowing devices, the present invention proposes a dry grass cutting device and method for gardens. The present invention first obtains a dry grass image and determines whether to adjust the device; further, when it is determined to adjust the device, according to the number distribution and edge strength of pixel points of each gray level, the difference in the number of pixel points between each gray level and the adjacent gray level, and combining the gray difference between the pixel points of each gray level and the pixel points within a preset neighborhood, gray level merging is performed on the current dry grass image; further, the texture complexity is obtained based on the gray level co-occurrence matrix of the current merged dry grass image; the dry grass cutting device is adjusted according to the texture complexity. By analyzing the number distribution, edge strength and local gray difference of pixel points of gray levels, the present invention merges gray levels to reduce the computational complexity, adjusts the dry grass cutting device in real time, reduces the energy and equipment loss costs, ensures the mowing effect, and reduces the time cost.

[0111] It should be noted that the above sequence of embodiments of the present invention is only for description and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the particular order shown or sequential order to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0112] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between each embodiment can be referred to each other. The key point of each embodiment is to illustrate the differences from other embodiments.

Claims

1. A method for cutting dry grass for gardens, characterized in that, The method includes: Obtain a hay image; determine whether to adjust the device according to the difference features between the current hay image and the historical adjacent hay images; When it is determined to adjust the device, obtain the merging acceptance degree of each gray level according to the quantity distribution and edge intensity of the pixel points of each gray level; obtain the merging priority of each gray level according to the quantity difference of the pixel points between each gray level and its adjacent gray levels, and combine the gray difference between the pixel points of each gray level and the pixel points within the preset neighborhood; perform gray level merging on the current hay image according to the merging acceptance degree and the merging priority of each gray level; Obtain the texture complexity based on the gray level co-occurrence matrix of the currently merged hay image; adjust the hay cutting device according to the texture complexity; The method for obtaining the merging acceptance degree includes: Obtain the merging acceptance degree of the corresponding gray level according to the difference between the quantity of pixel points of each gray level and the average value of the quantities of pixel points of all gray levels, and combine the quantity ratio of edge pixel points to non-edge pixel points within the corresponding gray level; the difference between the quantity of pixel points of each gray level and the average value of the quantities of pixel points of all gray levels is positively correlated with the merging acceptance degree; the quantity ratio of edge pixel points to non-edge pixel points is negatively correlated with the merging acceptance degree.

2. The method for cutting dry grass for garden use according to claim 1, wherein The method for determining whether to adjust the device includes: Obtain the adjustment necessity according to the cosine similarity between the current hay image and the historical adjacent hay images; the adjustment necessity is negatively correlated with the cosine similarity; When the adjustment necessity is greater than the preset adjustment threshold, it is determined that the device needs to be adjusted currently.

3. A method for cutting hay for garden use according to claim 1, characterized in that, The method for obtaining the merging priority includes: Select any one of the gray levels as the target gray level; Take the maximum ratio of the quantity of pixel points of the adjacent gray level to the quantity of pixel points of the target gray level as the numerator, take the average value of the squares of the differences between the gray values of the pixel points within the target gray level and the gray values of the pixel points within the preset neighborhood as the denominator, and take the fractional ratio as the merging priority of the target gray level.

4. A method for cutting hay for gardening according to claim 1, characterized in that, The method for performing gray level merging on the current hay image includes: Obtain the merging possibility of each gray level according to the merging acceptance degree and the merging priority of each gray level; both the merging acceptance degree and the merging priority are positively correlated with the merging possibility; When the merging possibility is greater than the preset merging threshold, it is determined that the corresponding gray level needs to be merged, and the corresponding gray level is merged into the gray level with the most adjacent pixel points.

5. A method for cutting hay for garden use according to claim 1, characterized in that, The method for obtaining the texture complexity includes: Obtain the texture complexity according to the entropy and contrast obtained from the gray level co-occurrence matrix; both the entropy and the contrast are positively correlated with the texture complexity.

6. A method for cutting hay for gardening according to claim 5, characterized in that, The method for adjusting the hay cutting device according to the texture complexity includes: Obtain the current moving speed according to the no-load speed of the hay cutting device and the texture complexity; the no-load speed is positively correlated with the current moving speed; the texture complexity is negatively correlated with the current moving speed; Obtain a power increment based on the cutting power range difference of the hay cutting device and the texture complexity; both the cutting power range difference and the texture complexity are positively correlated with the power increment; use the sum of the minimum cutting power and the power increment as the current cutting power.

7. A method for cutting hay for gardens according to claim 1, characterized in that The preset neighborhood is an eight-neighborhood.

8. A method for cutting hay for gardens according to claim 1, characterized in that, Obtain the edge pixel points through the Canny operator.

9. A hay cutting device for gardens, the device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, it implements the steps of a hay cutting method for gardening according to any one of claims 1 to 8.

Images