Parameter adaptive density filtering method and device for silo grain bulk volume point cloud data

Abstract

This invention relates to a parametric adaptive density filtering method and apparatus for point cloud data of grain pile volume in silos. The parametric adaptive density filtering method for point cloud data of grain pile volume in silos includes the following steps: S1, acquiring point cloud data of the grain pile volume measured inside the silo; S2, generating random samples from the point cloud data using the Monte Carlo method, and calculating the optimal total number of clusters K based on the random samples. best Then according to K best The adaptive mean Euclidean distance S3 is calculated first based on the calculated global density mean of the point cloud data and the local densities ρ1 to ρ of the N original points. N ; and then based on ρ1~ρ N The adaptive density threshold ρ is calculated. y This invention employs clustering to adaptively calculate the neighborhood radius parameter. Then, by calculating the neighborhood density of each original point and using statistical characteristics to dynamically adjust the adaptive density threshold, low-density noise points are eliminated, thereby ensuring the accuracy of point cloud data filtering and improving the accuracy of subsequent grain storage calculations.

Description

Technical Field

[0001] This invention relates to the field of point cloud data processing technology, and in particular to a parameter adaptive density filtering method and apparatus for point cloud data of grain pile volume in silos. Background Technology

[0002] With the continuous development of automated grain storage management, how to efficiently and accurately extract the characteristic parameters of grain pile volume within the grain storage has become an important issue in intelligent grain storage management systems. Traditional grain pile structure detection and modeling mainly rely on traditional manual measurement and pressure sensor methods. However, manual measurement methods are not only time-consuming and labor-intensive, but their accuracy cannot be guaranteed when faced with the complex measurement environment inside the grain storage. Pressure sensor measurement methods, due to the high cost and high failure rate resulting from the large number of sensors required, become increasingly unreliable, especially as the size and complexity of grain storage increases.

[0003] In recent years, with the increasing maturity of 3D laser technology, 3D modeling methods based on point cloud data have gradually emerged. These technologies can acquire 3D point cloud data of the upper surface of grain piles inside grain silos with high precision. Then, based on this 3D point cloud data, the grain storage capacity can be calculated, providing a new approach for structural analysis and feature extraction of grain storage within grain silos, thereby improving the accuracy of subsequent grain storage capacity calculations. However, point cloud data often suffers from problems such as noise, redundancy, and missing data, making feature extraction difficult, especially in large and complex environments like grain silos.

[0004] Currently, commonly used point cloud filtering methods mainly include statistical filtering, radius filtering, and voxel filtering. Statistical filtering filters based on the statistical features of a point's neighborhood, but its parameter settings lack adaptability, easily leading to the loss of effective data. Radius filtering filters by setting a threshold for the number of points within a fixed radius, but it struggles to adapt to uneven point cloud density. Voxel filtering divides the point cloud data into a regular grid for downsampling, but may lose important details. These methods improve the quality of point cloud data to some extent, but still have the following shortcomings: First, when the point cloud density is too high, traditional filtering methods may lose some important structural details. Second, existing point cloud filtering methods have poor adaptability to complex structures in different environments, especially in scenarios like grain silos containing multiple complex factors, often failing to effectively distinguish noise from true structural features. Finally, existing methods often fail to consider the noise source distribution in specific scenarios when processing point clouds, leading to significant deviations in processing results and difficulty in achieving accurate feature extraction, thus affecting the accuracy of subsequent grain reserve calculations. Furthermore, existing filtering algorithms generally suffer from a lack of adaptability in parameter settings, requiring repeated manual adjustments. At the same time, these algorithms cannot guarantee a high filtering rate for the special noise in the volume point cloud data of grain piles inside the grain warehouse, and do not adequately consider the special characteristics of the volume point cloud data of grain piles inside the grain warehouse. Summary of the Invention

[0005] Therefore, it is necessary to address the problem of low filtering accuracy in existing grain pile volume feature point cloud filtering methods by providing a parameter adaptive density filtering method and device for silo grain pile volume point cloud data.

[0006] In a first aspect, this invention proposes a parameter-adaptive density filtering method for point cloud data of grain pile volume in silos, which includes the following steps:

[0007] S1. Obtain point cloud data of the grain pile volume inside the silo; the point cloud data includes: N original points.

[0008] S2. Generate random samples from the point cloud data using the Monte Carlo method, and calculate the optimal total number of clusters K based on the random samples. best Then according to K best The adaptive mean Euclidean distance is calculated.

[0009] S3, First according to The global density mean of the point cloud data was calculated. and the local density ρ1~ρ of N original points N ;n∈[1,N].

[0010] Then according to and ρ1~ρ N The adaptive density threshold ρ is calculated.y .

[0011] S4. Determine ρ1~ρ N Is it greater than ρ? y .

[0012] If yes, the original points corresponding to the point cloud data are retained as information points; otherwise, the points corresponding to the point cloud data are filtered out as noise points.

[0013] S2 includes the following steps:

[0014] S21. Perform M clustering operations on the random samples.

[0015] The m-th clustering process generates K+m clusters; m∈[1,M].

[0016] S22. Calculate the compactly supported measure D of the (k+m)th cluster in the m-th clustering process. m,k+m k∈[1,K+m];

[0017] Traversing the K+m clusters in the m-th clustering process, we obtain K+m compactly supported measures D. m,2 ~D m,K+m ;

[0018] According to D m,2 ~D m,K+m Calculate the sum of squared errors W in the m-th clustering process. m,K+m .

[0019] S23. Perform M clustering operations to obtain M sums of squared errors W. 1,K+1 ~W M,K+M And calculate the average of their logarithms to obtain E(logW). m,K+m ).

[0020] S24. Construct the gap function Gap(K+m), and take the K+m corresponding to the maximum value of the function as the optimal total number of clusters K. best .

[0021]

[0022] S25. Divide N by K best The number of neighboring points t is calculated.

[0023] Calculate the average Euclidean distance from the nth original point to its t neighboring original points.

[0024] Traverse N original points to obtain N average Euclidean distances.

[0025] right Calculate the average to obtain the mean.

[0026] Secondly, the present invention also proposes a parameter-adaptive density filtering device for silo grain pile volume point cloud data, which uses the parameter-adaptive density filtering method for silo grain pile volume point cloud data from the first aspect. This parameter-adaptive density filtering device for silo grain pile volume point cloud data includes an acquisition module, an adaptive module, a threshold module, and a judgment module.

[0027] The acquisition module is used to acquire point cloud data of the volume of the grain pile inside the silo.

[0028] The adaptive module is used to generate random samples from point cloud data using the Monte Carlo method, and calculates the optimal total number of clusters K based on the random samples. best And according to K best The adaptive mean Euclidean distance is calculated.

[0029] The threshold module is used to determine the threshold based on the threshold. The global density mean of the point cloud data was calculated. and the local density ρ1~ρ of N original points N and according to and ρ1~ρ N The adaptive density threshold ρ is calculated. y .

[0030] The judgment module is used to judge ρ1~ρ N Is it greater than ρ? y If yes, then the original points corresponding to the point cloud data are retained as information points; otherwise, the points corresponding to the point cloud data are filtered out as noise points.

[0031] Thirdly, the present invention also proposes a software program product. This software program product includes program instructions that, when run on an electronic device, cause the electronic device to execute the steps of the parameter adaptive density filtering method for the silo grain pile volume point cloud data in the first aspect.

[0032] The beneficial effects of this invention are as follows:

[0033] 1. This invention uses clustering to achieve adaptive calculation of neighborhood radius parameters. Then, by calculating the neighborhood density of each original point and using statistical characteristics to dynamically adjust the adaptive density threshold, low-density noise points are eliminated, thereby ensuring the accuracy of point cloud data filtering and improving the accuracy of subsequent grain storage calculation.

[0034] 2. This invention derives an adaptive density threshold based on the statistical characteristics of point cloud data itself. This dynamic adjustment of the adaptive density threshold takes into account the overall density distribution pattern of point cloud data. Compared with the traditional fixed threshold method, this adaptive threshold determination method is more flexible and can better adapt to the characteristics of point cloud data in different scenarios.

[0035] 3. This invention can filter LiDAR point cloud data for complex structures and large-scale environments. The adaptive parameters in the filtering method make it applicable to most actual cylindrical grain silos, eliminating the need for manual adjustment of filtering parameters, reducing labor costs, improving the algorithm's adaptability to different scenarios, and enhancing the efficiency and intelligence level of grain silo point cloud data processing. Attached Figure Description

[0036] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0037] Figure 1 This is a flowchart of the parameter adaptive density filtering method for point cloud data of silo grain pile volume in the embodiment;

[0038] Figure 2 This is a rendering of the original grain pile volume point cloud data from the experiment.

[0039] Figure 3 The image shows the point cloud data obtained after denoising using the traditional statistical filtering algorithm in the experiment.

[0040] Figure 4 The image shows the point cloud data obtained after denoising using point cloud denoising algorithms based on statistical filtering and bilateral filtering in the experiment.

[0041] Figure 5 The image shows the point cloud data obtained after denoising using a parameter-adaptive statistical filtering denoising algorithm in the experiment.

[0042] Figure 6 The image shows the point cloud data obtained after denoising using the present invention in the experiment. Detailed Implementation

[0043] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the specification of this invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "or / and" as used herein includes any and all combinations of one or more of the associated listed items.

[0045] Please see Figure 1 This embodiment provides a parameter adaptive density filtering method for point cloud data of grain pile volume, which includes the following steps:

[0046] S1. Obtain point cloud data of the grain pile volume inside the silo; the point cloud data includes: N original points.

[0047] In this step, a laser sensor can scan the environment of the grain silo to automatically generate a point cloud dataset of the grain pile surface. This dataset contains information about the number of original points.

[0048] S2. Generate random samples from the point cloud data using the Monte Carlo method, and calculate the optimal total number of clusters K based on the random samples. best And according to K best The adaptive mean E(d) E(d) is calculated.

[0049] In this step, this embodiment employs clustering to divide the point cloud data into different clusters. Points representing the main information and noise points with less information generally belong to different clusters. This embodiment uses interval statistics, a method with low computational cost and insensitivity to dataset density and sparsity. Specifically, S2 includes the following steps:

[0050] S21. The random samples generated by the Monte Carlo method are uniformly distributed within their respective regions, randomly generating the same number of random points as the original points in the point cloud data. These random points are then subjected to M clustering operations. The m-th clustering operation will generate K+m clusters, where m∈[1,M].

[0051] S22. Calculate the compactly supported measure D of the (k+m)th cluster in the m-th clustering process. m,k+m Where k∈[1,K+m].

[0052] Dm,k+m The Euclidean distance is used to represent the clustering; the smaller the distance, the better the compact support of the cluster.

[0053] Specifically, D m,k+m The calculation formula is:

[0054]

[0055] In the formula, Q k+m P represents all the original points in the (k+m)th cluster; q For Q k+m The q-th original point within the cluster; μ is the center point of the (k+m)-th cluster.

[0056] Then, the K+m clusters in the m-th clustering process are traversed to obtain the compactly supported measure D. m,2 ~D m,K+m Then according to D m,2 ~D m,K+m Calculate the sum of squared errors W in the m-th clustering process. m,K+m The calculation formula is as follows:

[0057]

[0058] S23, take W m,K+m log(W) m,K+m The M clustering processes yielded M log(W) values. 1,K+1 )~log(W M,K+M ). For log(W) 1,K+1 )~log(W M,K+M Calculate the average value to obtain E(logW). m,K+m ).

[0059] S24. Construct the gap function Gap(K+m), and take the K+m corresponding to the maximum value of the function as the optimal total number of clusters K. best :

[0060]

[0061] S25, according to K best Cluster the point cloud data into K groups best There are clusters, at which point N is divided by K. best The number of neighboring points, t, is calculated. This t represents the average number of original points within each cluster. Using t as the number of neighboring points required by the KNN algorithm, the average Euclidean distance d from the nth original point to its t neighboring original points can be calculated. n Then, by traversing the N original points, N average Euclidean distances can be obtained. Will The desired mean is obtained by summing and averaging.

[0062] S3, according to The global density mean of the point cloud data was calculated. and the local density ρ1~ρ of N original points N n∈[1,N]. According to and ρ1~ρ N The adaptive density threshold ρ is calculated. y .

[0063] In this step, density-adaptive filtering parameters are established, primarily considering the local density features of each original point in the point cloud data. First, the density-adaptive filtering parameters are... The search radius is used as the radius of the nth original point in the point cloud data. This search radius forms a spherical space, and the number M of all original points located within this spherical space is calculated. The calculation method for M includes the following steps:

[0064] S61. Calculate the relationship between the nth original point and any other original point P. j The Euclidean distance d nj .

[0065] S62, Compare d nj and Size.

[0066] like Then the count value is incremented by 1. If Then the count value is incremented by 0.

[0067] S63. After traversing N-1 original points, the resulting count value is M.

[0068] In another embodiment, the calculation method for M is further optimized, differing from the above embodiment in that: the N-1 original points are traversed in order of distance from the nth original point from nearest to farthest, until... Then stop. Compared with the above embodiments, this embodiment only needs to calculate the original points near the nth original point, without needing to search all the original points, thus greatly reducing the amount of data calculation.

[0069] After obtaining M, according to Calculate the local density ρ of the nth original point with M. n The calculation formula is as follows:

[0070]

[0071] Then, traversing the N original points, we obtain the local densities ρ1 to ρ2 corresponding to the N original points. N This local density calculation fully considers the spatial distribution characteristics around each original point, accurately reflecting the density of the point cloud in different regions. Then, it determines ρ1 to ρ...N Does it satisfy the Gaussian distribution condition? If so, the cutoff threshold λ is obtained according to the properties of the Gaussian distribution (e.g., by consulting a normal distribution table). And to obtain a suitable ρ... y It is also necessary to convert ρ1 to ρ N After summing and averaging, we get get ρ can be calculated after adding λ. y The calculation formula is as follows:

[0072]

[0073] In the formula, σ is the standard deviation of the point cloud data.

[0074] The Jarque-Bera statistical method can be used to evaluate ρ1 to ρ N Does it follow a Gaussian distribution? Specifically, first calculate the density variance S of the point cloud data, which is calculated using the following formula:

[0075]

[0076] Then σ above is The statistic JB is then calculated using the following formula:

[0077]

[0078] In the formula, T represents skewness, L represents kurtosis, and E[·] represents the expected value of [·].

[0079] If N > 30, and JB ~ χ 2 (2), then ρ1~ρ N It satisfies the Gaussian distribution condition.

[0080] S4. For original points with local density below the adaptive density threshold, due to the sparse distribution of surrounding original points and a significant difference in distribution characteristics from normal original points, they are identified as noise points or outliers and need to be removed from the data. Original points with high local density, however, often reflect the true features of the object's surface and will be retained as valid point cloud data. Therefore, it is necessary to determine ρ. n Is it greater than ρ? y If so, the original points corresponding to the point cloud data are retained as information points; otherwise, the points corresponding to the point cloud data are filtered out as noise points. The final point cloud data composed of the retained information points is the filtered point cloud data.

[0081] In summary, the main principle of this invention is to determine the number of neighborhoods based on interval statistics, and to derive the neighborhood search radius and truncation threshold multiple from the statistics of the point cloud data itself. Then, based on the neighborhood density of each original point in the point cloud data, the local density of each original point is evaluated by calculating the average neighborhood density. An adaptive density threshold is then calculated based on statistical characteristics. All points below the adaptive density threshold are considered noise points and are removed, while the remaining points are retained as valid information points. This adaptive filtering method based on clustering and density features ensures the denoising effect of the algorithm, prevents the uncertainty and inefficiency caused by manual input parameters, and avoids the loss of valid information due to over-filtering. Furthermore, by fully utilizing the statistical characteristics of point cloud data, this method has strong robustness and adaptability, effectively handling point cloud data with different density distribution characteristics, providing a more reliable foundation for subsequent data analysis and applications. This invention is not only applicable to silos but also to flat warehouses. It can be widely used in fields such as 3D modeling and intelligent management of grain warehouses, especially in automated storage management and intelligent monitoring systems, providing important support for grain warehouse storage calculation and intelligent management. With the continuous development of intelligent technology, based on the technology of this invention, the grain warehouse management system will become more intelligent and efficient, promoting the grain warehouse industry towards greater automation and informatization.

[0082] To verify the effectiveness of the denoising effect of this invention, the denoising effect of this invention was compared with that of existing algorithms for simulated cylindrical grain surface models with added composite noise. Denoising experiments were then conducted on point cloud models under different noise intensities, and the results were analyzed. The specific analysis process is as follows:

[0083] This invention was conducted on an Ubuntu 24.04 LTS operating system environment, and the algorithm was implemented using the point cloud library (PCL 1.14.0). The visualization part used the PCL Visualizer built-in visualization module for point cloud rendering and display. For example... Figure 2 As shown, the original data used in this paper is 3D point cloud data derived from mathematical modeling based on relevant mathematical parameters of the actual silo and the actual conditions inside the grain silo using the PCL point cloud library. It consists of point cloud data of a conical surface with a base radius of 8m and a height of 4m, and several simulated temperature measuring cable noises. The total number of points in the initial simulated noise-added point cloud is 105,000. Figure 2 As shown in the figure. Four experimental schemes were set up to denoise the original point cloud data. To facilitate observation of the denoising effect, the point cloud data obtained by traditional statistical filtering denoising is shown in the figure. Figure 3 As shown. The point cloud data obtained by using a point cloud denoising algorithm based on statistical filtering and bilateral filtering (Algorithm 1) is as follows. Figure 4As shown. The point cloud data obtained by denoising using the parameter-adaptive statistical filtering denoising algorithm (Algorithm 2) is as follows. Figure 5 As shown. The point cloud data obtained by denoising using the algorithm of this invention is as follows. Figure 6 As shown in Table 1, the denoising effect achieved by the method used in this invention is superior to other traditional algorithms and those mentioned in the literature.

[0084] Table 1: Comparison of Denoising Results

[0085]

[0086] In other embodiments, a parameter adaptive density filtering device for silo grain pile volume point cloud data is also proposed, which uses the parameter adaptive density filtering method for silo grain pile volume point cloud data in the above embodiments.

[0087] The parameter adaptive density filtering device includes an acquisition module, an adaptive module, a threshold module, and a judgment module. Among them:

[0088] The acquisition module is configured to acquire point cloud data obtained from the volume measurement of the grain pile inside the silo. The adaptive module is configured to generate random samples from the point cloud data using the Monte Carlo method, and calculate the optimal total number of clusters K based on the random samples. best And according to K best The adaptive mean Euclidean distance is calculated. The threshold module is configured as follows: based on The global density mean of the point cloud data was calculated. and the local density ρ1~ρ of N original points N and according to and ρ1~ρ N The adaptive density threshold ρ is calculated. y The judgment module is configured to judge ρ1 to ρ N Is it greater than ρ? y If yes, then the original points corresponding to the point cloud data are retained as information points; otherwise, the points corresponding to the point cloud data are filtered out as noise points.

[0089] In other embodiments, an electronic device is also proposed. This electronic device includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, it implements the steps of the parameter adaptive density filtering method for the point cloud data of the silo grain pile volume described in the above embodiments.

[0090] In other embodiments, a computer-readable storage medium is also proposed. This computer-readable storage medium stores a computer program. When the computer program is executed by a processor, it implements the steps of the parameter adaptive density filtering method for the silo grain pile volume point cloud data described in the above embodiments.

[0091] In other embodiments, a software program product is also proposed. This software program product includes program instructions that, when run on an electronic device, cause the electronic device to perform the steps of the parameter adaptive density filtering method for the silo grain pile volume point cloud data described in the above embodiments.

[0092] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0093] The embodiments described above are merely examples of several implementations of the present invention, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these all fall within the protection scope of the present invention. Therefore, the protection scope of this invention patent should be determined by the appended claims.

Claims

1. A method for parameter adaptive density filtering of silo grain bulk volume point cloud data, characterized in that, It comprises the following steps: S1, acquiring point cloud data measured by a volume of the grain pile in the silo; the point cloud data includes: N one original point; S2, generating random samples from the point cloud data by a Monte Carlo method, and calculating an optimal total cluster number according to the random samples K best , and calculating an adaptive average Euclidean distance mean value K best , and calculating an adaptive average Euclidean distance mean value ; S3、first according to The global density mean value of the point cloud data is calculated And N The local density of the original point ; n ∈[1, N ]; Further according to and The adaptive density threshold is calculated ; S4, determining whether greater than ; If yes, the original point corresponding to the point cloud data is reserved as an information point; if not, the point corresponding to the point cloud data is filtered out as a noise point; S2 comprises the following steps: S21, performing on the random sample M sub-clustering process; First m sub-cluster processing generates K + m clusters; m ∈[1, M ]; S22、calculating the first m tight measure of the first k + m cluster in the second clustering process D m,k+m ; k ∈[1, K + m ]; traversing the first m second clustering process K m K m D m,2 ～D m,K+m ;​​​​ According to D m,2 ～D m,K+m Computing the sum of squared errors in the second clustering process m W m,K+m ;​ S23, proceed M Sub-cluster processing obtains M The sum of squares of errors W 1,K+1 W M,K+M , and calculates the average value of the logarithm, obtains E (log W m,K+m );​ S24, construct interval value function Gap K+m , take the maximum value of K+m as the optimal total cluster number K best ;​ ； S25、will N Divide by K best The number of neighborhood points is calculated t ; calculating the average Euclidean distance of the first original point to the second original point n t ;​​ traverse N original points, obtaining N average euclidean distances ; To average, the mean .

2. The method for parameter adaptive density filtering of silo grain bulk volume point cloud data according to claim 1, characterized in that, In S3, , The calculation method of the present application comprises the following steps: Firstly The search radius of the first original point in the point cloud data is calculated, and the number of all original points located within the search radius is calculated n M ;​ Further according to and M the local density of the n first original point is calculated ; traversing N original points, obtaining N local density ; To average, we get .

3. The method for parameter adaptive density filtering of silo grain bulk volume point cloud data according to claim 2, characterized in that, M The calculation method includes the following steps: S61, Calculate the first n Each origin point and any origin point P j European distance d nj ; S62, compare d nj with the size of If d nj ≤ then the count value is incremented by 1; S63, iterate N - 1 original point, resulting count value is M .

4. The method for parameter adaptive density filtering of silo grain bulk volume point cloud data according to claim 3, characterized in that, S63 further comprises: traverse the n N -1 original points in order from near to far from the original point until d nj stop.​​​​ 5. The method for parameter adaptive density filtering of silo grain bulk volume point cloud data according to claim 2, wherein, According to the first n The search radius of the original point constitutes a spherical space; The calculation formula is: 。 6. The method of claim 1, wherein, In S3, The calculation method comprises the following steps: S71, judging whether the Gaussian distribution condition is met; If satisfied, then according to the properties of the Gaussian distribution, the threshold multiple of the cutoff is obtained ; S72、According to The adaptive density threshold is calculated The formula is: ； In the formula, is the standard deviation of the point cloud data.

7. The method of claim 6, wherein, In S71, it is determined The method for determining whether the Gaussian distribution condition is satisfied includes the following steps: Computing the density variance S ; The Jarque-Bera statistic is calculated JB The formula is: ； In the formula, T is skewness, and L is kurtosis; If N > 30, and then the Gaussian distribution condition is satisfied.

8. The method of claim 1, wherein, In S22, D m,k+m The calculation formula is: ； In the formula, Q k+m For the first k+m All original points in a cluster; P q for Q k+m The first q One original point; For the first k+m The center point of each cluster.

9. A device for parameter adaptive density filtering of silo grain bulk volume point cloud data, characterized by The silo grain bulk volume point cloud data parameter adaptive density filtering method uses the silo grain bulk volume point cloud data parameter adaptive density filtering device; The silo grain bulk volume point cloud data parameter adaptive density filtering device comprises: An acquisition module is configured to acquire point cloud data measured by a silo grain bulk volume measurement device; An adaptive module is configured to generate random samples from the point cloud data by a Monte Carlo method, and to calculate an optimal total cluster number according to the random samples K best , and to calculate an adaptive average Euclidean distance mean K best ;​ a threshold module, configured to calculate an adaptive density threshold value according to a global density mean value of the point cloud data and N a local density of the original point , and according to and ;​ a judging module for judging whether the distance is greater than ; if yes, the original point corresponding to the point cloud data is reserved as an information point; if not, the point corresponding to the point cloud data is filtered out as a noise point.

10. A software program product characterized in that The software program product comprises program instructions which, when executed on an electronic device, cause the electronic device to perform the steps of the silo grain bulk volume point cloud data parameter adaptive density filtering method according to any one of claims 1 to 8.

Images