Point cloud data optimization method and system based on self-supervised training

Abstract

The invention discloses a point cloud data optimization method and system based on self-supervised training. The optimization method is used for optimizing to-be-processed actual production scene target point cloud data to obtain optimized actual production scene target point cloud data, and the to-be-processed actual production scene target point cloud data is characterized by being incomplete, sparse or edge fuzzy. The optimized actual production scene target point cloud data has the characteristics of completeness, denseness or smoothness; structural integrity, data density and data smoothness are optimized, and the method comprises the following steps: training a point cloud data cascade optimization network TP-PCONet through a data pair construction strategy and a self-supervision method; through snowflake diffusion-based point cloud completion, gradually refined point cloud up-sampling and edge thermodynamic value-based point cloud refinement, to-be-processed actual production scene target point cloud data is subjected to trunk, surface and contour constraint in sequence, and optimized actual production scene target point cloud data is obtained. The invention further discloses corresponding electronic equipment and a computer readable storage medium.

Description

Technical Field

[0001] The present invention relates to the technical field of deep learning and point cloud data optimization processing, and in particular to a point cloud data optimization method and system based on self-supervised training. Background Art

[0002] In the segmentation of different parts of the target in actual production scenarios, for example, the segmentation of the stems and leaves of plants, the density and completeness of the application data are important influencing factors. Since plants are affected by multiple biological, physical and environmental factors during their natural growth, they show a high degree of morphological complexity and irregularity, which makes it difficult for the acquisition equipment to obtain uniform and dense point cloud data during the acquisition process. There are studies on the optimization of point cloud data, among which the point cloud upsampling method based on deep learning is the current mainstream solution. Point cloud upsampling refers to the task of generating a dense point cloud given a set of sparse point cloud inputs, which is a necessary task to complete downstream visual tasks. These include: using the point cloud upsampling network PU-Net, which learns the multi-level features of each point and implicitly expands the point set in the feature space through multi-branch convolutional units, thereby achieving point cloud upsampling; or using PU-Transformer, which improves the feature map connection between points and channels through the encoding and decoding structure and multi-head self-attention structure, thereby generating a more realistic point cloud; or using the point cloud upsampling network Grad-PU, which first interpolates the point cloud, and then refines the position of the interpolation point through iterative optimization, achieving more accurate point cloud upsampling. It can be seen that point cloud upsampling technology can be used to optimize plant point cloud data collected in actual production scenarios, and provide data support for subsequent plant stem and leaf segmentation.

[0003] However, the existing point cloud upsampling network adds points based on the existing points. When the input point cloud is incomplete, resulting in an incomplete morphological structure, such a network is limited by the existing data points and it is difficult to accurately restore the real structure of the point cloud, thereby limiting its application accuracy on complex structure point clouds; and when the input point cloud is noisy, the redundant noise features of the existing data points cause the edge of the point cloud to be blurred, which will interfere with the feature extraction of the network, resulting in inaccurate upsampling results, thereby limiting its application effect. Due to the complex structure of the leaves, the tortuous direction of the stems, and the randomness of the overall growth of the plant point cloud to be optimized, it is inevitable that the mutual occlusion between leaves and leaves, stems and leaves will be generated during the collection; and because the plant itself is a flexible object, the plant itself will shake due to external factors such as wind and movement of the collection equipment during the collection, resulting in serious point cloud noise and blurred edges. Therefore, the existing point cloud upsampling algorithm has difficulties in optimizing the plant point cloud, and the degree of data optimization directly affects the accuracy of stem and leaf segmentation, which is not conducive to the accurate segmentation of plant stems and leaves in actual production scenarios. How to construct a point cloud data optimization model that can achieve highly optimized point clouds in actual production scenarios and provide analyzable data for the segmentation of different parts of subsequent targets (such as stems and leaves of plants) and calculation of phenotypic parameters is one of the research focuses of the present invention.

[0004] In summary, the following key challenges still exist in point cloud data work at this stage:

[0005] Low quality of point cloud data: Due to the morphological complexity and irregularity of targets in actual production scenarios (for example, plant growth), the collected point cloud data inevitably contains a lot of noise, resulting in blurred edges and large incompleteness. The point cloud data quality is low, which affects the segmentation accuracy of different parts of the target (for example, the stems and leaves of the plant). Summary of the invention

[0006] In order to solve the problems existing in the prior art, a point cloud data optimization method and system based on self-supervised training are provided to solve the technical problem of optimizing point cloud data collected in actual production scenarios. In order to reasonably guide the process of optimizing point cloud data collected in actual production scenarios to high-quality point clouds, and at the same time ensure that the optimized data can fully and accurately reflect the structural information of the target (e.g., plants), the present invention is based on the morphology and data characteristics analysis of the target (e.g., plants) in the actual production scene, and proposes a point cloud data cascade optimization network TP-PCONet based on self-supervised training. The point cloud data of the actual production scene is optimized in terms of structural integrity, data density, and data smoothness, so that the segmentation model can effectively process the optimized high-quality point cloud data.

[0007] The first aspect of the present invention is to provide a point cloud data optimization method based on self-supervised training, which is used to optimize the actual production scene target point cloud data to be processed to obtain the optimized actual production scene target point cloud data, wherein the data characteristics of the actual production scene target point cloud data to be processed are one or more of incompleteness, sparseness and edge fuzziness; the data characteristics of the actual production scene target point cloud data after optimization are one or more of completeness, density and smoothness; the optimization is performed from the perspective of structural integrity, data density and data smoothness, and the method comprises:

[0008] S1, training the point cloud data cascade optimization network TP-PCONet through data pair construction strategy and self-supervision method;

[0009] S2, through the three steps of point cloud completion based on snowflake diffusion, point cloud upsampling based on gradual refinement, and point cloud refinement based on edge thermal value, the actual production scene target point cloud data to be processed are constrained in three aspects: backbone, surface, and contour, to obtain the optimized actual production scene target point cloud data.

[0010] Preferably, the S1 comprises:

[0011] S11, taking indoor high-quality point cloud data as input, obtaining simulated production scene point cloud data through a data pair construction strategy based on the point cloud characteristics of the actual production scene, and constructing a training data pair based on the simulated production scene point cloud data; the data pair construction strategy of the point cloud characteristics of the actual production scene includes: taking indoor high-quality point cloud data as input, and presetting the incompleteness coefficient n, sparseness coefficient m and noise coefficient k during processing to respectively implement surface incompleteness simulation, overall sparseness and noise addition, so as to simulate the occlusion and shaking of the target and the point cloud data changes caused during the data acquisition process, and processing the simulated production scene point cloud obtained Data is taken as output; the method of obtaining the simulated production scene point cloud data by the data pair construction strategy based on the point cloud characteristics of the actual production scene includes: first, normalizing the indoor high-quality point cloud data; then randomly selecting n points in the indoor high-quality point cloud data as the center of the cube generated in the point cloud space, and setting the edge length; then removing the points belonging to the cube from the input indoor high-quality point cloud data to obtain the processed residual point cloud data; finally, randomly downsampling the processed residual point cloud data by m times, and adding Gaussian noise with k times the standard deviation, and finally obtaining the simulated production scene point cloud data;

[0012] S12, training the point cloud data cascade optimization network TP-PCONet based on the training data pair and the self-supervised learning method, including: setting the upsampling network PUCRN as the baseline model; introducing a point cloud completion module based on snowflake diffusion and a point cloud refinement module based on edge thermal value on the basis of the point cloud upsampling module based on stepwise refinement in the baseline model to constitute the point cloud data cascade optimization network TP-PCONet, the input and output of the point cloud completion module based on snowflake diffusion, the point cloud upsampling module based on stepwise refinement and the point cloud refinement module based on edge thermal value correspond to the low-quality point cloud data before optimization and the high-quality point cloud data after optimization for the same target in the training data pair; training the point cloud data cascade optimization network TP-PCONet based on the training data pair and the self-supervised learning method, the training data pair are respectively the network input value and the training true value of the point cloud data cascade optimization network TP-PCONet before training, and the training is used to mine the internal inherent structure and mapping relationship of the training data pair to obtain the point cloud data cascade optimization network TP-PCONet.

[0013] Preferably, the incomplete coefficient n is 5, the sparse coefficient m is 4, the noise coefficient k is 0.02; the edge length is 0.05; the point cloud data cascade optimization network TP-PCONet uses the upsampling network PUCRN as the baseline model, and the upsampling network PUCRN avoids distortion and discontinuity problems caused by excessive one-time filling of the point cloud by gradually increasing the point cloud density.

[0014] Preferably, S2 includes:

[0015] S21, processing of the point cloud completion module based on snowflake diffusion, is used to predict the complete form of the actual production scene point cloud to be optimized to obtain a structurally complete actual production scene target point cloud, and provide complete data for the point cloud optimization, including: taking the point cloud data to be optimized as input to obtain the completed point cloud data, as shown in formula (2) and formula (3):

[0016]

[0017] in Represents the input point cloud data to be optimized, Represents the output completed point cloud data, N and K represent the number of point cloud points of the input point cloud data to be optimized and the number of point cloud points of the output completed point cloud data, respectively, K>N, It consists of three steps, among which MFE(x i ,y i , z i ) represents the point cloud backbone feature extraction, SG(x i ,y i , zi ) indicates the missing part of the seed point generation, SPD(x i ,y i , z i ) represents snowflake point diffusion;

[0018] S22, based on the processing of the gradually refined point cloud upsampling module, is used to receive the completed complete target point cloud, and perform an upsampling operation on it to obtain the final dense point cloud data, and provide the dense point cloud data for subsequent point cloud optimization. The gradually refined point cloud upsampling module takes the completed point cloud as input and outputs the dense point cloud, as shown in formula (7) and formula (8):

[0019]

[0020] in, Represents the complete point cloud data of the input, Represents the output dense point cloud data, r represents the upsampling multiple; It consists of two steps, FE(x i ,y i , z i ) represents the feature expansion of point cloud data, CR(x i ,y i , z i ) represents the coordinate reconstruction of point cloud data;

[0021] S23, processing of the point cloud refinement module based on edge thermal value, is used to receive the dense target point cloud data, refine it to generate a smoother point cloud, and provide data for the subsequent segmentation of each part of the target. The point cloud refinement module based on edge thermal value takes the dense point cloud data as input and outputs the smoothed point cloud data, as shown in formula (11) and formula (12):

[0022]

[0023] in, represents the input dense point cloud data, Represents the output refined point cloud data, It consists of three steps: GC(x i ,y i , z i ) indicates the gradient value calculation, HMG(x i ,y i , z i ) represents the generation of heat map, EPC(x i ,y i , z i ) indicates that the edge points shrink.

[0024] Preferably, the S21 includes:

[0025] Step 1: Extract the backbone features of the point cloud. This is used to extract the backbone features of the point cloud P0 to be optimized. The encoder in the Transformer architecture is used to extract and encode the input point cloud to obtain the backbone feature vector of the point cloud, as shown in formula (4):

[0026] V m =T e (P0) (4)

[0027] in, represents the backbone feature vector, N represents the total number of points in the input point cloud, T e Represents the encoder of Transformer;

[0028] Step 2: Generate missing seed points, including: based on the extracted backbone feature vector V m , the Transformer decoder predicts the seed point set of the missing part, as shown in formula (5). The seed points in the seed point set of the missing part represent the approximate position and shape of the missing area;

[0029] P s =T d (V m ) (5)

[0030] in, n represents the number of points in the generated point cloud subset, T d Represents the decoder of Transformer;

[0031] Step 3: Snowflake point diffusion, including: using multiple SPD modules in the SnowFlakeNet network, taking the seed point in the seed point set of the missing part as the center, gradually diffusing outward in a snowflake shape to generate new points until the entire point cloud data becomes complete, and obtaining complete point cloud data. The diffusion process of the snowflake-shaped gradual outward diffusion is shown in formula (6):

[0032] P s(i+1) =SPD(P si ) (6)

[0033] Among them, P si represents the input point cloud of the i+1th SPD module, P s(i+1) Represents the output point cloud of the i+1th SPD module. When i=predetermined value, the point diffusion is completed to obtain the actual production scene target point cloud with complete structure;

[0034] Among them, the point cloud completion module based on snowflake diffusion obtains the backbone points through the farthest point sampling of the true point cloud, and calculates the loss with the predicted point cloud through CD loss, and updates the point cloud completion module based on snowflake diffusion.

[0035] Preferably, the S22 includes:

[0036] Step 1: feature expansion of point cloud data, including: after the completed point cloud data is input into a point cloud upsampling module based on gradual refinement, feature expansion is performed through a multi-layer perceptron MLP to obtain an expanded point cloud feature vector, as shown in formula (9):

[0037] F up =MLP(P1) (9)

[0038] in, represents the extended feature vector;

[0039] Step 2: coordinate reconstruction of point cloud data, including: fitting the distribution and regularity of the existing point cloud data, and gradually generating new point cloud subsets from the point cloud feature vector after the feature expansion through the multi-layer perceptron MLP again to fill the sparse area in the original point cloud, as shown in formula (10):

[0040] S=MLP(F up ) (10)

[0041] in, Represents a new point cloud subset generated based on the use of a step-by-step upsampling method, where α=r / 2; repeating the coordinate reconstruction operation of the point cloud data to obtain the final dense point cloud data.

[0042] Based on the gradually refined point cloud upsampling module, the CD loss of the true point cloud and the predicted point cloud is directly calculated to constrain the generated dense point cloud. Through the above operations, a complete and dense target point cloud of the actual production scene can be obtained.

[0043] Preferably, the S23 includes:

[0044] Step 1: Gradient value calculation, including: using the KD tree to perform nearest neighbor query on the input dense point cloud data to obtain the neighbor point set of each point in the point cloud, as shown in formula (13):

[0045] PN = KDTree(P i ) (13)

[0046] Where P i represents the input point cloud, represents the set of neighboring points of each point, and β represents the number of neighboring points of each point;

[0047] Step 2: Heat map generation, including: calculating the geometric center PC of the neighboring point set of each point as the center coordinate of the center point of the current point, as shown in formula (14):

[0048]

[0049] Calculate the distance between each point and its corresponding center point to obtain the position offset of each point in the point cloud, and calculate the difference between the offset value of the current point and the offset value of its neighboring points. The largest difference is the gradient value d of each point. i , as shown in formula (15):

[0050] d i =max||P i -PC i |-|PN i -PC i || (15)

[0051] In the point cloud data, the larger the gradient value of the point, the closer the point is to the edge. Conversely, the closer the point is to the inside of the point cloud, the calculated point cloud gradient value is normalized to [0,1] to obtain the edge thermal value of the point cloud. The closer the thermal value is to 1, the closer the point is to the edge, thus forming an edge thermal map;

[0052] Step 3: shrinking edge points, including: shrinking edge points based on the generated edge heat map, and obtaining a final optimized point cloud;

[0053] Among them, the point cloud refinement module based on edge thermal value calculates the edge thermal values ​​of the predicted point cloud and the true point cloud, and constrains the point cloud through the L1 loss of the two, so as to obtain a complete, dense and smooth target point cloud of the actual production scene.

[0054] The second aspect of the present invention is to provide a point cloud data optimization system based on self-supervised training, which is used to implement the point cloud data optimization method based on self-supervised training described in the first aspect, characterized in that the actual production scene target point cloud data to be processed is optimized to obtain the optimized actual production scene target point cloud data, the data characteristics of the actual production scene target point cloud data to be processed are one or more of incompleteness, sparseness and edge fuzzy; the data characteristics of the actual production scene target point cloud data after optimization are one or more of completeness, density and smoothness; the optimization is performed from the perspective of structural integrity, data density and data smoothness, and the system includes:

[0055] Self-supervised training module, used to train the point cloud data cascade optimization network TP-PCONet through data pair construction strategy and self-supervised method;

[0056] The point cloud data optimization module is used to sequentially constrain the backbone, surface and contour of the actual production scene target point cloud data to be processed through three steps: point cloud completion based on snowflake diffusion, point cloud upsampling based on gradual refinement, and point cloud refinement based on edge thermal value, to obtain the optimized actual production scene target point cloud data.

[0057] A third aspect of the present invention provides an electronic device, comprising a processor and a memory, wherein the memory stores a plurality of instructions, and the processor is configured to read the instructions and execute the method described in the first aspect.

[0058] A fourth aspect of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a plurality of instructions, and the plurality of instructions can be read by a processor to execute the method described in the first aspect.

[0059] The method, system, electronic device, and computer-readable storage medium provided by the present invention have the following beneficial technical effects:

[0060] A point cloud data optimization method is proposed, which is segmented by using a target segmentation framework based on actual production scene acquisition with zero annotation, providing the possibility for phenotypic parameter calculation. At the same time, the point cloud data collected in the actual production scene is optimized, and a point cloud cascade optimization network TP-PCONet based on self-supervised training is designed. The experimental results show that the proposed TP-PCONet achieves a CD distance of 8.45 and an F1-score of 73.5%, respectively. Therefore, the proposed method has the potential to significantly improve the segmentation accuracy and efficiency of different parts of the target (for example, the stems and leaves of plants), and has similar excellent performance and application scenarios in other related fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Figure 1 This is a flow chart of the point cloud data optimization method based on self-supervised training according to the present invention;

[0062] Figure 2 It is a principle diagram of the zero-annotation target segmentation method based on point cloud data optimization described in the present invention;

[0063] Figure 3 Schematic diagram of the point cloud optimization method based on self-supervised training according to the present invention;

[0064] Figure 4 This is an example diagram of the point cloud data characteristics of an actual production scenario described in the present invention;

[0065] Figure 5 This is a schematic diagram of the structure of the point cloud data cascade optimization network TP-PCONet of the present invention;

[0066] Figure 6This is a schematic diagram of the point cloud edge heat map described in the present invention;

[0067] Figure 7 This is a diagram of the architecture of the point cloud data optimization system based on self-supervised training according to the present invention;

[0068] Figure 8 It is a schematic diagram of the structure of the electronic device described in the present invention. DETAILED DESCRIPTION

[0069] In order to better understand the above technical solution, the above technical solution will be described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0070] The method provided by the present invention can be implemented in the following terminal environment, and the terminal may include one or more of the following components: a processor, a memory, and a display screen. The memory stores at least one instruction, and the instruction is loaded and executed by the processor to implement the method described in the following embodiment.

[0071] The processor may include one or more processing cores. The processor uses various interfaces and lines to connect various parts of the entire terminal, and executes various functions of the terminal and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory, and calling data stored in the memory.

[0072] The memory may include a random access memory (RAM) or a read-only memory (ROM). The memory may be used to store instructions, programs, codes, code sets or instructions.

[0073] The display screen is used to display the user interface of each application.

[0074] In addition, those skilled in the art can understand that the structure of the above terminal does not constitute a limitation on the terminal, and the terminal may include more or fewer components, or combine certain components, or arrange the components differently. For example, the terminal also includes components such as a radio frequency circuit, an input unit, a sensor, an audio circuit, and a power supply, which will not be described in detail here.

[0075] Embodiment 1

[0076] like Figure 1 As shown, this embodiment provides a point cloud data optimization method based on self-supervised training, which is used to optimize the actual production scene target point cloud data to be processed to obtain optimized actual production scene target point cloud data, including:

[0077] S1, training the point cloud data cascade optimization network TP-PCONet through data pair construction strategy and self-supervision method;

[0078] In this embodiment, Figure 2 As shown in ①, the point cloud data optimization method based on self-supervised training is used to solve the problem that the algorithm cannot effectively extract structural features due to quality factors such as incompleteness, sparseness and blurred edges of point clouds collected in actual production scenarios.

[0079] In this embodiment, the optimization task for point cloud data of actual production scenarios requires that the model be able to learn the morphological characteristics of the plants and have the ability to extract and characterize complete and detailed three-dimensional structural information of the plants. How to set plant morphological supervision information for the model is the primary consideration. Among various supervision methods, the self-supervised learning method uses the data itself as a supervision signal, deeply mines the information contained in the data through the network, and updates the knowledge representation in real time. It can accurately capture the subtle spatial structural characteristics of tomato plants and is suitable for the optimization task of point cloud data of actual production scenarios. Therefore, the present invention uses the self-supervised training method to achieve a high degree of optimization of point clouds in actual production scenarios, such as Figure 3 The method includes a data pair construction strategy based on the point cloud characteristics of actual production scenarios and a point cloud data cascade optimization network TP-PCONet, which are introduced below.

[0080] In the model training stage: first, indoor high-quality point clouds are used as input, and through ① a data pair construction strategy based on the characteristics of actual production scene point clouds, simulated production scene point clouds are obtained to construct training data pairs; secondly, the produced training data pairs are used to self-supervise the training of ② point cloud data cascade optimization network TP-PCONet. In the model application stage: the actual production scene point cloud to be processed is input into the trained point cloud data optimization model to obtain the optimized actual production scene point cloud.

[0081] The data characteristics of the actual production scene target point cloud data to be processed are one or more of incompleteness, sparseness and edge fuzziness; the data characteristics of the actual production scene target point cloud data after optimization are one or more of completeness, density and smoothness; the optimization is performed from three aspects: structural integrity, data density and data smoothness;

[0082] As a preferred embodiment, the S1 includes:

[0083] S11, taking indoor high-quality point cloud data as input, obtaining simulated production scene point cloud data through a data pair construction strategy based on the point cloud characteristics of the actual production scene, and constructing a training data pair based on the simulated production scene point cloud data; the data pair construction strategy of the point cloud characteristics of the actual production scene includes: taking indoor high-quality point cloud data as input, and presetting the incompleteness coefficient n, sparseness coefficient m and noise coefficient k during processing to respectively implement surface incompleteness simulation, overall sparseness and noise addition, so as to simulate the occlusion and shaking of the target and the point cloud data changes caused during the data acquisition process, and processing the simulated production scene point cloud obtained Data is taken as output; the method of obtaining the simulated production scene point cloud data through the data pair construction strategy based on the point cloud characteristics of the actual production scene includes: first, normalizing the indoor high-quality point cloud data; then randomly selecting n points in the indoor high-quality point cloud data as the center of the cube generated in the point cloud space, and setting the edge length; then removing the points belonging to the cube from the input indoor high-quality point cloud data to obtain the processed residual point cloud data; finally, randomly downsampling the processed residual point cloud data by m times, and adding Gaussian noise with k times the standard deviation to obtain the simulated production scene point cloud data.

[0084] In this embodiment, the point cloud data optimization network based on self-supervised training needs to set sample training pairs of low-quality point clouds before optimization and high-quality point clouds after optimization for the same target, respectively as network input values ​​and training truth values, to mine the inherent structure and mapping relationship within the data. However, in actual production scenarios, the collected point cloud data cannot obtain high-quality point clouds as supervision truth values ​​due to the complex environment. Therefore, it is necessary to use high-quality point cloud data sets that can be obtained under indoor conditions as supervision truth values ​​to provide sufficient supervision information to guide the model to perform correct self-learning. Furthermore, in order to enable the network to have the ability to optimize the tomato plant point cloud in actual production scenarios, the present invention sets a simulated production scenario data set based on the characteristics of the point cloud data in actual production scenarios.

[0085] Examples of point cloud data features in actual production scenarios include: Figure 4 As shown in the figure, there are mainly the following aspects: First, due to the mutual occlusion between leaves and between leaves and stems, the collected point cloud is incomplete [(A): incomplete characteristic]; second, due to the complex environment in which the plant is located and the irregular growth form, it is difficult to collect a uniform and dense point cloud [(B): sparse characteristic]; in addition, in the actual production scene collection process, external factors such as wind disturbance and shaking of the collection equipment will cause the plant to be unable to fix a posture, resulting in blurred edges of the collected point cloud [(C): blurred edge characteristic].

[0086] Therefore, based on the above analysis of the characteristics of the actual production scene point cloud data, the specific manifestation of edge blur is that the points at the boundary are not distributed in a certain linear shape, and the local distribution is Gaussian noise. Therefore, this method performs a series of processing on the input indoor high-quality point cloud, including surface defect simulation, overall sparseness, and noise addition, to simulate plant occlusion, shaking, and point cloud data changes that may be caused during data collection. The simulated production scene point cloud data with the characteristics of the actual production scene point cloud is obtained. The construction strategy of the simulated production scene point cloud data set is shown in Algorithm 1.

[0087]

[0088] As can be seen from Algor ithm 1, in the process of constructing the data pair, the present invention uses indoor high-quality point clouds as input, and presets the incomplete coefficient n (set to 5 in the embodiment of the present invention), the sparse coefficient m (set to 4 in the embodiment of the present invention) and the noise coefficient k (set to 0.02 in the embodiment of the present invention) during processing, and uses the simulated production scene point cloud obtained after processing as output. In this process, first, the input point cloud data is normalized for subsequent processing; then n points are randomly selected from the input point cloud as the center of the cube generated in the point cloud space, and the edge length is set to 0.05 (the empirical value of the embodiment of the present invention); then the points belonging to the cube are removed from the input point cloud to obtain the processed residual point cloud; finally, it is randomly downsampled m times, and Gaussian noise with k times the standard deviation is added to finally obtain the simulated production scene point cloud. Through the above operations, the simulated production scene plant point cloud samples can be obtained, thereby constructing the low-quality point cloud data pairs before optimization and the high-quality point cloud data pairs after optimization that can be used for point cloud optimization network training.

[0089] S12, training the point cloud data cascade optimization network TP-PCONet based on the training data pair and the self-supervised learning method, including: setting the upsampling network PUCRN as the baseline model; introducing a point cloud completion module based on snowflake diffusion and a point cloud refinement module based on edge thermal value on the basis of the point cloud upsampling module based on stepwise refinement in the baseline model to constitute the point cloud data cascade optimization network TP-PCONet, the input and output of the point cloud completion module based on snowflake diffusion, the point cloud upsampling module based on stepwise refinement and the point cloud refinement module based on edge thermal value correspond to the low-quality point cloud data before optimization and the high-quality point cloud data after optimization for the same target in the training data pair; training the point cloud data cascade optimization network TP-PCONet based on the training data pair and the self-supervised learning method, the training data pair are respectively the network input value and the training true value of the point cloud data cascade optimization network TP-PCONet before training, and the training is used to mine the internal inherent structure and mapping relationship of the training data pair to obtain the point cloud data cascade optimization network TP-PCONet.

[0090] In this embodiment, the self-supervised learning method enables the network to actively learn the mapping relationship between data pairs, so that it has the ability to optimize point clouds in actual production scenarios. In order to further guide the network to optimize from low-quality point clouds to high-quality point clouds, capture the fine structure of the target (in this embodiment, tomato plants), and avoid optimization distortion problems, this embodiment designs the network structure as follows.

[0091] First, this embodiment uses the upsampling network PUCRN as the baseline model. This model effectively avoids the distortion and discontinuity caused by excessive filling of point clouds at one time by gradually increasing the density of point clouds, which is suitable for the complex structure and rich details of tomato plants. Secondly, in order to enable the optimization network to have the ability to complete the incomplete point cloud, this embodiment introduces the point cloud completion module based on snowflake diffusion in SnowFlakeNet in the baseline model. This module uses the principle of snowflake diffusion to achieve smooth transition and natural connection of point cloud data, avoiding the problem of abrupt distortion in completion, so as to meet the different incomplete characteristics of stems and leaves in the plant point cloud. In view of the problem of blurred edges of stems and leaves caused by the data itself and the upsampling and completion operations, this embodiment constrains the surface of the point cloud and designs a point cloud upsampling module based on gradual refinement. At the same time, by constraining the contour of the point cloud, a point cloud refinement module based on edge thermal value is designed. The above three modules constitute the point cloud data cascade optimization network TP-PCONet designed in this embodiment. Figure 5 The specific structural diagram of the point cloud data cascade optimization network TP-PCONet is shown. It is mainly composed of three sub-modules (A-1), (A-2), and (A-3) in cascade. Among them, (A-1): constrains the point cloud backbone to effectively fill in the missing parts and provide complete data for subsequent processing; (A-2): constrains the point cloud surface to optimize sparse data into dense data and improve data density; (A-3): constrains the contour of the completed and dense point cloud to obtain smooth point cloud data.

[0092] S2, through the three steps of point cloud completion based on snowflake diffusion, point cloud upsampling based on gradual refinement, and point cloud refinement based on edge thermal value, the actual production scene target point cloud data to be processed are constrained in three aspects: backbone, surface, and contour, to obtain the optimized actual production scene target point cloud data.

[0093] In practical applications, the optimized actual production scene point cloud data is input into Figure 2 The standard point cloud segmentation training model in ② is used to further segment the target (the tomato plant in this embodiment) into different parts (the stem and leaf segmentation in this embodiment).

[0094] In the actual production scenario tomato plant point cloud data optimization task, it is assumed that the input point cloud subset is represented as Where P 0i represents the i-th point of the point cloud, N is the total number of points in the input point cloud, and the task aims to obtain an optimized point cloud subset Where M represents the total number of points in the optimized point cloud, and M>N. The optimized point cloud subset P3 is obtained by formula (1), as shown below:

[0095]

[0096] In formula (1), express Figure 5 The completion module processing process based on snowflake diffusion shown in (A-1) is as follows: express Figure 5 (A-2) shows the upsampling module processing process based on step-by-step refinement, express Figure 5 (A-3) shows the refinement module processing process based on edge thermal value.

[0097] In this embodiment, the processing flow after the point cloud data is input into the data optimization network is processed into three modules: a point cloud completion module based on snowflake diffusion, a point cloud upsampling module based on gradual refinement, and a point cloud refinement module based on edge thermal value.

[0098] Therefore, as a preferred embodiment, the S2 includes:

[0099] S21, processing of point cloud completion module based on snowflake diffusion

[0100] Figure 5 (A-1) shows the point cloud completion module based on snowflake diffusion introduced by SnowFlakeNet in this embodiment, which is used to predict the complete form of the actual production scene point cloud to be optimized and provide complete data for subsequent point cloud optimization. This module takes the point cloud to be optimized as input and obtains the completed point cloud, as shown in formula (2) and formula (3).

[0101]

[0102] in Represents the input point cloud data to be optimized, Represents the output completed point cloud data, N and K represent the number of point cloud points of the input point cloud data to be optimized and the number of point cloud points of the output completed point cloud data, respectively, K>N, It consists of three steps, among which MFE(x i ,y i , z i ) represents the point cloud backbone feature extraction, SG(x i ,yi , z i ) indicates the missing part of the seed point generation, SPD(x i ,y i , z i ) represents snowflake point diffusion. The specific steps are as follows.

[0103] Step 1: Extract the backbone features of the point cloud, which is used to extract the backbone features of the point cloud P0 to be optimized. Use the encoder in the Transformer architecture to extract and encode the input point cloud to obtain the backbone feature vector of the point cloud, as shown in formula (4).

[0104] V m =T e (P0) (4)

[0105] in, represents the backbone feature vector, N represents the total number of points in the input point cloud, T e Represents the encoder of Transformer.

[0106] Step 2: Generate missing seed points, including: based on the extracted backbone feature vector V m , the Transformer decoder predicts the seed point set of the missing part, as shown in formula (5). The seed points in the seed point set of the missing part represent the approximate position and shape of the missing area.

[0107] P s =T d (V m ) (5)

[0108] in, n represents the number of points in the generated point cloud subset, T d Represents the decoder of Transformer.

[0109] Step 3: Snowflake point diffusion, including: using multiple SPD modules in the SnowFlakeNet network, with the seed point in the missing part of the seed point set in step 2 as the center, gradually diffuse outward in a snowflake shape to generate new points until the entire point cloud data becomes complete, and the complete point cloud data is obtained. The module diffusion process is shown in formula (6).

[0110] P s(i+1) =SPD(P si ) (6)

[0111] Among them, P si represents the input point cloud of the i+1th SPD module, P s(i+1) represents the output point cloud of the i+1th SPD module. In this embodiment, when i=2, point diffusion is completed.

[0112] The point cloud completion module based on snowflake diffusion obtains the backbone points by sampling the farthest point of the true point cloud, and calculates the loss with the predicted point cloud through CDloss to update the point cloud completion module based on snowflake diffusion. Through the above three steps, a complete point cloud of plants in the actual production scene can be obtained.

[0113] S22, processing of point cloud upsampling module based on gradual refinement

[0114] Figure 5 (A-2) shows the point cloud upsampling module based on gradual refinement in the baseline model PUCRN, which is mainly used to receive the completed plant point cloud and perform upsampling operations on it to provide dense data for subsequent point cloud optimization. This module takes the completed point cloud as input and outputs the dense point cloud, as shown in formulas (7) and (8).

[0115]

[0116] in, Represents the complete point cloud data of the input, represents the output dense point cloud data, r represents the upsampling multiple, which is set to 4 in this embodiment; It consists of two steps, FE(x i ,y i , z i ) represents the feature expansion of point cloud data, CR(x i ,y i , z i ) represents the coordinate reconstruction of point cloud data. The specific steps are as follows.

[0117] Step 1: feature expansion of point cloud data, including: after the completed point cloud data is input into a point cloud upsampling module based on gradual refinement, feature expansion is performed through a multi-layer perceptron MLP to obtain an expanded point cloud feature vector, as shown in formula (9).

[0118] F up =MLP(P1) (9)

[0119] in, Represents the extended feature vector.

[0120] Step 2: coordinate reconstruction of point cloud data, including: fitting the distribution and regularity of the existing point cloud data, and gradually generating new point cloud subsets from the point cloud feature vector after the feature expansion through the multi-layer perceptron MLP again to fill the sparse areas in the original point cloud, as shown in formula (10).

[0121] S=MLP(F up) (10)

[0122] in, Represents the newly generated point cloud subset. Since this embodiment uses a step-by-step upsampling method, α=r / 2 here; repeat the operation of step 2 to obtain the final dense point cloud data.

[0123] Based on the gradually refined point cloud upsampling module, the CD loss of the true point cloud and the predicted point cloud is directly calculated, and the generated dense point cloud is constrained. Through the above operations, a complete and dense point cloud of the actual production scene plant can be obtained.

[0124] S23, processing of point cloud refinement module based on edge thermal value

[0125] Figure 5 (A-3) shows the point cloud refinement module based on edge thermal value designed in this embodiment, which is mainly used to receive the dense plant point cloud data and refine it to generate a smoother point cloud to provide data for subsequent stem and leaf segmentation. The point cloud refinement module based on edge thermal value takes dense point cloud data as input and outputs smoothed point cloud data, as shown in formula (11) and formula (12).

[0126]

[0127] in, represents the input dense point cloud data, Represents the output refined point cloud data, It consists of three steps: GC(x i ,y i , z i ) indicates the gradient value calculation, HMG(x i ,y i , z i ) represents the generation of heat map, EPC(x i ,y i , z i ) indicates the edge point shrinkage. The specific implementation steps are as follows.

[0128] Step 1: Gradient value calculation, including: using the KD tree to perform nearest neighbor query on the input dense point cloud data to obtain the neighbor point set of each point in the point cloud, as shown in formula (13).

[0129] PN = KDTree(P i ) (13)

[0130] Where P i represents the input point cloud, represents the neighbor point set of each point, β represents the number of neighbor points of each point, which is set to 350 in this embodiment.

[0131] Step 2: Heat map generation, including: calculating the geometric center PC of the neighboring point set of each point as the center coordinate of the center point of the current point, as shown in formula (14).

[0132]

[0133] Calculate the distance between each point and its corresponding center point to obtain the position offset of each point in the point cloud, and calculate the difference between the offset value of the current point and the offset value of its neighboring points. The largest difference is the gradient value d of each point. i , as shown in formula (15).

[0134] d i =max||P i -PC i |-|PN i -PC i || (15)

[0135] In the point cloud data, the larger the gradient value of the point, the closer the point is to the edge. Conversely, the closer the point is to the inside of the point cloud, the calculated point cloud gradient value is normalized to [0,1] to obtain the edge thermal value of the point cloud. The closer the thermal value is to 1, the closer the point is to the edge, thus forming a clear edge thermal map. The point cloud edge thermal map is as follows: Figure 6 shown, specifically, Figure 6 Middle (A) represents the original input point cloud; Figure 6 Middle (B) shows the corresponding heat map visualization generated by the encoding.

[0136] Step 3: Edge point shrinkage, including: based on the generated edge heat map, shrinking the edge points to reduce the edge noise of the point cloud, making the point cloud outline clear, and obtaining the final optimized point cloud. This module calculates the edge heat value of the predicted point cloud and the true point cloud, and constrains the point cloud through the L1 loss of the two. Through the above operations, a complete, dense and smooth point cloud of plants in actual production scenes can be obtained.

[0137] After three modules, the actual production scene point cloud data to be optimized is processed by TP-PCONet step by step, and finally a point cloud with high-quality point cloud distribution characteristics is formed. The optimized point cloud structure is complete, dense and smooth, thus providing effective data for the segmentation of tomato plant stems and leaves.

[0138] Embodiment 2

[0139] like Figure 7As shown, this embodiment provides a point cloud data optimization system based on self-supervised training, which is used to implement the method of the aforementioned embodiment 1, optimize the actual production scene target point cloud data to be processed to obtain optimized actual production scene target point cloud data, the data characteristics of the actual production scene target point cloud data to be processed are one or more of incomplete, sparse and edge fuzzy; the data characteristics of the actual production scene target point cloud data after optimization are one or more of complete, dense and smooth; the optimization is performed from the perspective of structural integrity, data density and data smoothness, and the system includes:

[0140] A self-supervised training module 101, used for training the point cloud data cascade optimization network TP-PCONet through a data pair construction strategy and a self-supervised method;

[0141] The point cloud data optimization module 102 is used to sequentially constrain the actual production scene target point cloud data to be processed in terms of backbone, surface and contour through three steps: point cloud completion based on snowflake diffusion, point cloud upsampling based on gradual refinement, and point cloud refinement based on edge thermal value, so as to obtain the optimized actual production scene target point cloud data.

[0142] The present invention also provides a memory storing a plurality of instructions, wherein the instructions are used to implement the method described in the first embodiment.

[0143] like Figure 8 As shown, the present invention also provides an electronic device, including a processor 301 and a memory 302 connected to the processor 301, wherein the memory 302 stores a plurality of instructions, which can be loaded and executed by the processor so that the processor can execute the method described in Example 1.

[0144] Although preferred embodiments of the present invention have been described, additional changes and modifications may be made to these embodiments by those skilled in the art once the basic inventive concepts are known. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention. Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A point cloud data optimization method based on self-supervised training, characterized in that: Used to optimize the actual production scene target point cloud data to be processed to obtain optimized actual production scene target point cloud data, wherein the data characteristics of the actual production scene target point cloud data to be processed are one or more of incomplete, sparse and edge fuzzy; the data characteristics of the actual production scene target point cloud data after optimization are one or more of complete, dense and smooth; The optimization is performed from the perspectives of structural integrity, data density and data smoothness, and the method comprises: S1, training the point cloud data cascade optimization network TP-PCONet through data pair construction strategy and self-supervision method; S2, through the three steps of point cloud completion based on snowflake diffusion, point cloud upsampling based on gradual refinement, and point cloud refinement based on edge thermal value, the actual production scene target point cloud data to be processed are constrained in three aspects: backbone, surface, and contour, to obtain the optimized actual production scene target point cloud data.

2. The point cloud data optimization method based on self-supervised training according to claim 1, characterized in that: The S1 includes: S11, taking indoor high-quality point cloud data as input, obtaining simulated production scene point cloud data through a data pair construction strategy based on the point cloud characteristics of the actual production scene, and constructing a training data pair based on the simulated production scene point cloud data; the data pair construction strategy of the point cloud characteristics of the actual production scene includes: taking indoor high-quality point cloud data as input, and presetting the incompleteness coefficient n, sparseness coefficient m and noise coefficient k during processing to respectively implement surface incompleteness simulation, overall sparseness and noise addition, so as to simulate the occlusion and shaking of the target and the point cloud data changes caused during the data acquisition process, and processing the simulated production scene point cloud obtained Data is taken as output; the method of obtaining the simulated production scene point cloud data by the data pair construction strategy based on the point cloud characteristics of the actual production scene includes: first, normalizing the indoor high-quality point cloud data; then randomly selecting n points in the indoor high-quality point cloud data as the center of the cube generated in the point cloud space, and setting the edge length; then removing the points belonging to the cube from the input indoor high-quality point cloud data to obtain the processed residual point cloud data; finally, randomly downsampling the processed residual point cloud data by m times, and adding Gaussian noise with k times the standard deviation, and finally obtaining the simulated production scene point cloud data; S12, training the point cloud data cascade optimization network TP-PCONet based on the training data pair and the self-supervised learning method, including: setting the upsampling network PUCRN as the baseline model; introducing a point cloud completion module based on snowflake diffusion and a point cloud refinement module based on edge thermal value on the basis of the point cloud upsampling module based on stepwise refinement in the baseline model to constitute the point cloud data cascade optimization network TP-PCONet, the input and output of the point cloud completion module based on snowflake diffusion, the point cloud upsampling module based on stepwise refinement and the point cloud refinement module based on edge thermal value correspond to the low-quality point cloud data before optimization and the high-quality point cloud data after optimization for the same target in the training data pair; training the point cloud data cascade optimization network TP-PCONet based on the training data pair and the self-supervised learning method, the training data pair are respectively the network input value and the training true value of the point cloud data cascade optimization network TP-PCONet before training, and the training is used to mine the internal inherent structure and mapping relationship of the training data pair to obtain the point cloud data cascade optimization network TP-PCONet.

3. The point cloud data optimization method based on self-supervised training according to claim 2, characterized in that: The incomplete coefficient n is 5, the sparse coefficient m is 4, the noise coefficient k is 0.02; the edge length is 0.05; the point cloud data cascade optimization network TP-PCONet uses the upsampling network PUCRN as the baseline model, and the upsampling network PUCRN avoids the distortion and discontinuity caused by excessive filling of the point cloud at one time by gradually increasing the point cloud density.

4. The point cloud data optimization method based on self-supervised training according to claim 3, characterized in that: The S2 includes: S21, processing of the point cloud completion module based on snowflake diffusion, is used to predict the complete form of the actual production scene point cloud to be optimized to obtain a structurally complete actual production scene target point cloud, and provide complete data for the point cloud optimization, including: taking the point cloud data to be optimized as input to obtain the completed point cloud data, as shown in formula (2) and formula (3): in Represents the input point cloud data to be optimized, Represents the output completed point cloud data, N and K represent the number of point cloud points of the input point cloud data to be optimized and the number of point cloud points of the output completed point cloud data, respectively, K>N, It consists of three steps, among which MFE(x i ,y i , z i ) represents the point cloud backbone feature extraction, SG(x i ,y i , z i ) indicates the missing part of the seed point generation, SPD(x i ,y i , z i ) represents snowflake point diffusion; S22, based on the processing of the gradually refined point cloud upsampling module, is used to receive the completed complete target point cloud, and perform an upsampling operation on it to obtain the final dense point cloud data, and provide the dense point cloud data for subsequent point cloud optimization. The gradually refined point cloud upsampling module takes the completed point cloud as input and outputs the dense point cloud, as shown in formula (7) and formula (8): P2=F2(P1) (7) in, Represents the complete point cloud data of the input, Represents the output dense point cloud data, r represents the upsampling multiple; It consists of two steps, FE(x i ,y i , z i ) represents the feature expansion of point cloud data, CR(x i ,y i , z i ) represents the coordinate reconstruction of point cloud data; S23, processing of the point cloud refinement module based on edge thermal value, is used to receive the dense target point cloud data, refine it to generate a smoother point cloud, and provide data for the subsequent segmentation of each part of the target. The point cloud refinement module based on edge thermal value takes the dense point cloud data as input and outputs the smoothed point cloud data, as shown in formula (11) and formula (12): in, represents the input dense point cloud data, Represents the output refined point cloud data, It consists of three steps: GC(x i ,y i , z i ) indicates the gradient value calculation, HMG(x i ,y i , z i ) represents the generation of heat map, EPC(x i ,y i , z i ) indicates that the edge points shrink.

5. The point cloud data optimization method based on self-supervised training according to claim 4, characterized in that: The S21 includes: Step 1: Extract the backbone features of the point cloud. This is used to extract the backbone features of the point cloud P0 to be optimized. The encoder in the Transformer architecture is used to extract and encode the input point cloud to obtain the backbone feature vector of the point cloud, as shown in formula (4): V m =T e (P0) (4) in, represents the backbone feature vector, N represents the total number of points in the input point cloud, T e Represents the encoder of Transformer; Step 2: Generate missing seed points, including: based on the extracted backbone feature vector V m , the Transformer decoder predicts the seed point set of the missing part, as shown in formula (5). The seed points in the seed point set of the missing part represent the approximate position and shape of the missing area; P s =T d (V m ) (5) in, n represents the number of points in the generated point cloud subset, T d Represents the decoder of Transformer; Step 3: Snowflake point diffusion, including: using multiple SPD modules in the SnowFlakeNet network, taking the seed point in the seed point set of the missing part as the center, gradually diffusing outward in a snowflake shape to generate new points until the entire point cloud data becomes complete, and obtaining complete point cloud data. The diffusion process of the snowflake-shaped gradual outward diffusion is shown in formula (6): P s(i+1) =SPD(P si ) (6) Among them, P si represents the input point cloud of the i+1th SPD module, P s(i+1) Represents the output point cloud of the i+1th SPD module. When i=predetermined value, the point diffusion is completed to obtain the actual production scene target point cloud with complete structure; Among them, the point cloud completion module based on snowflake diffusion obtains the backbone points through the farthest point sampling of the true point cloud, and calculates the loss with the predicted point cloud through CD loss, and updates the point cloud completion module based on snowflake diffusion.

6. The point cloud data optimization method based on self-supervised training according to claim 5, characterized in that: The S22 includes: Step 1: feature expansion of point cloud data, including: after the completed point cloud data is input into a point cloud upsampling module based on gradual refinement, feature expansion is performed through a multi-layer perceptron MLP to obtain an expanded point cloud feature vector, as shown in formula (9): F up =MLP(P1) (9) in, represents the extended feature vector; Step 2: coordinate reconstruction of point cloud data, including: fitting the distribution and regularity of the existing point cloud data, and gradually generating new point cloud subsets from the point cloud feature vector after the feature expansion through the multi-layer perceptron MLP again to fill the sparse area in the original point cloud, as shown in formula (10): S=MLP(F up ) (10) in, Represents a new point cloud subset generated based on the use of a step-by-step upsampling method, where α=r / 2; repeating the coordinate reconstruction operation of the point cloud data to obtain the final dense point cloud data; Based on the gradually refined point cloud upsampling module, the CD loss of the true point cloud and the predicted point cloud is directly calculated, and the generated dense point cloud is constrained. Through the above operations, a complete and dense target point cloud of the actual production scene can be obtained.

7. The point cloud data optimization method based on self-supervised training according to claim 6, characterized in that: The S23 includes: Step 1: Gradient value calculation, including: using the KD tree to perform nearest neighbor query on the input dense point cloud data to obtain the neighbor point set of each point in the point cloud, as shown in formula (13): PN=KDTree(P i ) (13) Where P i represents the input point cloud, represents the set of neighboring points of each point, and β represents the number of neighboring points of each point; Step 2: Heat map generation, including: calculating the geometric center PC of the neighboring point set of each point as the center coordinate of the center point of the current point, as shown in formula (14): Calculate the distance between each point and its corresponding center point to obtain the position offset of each point in the point cloud, and calculate the difference between the offset value of the current point and the offset value of its neighboring points. The largest difference is the gradient value d of each point. i , as shown in formula (15): d i =max||P i -PC i |-|PN i -PC i || (15) In the point cloud data, the larger the gradient value of the point, the closer the point is to the edge. Conversely, the closer the point is to the inside of the point cloud, the calculated point cloud gradient value is normalized to [0, 1] to obtain the edge thermal value of the point cloud. The closer the thermal value is to 1, the closer the point is to the edge, thus forming an edge thermal map; Step 3: shrinking edge points, including: shrinking edge points based on the generated edge heat map, and obtaining a final optimized point cloud; Among them, the point cloud refinement module based on edge thermal value calculates the edge thermal values ​​of the predicted point cloud and the true point cloud, and constrains the point cloud through the L1 loss of the two, so as to obtain a complete, dense and smooth target point cloud of the actual production scene.

8. A point cloud data optimization system based on self-supervised training, used to implement the point cloud data optimization method based on self-supervised training according to any one of claims 1 to 7, characterized in that: The actual production scene target point cloud data to be processed is optimized to obtain optimized actual production scene target point cloud data, wherein the data characteristics of the actual production scene target point cloud data to be processed are one or more of incompleteness, sparseness and edge fuzziness; the data characteristics of the actual production scene target point cloud data after optimization are one or more of completeness, density and smoothness; The optimization is performed from the perspectives of structural integrity, data density and data smoothness, and the system comprises: A self-supervised training module (101), used for training the point cloud data cascade optimization network TP-PCONet through a data pair construction strategy and a self-supervised method; The point cloud data optimization module (102) is used to sequentially constrain the actual production scene target point cloud data to be processed in terms of backbone, surface and contour through three steps: point cloud completion based on snowflake diffusion, point cloud upsampling based on gradual refinement, and point cloud refinement based on edge thermal value, so as to obtain the optimized actual production scene target point cloud data.

9. An electronic device, characterized in that: It includes a processor and a memory, wherein the memory stores a plurality of instructions, and the processor is used to read the instructions and execute the point cloud data optimization method based on self-supervised training as described in claims 1-7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the plurality of instructions can be read by a processor and executed by the point cloud data optimization method based on self-supervised training as described in claims 1-7.

Images