Intelligent monitoring system for sandblasting equipment operation trajectory data based on artificial intelligence

Abstract

The present invention relates to the field of data processing technology, and in particular to an intelligent monitoring system for sandblasting equipment running trajectory data based on artificial intelligence, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the following steps when executing the computer program: obtaining a raster image of a large workpiece to be sandblasted; dividing the raster image into a plurality of block areas, and using an ant colony algorithm to obtain a first optimal path between all the block areas; obtaining at least one growth area in the block area, and using an ant colony algorithm to obtain a second optimal path between all the growth areas and a third optimal path for each growth area; performing sandblasting on the large workpiece to be sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each block area, and obtaining a sandblasting path with better sandblasting quality, so that the local sandblasting operation is more delicate and the efficiency of the sandblasting operation is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an intelligent monitoring system for sandblasting equipment operation trajectory data based on artificial intelligence. Background Art

[0002] At present, the sandblasting operation of large steel structures is mainly based on manual operation. The efficiency of manual sandblasting is low and the quality of sandblasting is poor, especially it will have a serious impact on the health of operators. Therefore, the use of automatic and intelligent sandblasting equipment (such as sandblasting robots) to replace manual sandblasting operations will become the main development trend of the sandblasting industry in the future.

[0003] Intelligent sandblasting equipment needs to set the running track in advance to complete the automated sandblasting operation. Among them, reasonable path planning can help the intelligent sandblasting equipment choose the shortest and most effective path for operation, which can reduce a lot of redundant movements, not only shorten the sandblasting time, but also effectively reduce energy consumption and equipment wear, thereby improving the overall sandblasting operation efficiency.

[0004] In the prior art, an ant colony algorithm is usually used to obtain the nearest sandblasting path. However, the surface environment of the sandblasted workpiece is complex. When traversing a large workpiece, the traditional ant colony algorithm consumes a lot of calculation and time, resulting in poor calculation efficiency of the sandblasting trajectory. In addition, the algorithm only considers the nearest sandblasting path, and rarely combines the actual surface characteristics of the workpiece, resulting in low quality of the obtained sandblasting path.

[0005] Therefore, how to improve the quality of the sandblasting path obtained by using the ant colony algorithm to improve the operating efficiency of the intelligent sandblasting equipment has become an urgent problem to be solved. Summary of the invention

[0006] In view of this, an embodiment of the present invention provides an intelligent monitoring system for sandblasting equipment operation trajectory data based on artificial intelligence to solve the problem of how to improve the quality of the sandblasting path obtained using the ant colony algorithm to improve the operating efficiency of the intelligent sandblasting equipment.

[0007] An embodiment of the present invention provides an intelligent monitoring system for sandblasting equipment operation trajectory data based on artificial intelligence, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the following steps when executing the computer program:

[0008] Acquire a surface structure image of a large workpiece to be sandblasted, mark and distinguish the sprayable area and the obstacle area in the surface structure image to obtain a target image, and acquire a grid image corresponding to the target image according to a grid method and a preset sandblasting diameter;

[0009] According to the obstacle area in the grid image, an adaptive block size when dividing the grid image is obtained, the grid image is divided into a plurality of block areas according to the adaptive block size, and an ant colony algorithm is used to obtain a first optimal path between all the block areas;

[0010] For any block area, respectively obtain the regional growth standard value between every two adjacent grids in the block area, perform regional growth on the block area according to the regional growth standard value to obtain at least one growth area in the block area, and use the ant colony algorithm to respectively obtain the second optimal path between all the growth areas and the third optimal path of each of the growth areas;

[0011] The large workpiece to be sandblasted is sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each of the block areas.

[0012] Preferably, the step of obtaining an adaptive block size when dividing the raster image into blocks according to the obstacle area in the raster image comprises:

[0013] For any obstacle region in the raster image, count the number of side lengths of the obstacle region, calculate the area ratio between the minimum circumscribed rectangle of the obstacle region and the obstacle region, use the product of the reciprocal of the number of side lengths and the area ratio as a weight, and based on the weight, perform weighted summation on the square of each side length in the obstacle region to obtain a side length characteristic value of the obstacle region;

[0014] Obtain a side length characteristic value of each obstacle area in the raster image, accumulate all side length characteristic values ​​to obtain an obstacle distribution characteristic value in the raster image, and obtain an adaptive block size when dividing the raster image into blocks according to the obstacle distribution characteristic value.

[0015] Preferably, the obtaining of the adaptive block size when dividing the raster image into blocks according to the obstacle distribution characteristic value comprises:

[0016] Among the side lengths of each obstacle area in the raster image, the maximum side length and the minimum side length are obtained, the side length difference between the maximum side length and the minimum side length is calculated, the product value of the obstacle distribution characteristic value and the side length difference of a preset multiple is obtained, the sum of the product value and the minimum side length is rounded, and the adaptive block size is obtained accordingly.

[0017] Preferably, the method of using an ant colony algorithm to obtain the first optimal path between all the block areas includes:

[0018] The centroid of each of the sub-block areas is obtained respectively, and the shortest path between all the centroids is obtained by using an ant colony algorithm as the first optimal path between all the sub-block areas.

[0019] Preferably, the step of respectively obtaining the regional growth standard value between every two adjacent grids in the block region comprises:

[0020] For any grid in the block area, the material feature value corresponding to the grid is obtained according to the grayscale values ​​of all pixels corresponding to the grid in the surface structure image, and the regional roughness feature value corresponding to the grid is obtained according to the grayscale values ​​of the neighboring pixels of each pixel corresponding to the grid in the surface structure image;

[0021] For any two adjacent grids in the block area, the first absolute difference value of the material characteristic values ​​and the second absolute difference value of the regional roughness characteristic values ​​between the two adjacent grids are calculated respectively, the first absolute difference value and the second absolute difference value are normalized respectively to obtain a first normalized value and a second normalized value respectively, and the regional growth standard value between the two adjacent grids is obtained according to the mean of the first normalized value and the second normalized value.

[0022] Preferably, obtaining the material feature value corresponding to the grid according to the grayscale values ​​of all pixels corresponding to the grid in the surface structure image includes:

[0023] The grayscale value average of all pixels corresponding to the grid in the surface structure image is calculated as the material feature value corresponding to the grid.

[0024] Preferably, obtaining the regional roughness feature value corresponding to the grid according to the grayscale value of the neighboring pixel points of each pixel point corresponding to the grid in the surface structure image includes:

[0025] For any pixel point corresponding to the grid in the surface structure image, among the pixels within a preset neighborhood range of the pixel point, a pixel point existing in the grid is obtained as a target pixel point, and the absolute value of the grayscale difference between the pixel point and each of the target pixel points is obtained to obtain the sum of the absolute values ​​of the differences;

[0026] The sum of the absolute values ​​of the differences of each pixel point corresponding to the grid in the surface structure image is accumulated to obtain the regional roughness characteristic value corresponding to the grid.

[0027] Preferably, performing region growing on the block regions according to the region growing standard value to obtain at least one growing region in the block regions comprises:

[0028] Taking the first grid in the upper left corner of the block area as the initial seed point, and taking the regional growth standard value between two adjacent grids being less than or equal to a preset standard value threshold as the growth criterion, regional growth is performed on the block area to obtain at least one growth area in the block area.

[0029] Preferably, the using the ant colony algorithm to respectively obtain the second optimal path between all growth regions and the third optimal path of each growth region includes:

[0030] Obtaining the centroid of each of the growth regions, and using the ant colony algorithm to obtain the shortest path between all centroids as the second optimal path between all the growth regions;

[0031] For any growth region, the ant colony algorithm is used to obtain the shortest path between all grids in the growth region as the third optimal path of the growth region.

[0032] Preferably, the sandblasting operation is performed on the large workpiece to be sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each of the block areas, including:

[0033] When performing sandblasting operation on the large workpiece to be sandblasted, based on the first optimal path, a first traversed block area is obtained, based on the second optimal path of the first traversed block area, a first traversed growth area in the first traversed block area is obtained, and based on the third optimal path of the first traversed growth area, sandblasting operation is performed on the first traversed growth area;

[0034] After the sandblasting operation of the first traversed growth area is completed, based on the second optimal path of the first traversed block area, a second traversed growth area in the first traversed block area is obtained, and based on the third optimal path of the second traversed growth area, the second traversed growth area is sandblasted, and so on, until the sandblasting operation of the first traversed block area is completed;

[0035] Based on the first optimal path, a second traversed block area is obtained, and sandblasting is performed on the second traversed block area according to the sandblasting method of the first traversed block area until the sandblasting of all the block areas is completed.

[0036] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0037] The present invention obtains a surface structure image of a large workpiece to be sandblasted, marks and distinguishes a sprayable area and an obstacle area in the surface structure image to obtain a target image, and obtains a grid image corresponding to the target image according to a grid method and a preset sandblasting diameter; obtains an adaptive block size when dividing the grid image according to the obstacle area in the grid image, divides the grid image into multiple block areas according to the adaptive block size, and uses an ant colony algorithm to obtain a first optimal path between all block areas; for any block area, obtains a regional growth standard value between every two adjacent grids in the block area, performs regional growth on the block area according to the regional growth standard value to obtain at least one growth area in the block area, and uses the ant colony algorithm to obtain a second optimal path between all growth areas and a third optimal path for each growth area; and performs a sandblasting operation on the large workpiece to be sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each of the block areas. Among them, by combining the distribution characteristics of obstacle areas in the raster image, the sandblasting equipment is divided into three layers of sandblasting path trajectories, namely the first optimal path between all block areas in the raster image, the second optimal path between all growth areas in each block area, and the third optimal path of each growth area. Then, based on these three layers of sandblasting path trajectories, sandblasting operations of large workpieces to be sandblasted are carried out, avoiding the problem of high computational complexity of the ant colony algorithm when facing a large number of grids, so as to obtain sandblasting paths with better sandblasting quality, making local sandblasting operations more delicate and improving the efficiency of sandblasting operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0039] Figure 1 It is a flow chart of a method for intelligently monitoring the operation trajectory data of sandblasting equipment based on artificial intelligence provided in the first embodiment of the present invention;

[0040] Figure 2 A schematic diagram of a raster image provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0041] Embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, but should not be understood as limiting the present disclosure.

[0042] It should be noted that the terms "first", "second", etc. in the specification of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure.

[0043] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for illustration.

[0044] The specific scenario targeted by the present invention is: when a sandblasting robot is used to perform sandblasting operations on a large workpiece to be sandblasted, a surface image of the large workpiece to be sandblasted is obtained, and the traditional ant colony algorithm is used to obtain the nearest path of the surface image to serve as the sandblasting trajectory of the sandblasting robot. However, due to the large area and complex surface conditions of large workpieces, the direct use of the ant colony algorithm has a large amount of calculation, so that the sandblasting trajectory of the sandblasting robot only considers the nearest path, and the actual surface conditions of the workpiece are less combined, resulting in poor quality of the obtained sandblasting trajectory. Therefore, the present invention obtains the optimal sandblasting trajectory of the sandblasting robot by combining the characteristics of the workpiece surface, thereby improving the sandblasting efficiency.

[0045] An embodiment of the present invention provides an intelligent monitoring system for sandblasting equipment operation trajectory data based on artificial intelligence, comprising a processor and a memory, wherein the processor executes a computer program stored in the memory to implement an intelligent monitoring method for sandblasting equipment operation trajectory data based on artificial intelligence, such as Figure 1 As shown, the intelligent monitoring method of sandblasting equipment operation trajectory data based on artificial intelligence includes the following steps:

[0046] Step S101, obtaining a surface structure image of a large workpiece to be sandblasted, marking and distinguishing the sprayable area and the obstacle area in the surface structure image to obtain a target image, and obtaining a grid image corresponding to the target image according to a grid method and a preset sandblasting diameter.

[0047] Before the sandblasting robot performs sandblasting on a large workpiece to be sandblasted, it first obtains the surface structure image of the large workpiece to be sandblasted through a camera, and then uses the trained convolutional neural network (CNN) model to simply identify and mark the sprayable area and obstacle area in the surface structure image. The corresponding surface structure image after distinction is the target image.

[0048] The target image area is discretized using the grid method to simplify it into a two-dimensional array or matrix that can be recognized by the computer. Specifically, the sandblasting diameter D of the sandblasting robot is obtained as the unit length of the grid. It should be noted that the sandblasting diameter belongs to the pre-set operating parameters of the sandblasting robot. It is not limited here and can be set according to the implementation scenario. Assuming that the length of the target image is L and the width is W, the total number of grids corresponding to the rasterization of the target image is obtained according to the length and width of the target image and the sandblasting diameter D:

[0049]

[0050] Where n represents the total number of grids, L represents the length of the target image, W represents the width of the target image, and D represents the sandblasting diameter.

[0051] According to the total number of grids, the grid method is used to obtain the grid image corresponding to the target image, such as Figure 2 As shown, it is a schematic diagram of a raster image. Figure 2 The black area is the obstacle area and does not need to be sprayed, and the white area is the workpiece surface area to be sprayed.

[0052] At this point, the raster image corresponding to the large workpiece to be sandblasted is obtained.

[0053] Step S102, according to the obstacle area in the raster image, obtain the adaptive block size when dividing the raster image, divide the raster image into multiple block areas according to the adaptive block size, and use the ant colony algorithm to obtain the first optimal path between all the block areas.

[0054] Since there are a large number of grids in a grid image, if each grid is used as a path node, the amount of calculation when using the ant colony algorithm to obtain the optimal path is too large, resulting in poor quality of the obtained optimal path. Therefore, in an embodiment of the present invention, by dividing the grid image into regions and using the divided regions as path nodes, the amount of calculation of the ant colony algorithm can be reduced. Then, it is necessary to determine the size of the blocks. If the blocks are too small, the degree of optimization of the calculation amount is low. If the blocks are too large, the optimal path obtained by the ant colony algorithm is not representative for smaller grids.

[0055] Since the optimal path needs to consider the avoidance of obstacles, the divided block area is preferably able to include the entire obstacle or exclude the obstacle. In the embodiment of the present invention, according to the obstacle distribution characteristics in the raster image, an adaptive block size is obtained for the block division of the raster image. Wherein, according to the obstacle area in the raster image, the adaptive block size when the raster image is divided into blocks is obtained, including:

[0056] For any obstacle region in the raster image, count the number of side lengths of the obstacle region, calculate the area ratio between the minimum circumscribed rectangle of the obstacle region and the obstacle region, use the product of the reciprocal of the number of side lengths and the area ratio as a weight, and based on the weight, perform weighted summation on the square of each side length in the obstacle region to obtain a side length characteristic value of the obstacle region;

[0057] Obtain a side length characteristic value of each obstacle area in the raster image, accumulate all side length characteristic values ​​to obtain an obstacle distribution characteristic value in the raster image, and obtain an adaptive block size when dividing the raster image into blocks according to the obstacle distribution characteristic value.

[0058] In one embodiment, the minimum bounding rectangle of each obstacle area in the raster image is obtained respectively, and the area ratio between the minimum bounding rectangle and the corresponding obstacle area is calculated. ,in, represents the area of ​​the ith obstacle region, The area of ​​the minimum circumscribed rectangle of the ith obstacle area is represented. The side length and the number of side lengths of each obstacle area are counted. Then, according to the side length, the number of side lengths and the area ratio of each obstacle area, the obstacle distribution characteristic value in the raster image is obtained. The calculation expression of the obstacle distribution characteristic value is:

[0059]

[0060] Among them, R represents the characteristic value of obstacle distribution, N represents the number of obstacle areas, represents the number of side lengths of the i-th obstacle area, represents the j-th edge length in the i-th obstacle area, represents the area of ​​the ith obstacle region, represents the area of ​​the minimum bounding rectangle of the ith obstacle area, Represents the area ratio of the i-th obstacle area, and 1 represents a constant.

[0061] It should be noted that It is used to increase the influence of each side length. When the cumulative sum of the side lengths of all obstacle areas is smaller, it means that most of the side lengths of the obstacle areas are smaller, and the corresponding adaptive block size should be smaller; The larger the value of is, the better the integrity of the corresponding obstacle area contained by its corresponding minimum enclosing rectangle is, and the more representative each side length in the corresponding obstacle area is. Similarly, The smaller the value of is, the less complex the shape of the corresponding obstacle is, and the more representative each side length of the corresponding obstacle area is. As the weight of each edge length in the corresponding obstacle area, that is, The larger the value of The smaller the value of , the higher the weight of each side length in the corresponding obstacle area. The weighted sum of the side lengths of all obstacle areas will result in a larger obstacle distribution characteristic value.

[0062] Considering the side lengths of all obstacles in the comprehensive raster image, if the side lengths of multiple obstacles in the raster image are smaller, the corresponding blocks need to be smaller. If the area ratio between multiple obstacles and their minimum enclosing rectangles is larger and the number of side lengths is smaller, it means that the minimum enclosing rectangle can more completely contain the entire obstacle, and each side length of the corresponding obstacle is more representative, that is, the larger the obstacle distribution characteristic value, then based on the obstacle distribution characteristic value, the adaptive block size when dividing the raster image can be obtained. The specific method is as follows:

[0063] Among the side lengths of each obstacle area in the raster image, the maximum side length and the minimum side length are obtained, the side length difference between the maximum side length and the minimum side length is calculated, the product value of the obstacle distribution characteristic value and the side length difference of a preset multiple is obtained, the sum of the product value and the minimum side length is rounded, and the adaptive block size is obtained accordingly.

[0064] In one implementation, the maximum side length and the minimum side length of the side lengths corresponding to all obstacles in the raster image are used as a range, and the adaptive block size is obtained in combination with the obstacle distribution characteristic value:

[0065]

[0066]

[0067] Among them, d is the side length of the adaptive block, R represents the characteristic value of obstacle distribution, represents the normalized value of the obstacle distribution characteristic value, represents the maximum side length, represents the minimum side length, Indicates the floor symbol.

[0068] It should be noted that Represents the result of normalizing the side length of the adaptive block. The result corresponding to the normalization of the obstacle distribution characteristic value is represented. By making the normalized results of the two equal or close to equal, the side length value corresponding to the obstacle distribution characteristic value within the obstacle side length range can be obtained, and then the final side length of the adaptive block is obtained by rounding down.

[0069] Since the obtained side length d of the adaptive block combines the comprehensive side length and shape characteristics of the obstacles in the raster image, it can better avoid the influence of obstacles during the path traversal process. The grid image is divided into blocks with the adaptive block size to obtain multiple block areas. It is worth noting that the boundary of the grid image does not meet the The range is based on the area actually located in the raster image.

[0070] At this point, the raster image is divided into multiple block areas, and then the first layer of operation trajectory of the sandblasting robot during sandblasting operation is obtained between the multiple block areas. Specifically, the center of gravity of each block area is obtained respectively, and the ant colony algorithm is used to obtain the shortest path between all the centers of gravity as the first optimal path between all the block areas.

[0071] Step S103, for any block area, obtain the regional growth standard value between every two adjacent grids in the block area, perform regional growth on the block area according to the regional growth standard value, obtain at least one growth area in the block area, and use the ant colony algorithm to obtain the second optimal path between all growth areas and the third optimal path of each growth area.

[0072] During the sandblasting process, different material parts of the same workpiece surface may present different degrees of roughness, and the sandblasting robot will adjust the sandblasting height and angle, etc. according to the material characteristics and roughness of the area to be sandblasted during the operation of the sandblasting trajectory. Therefore, during the sandblasting process, if the adjacent sandblasting areas in the sandblasting path are of the same material or have the same degree of roughness, it can greatly reduce the wear and function loss of the mechanical arm caused by the sandblasting robot adjusting the sandblasting height or angle, and reduce the inaccuracy of sandblasting at the boundary caused by switching the sandblasting mode. In this embodiment of the present invention, for each block area in the grid image, it is necessary to obtain the optimal path of its grid, that is, the specific operation trajectory of the sandblasting robot in the block area.

[0073] Taking a block area as an example, a regional growing algorithm is used to perform regional growth on the grids in the block area to obtain regions with similar features. Specifically, the first grid in the upper left corner of the block area is used as the initial seed point, and the regional growth standard value between two adjacent grids is less than or equal to the preset standard value threshold as the growth criterion, and the block area is regionally grown to obtain at least one growth region in the block area. Among them, the preset standard value threshold is 0.4, which is not limited here and can be set according to the implementation scenario.

[0074] It should be noted that the embodiment of the present invention replaces the growth criterion in the region growing algorithm with the region growth standard value between two adjacent grids being less than or equal to a preset standard value threshold. The corresponding two grids can be merged, and other steps remain unchanged. Since the region growing algorithm belongs to the prior art, it will not be repeated here.

[0075] Among them, the method for obtaining the regional growth standard value between two adjacent grids is:

[0076] (1) For any grid in the block area, the material feature value corresponding to the grid is obtained according to the grayscale values ​​of all pixels corresponding to the grid in the surface structure image.

[0077] Specifically, the mean grayscale value of all pixels corresponding to the grid in the surface structure image is calculated as the material feature value corresponding to the grid.

[0078] In one implementation, taking a grid as an example, the grid image is projected into the surface structure image to obtain a plurality of pixel points corresponding to the grid in the surface structure image, and then the material feature value corresponding to the grid is obtained according to the grayscale values ​​of these pixel points:

[0079]

[0080] Among them, G represents the material characteristic value corresponding to the grid, M represents the number of pixels corresponding to the grid in the surface structure image, Represents the gray value of the i-th pixel corresponding to the grid in the surface structure image.

[0081] It should be noted that, considering that the different materials on the workpiece surface reflect different grayscale values ​​in the grayscale image, that is, there are grayscale value differences between pixels, therefore, the grayscale value mean of all pixels in the surface structure image corresponding to the grid is used as one of the regional growth conditions. The closer the grayscale value means between adjacent grids are, the more suitable it is for regional growth, that is, the more it satisfies the growth criteria, and the corresponding growth area is less likely to need frequent adjustments during the sandblasting process.

[0082] (2) Obtaining a regional roughness feature value corresponding to the grid according to the grayscale values ​​of the neighboring pixels of each pixel corresponding to the grid in the surface structure image.

[0083] Specifically, for any pixel point corresponding to the grid in the surface structure image, among the pixel points within a preset neighborhood range of the pixel point, a pixel point existing in the grid is obtained as a target pixel point, and the absolute value of the grayscale difference between the pixel point and each of the target pixel points is obtained to obtain the sum of the absolute values ​​of the differences;

[0084] The sum of the absolute values ​​of the differences of each pixel point corresponding to the grid in the surface structure image is accumulated to obtain the regional roughness characteristic value corresponding to the grid.

[0085] In one embodiment, the preset neighborhood range is set to eight neighborhoods, and the regional roughness feature value corresponding to the grid is obtained by analyzing the gray value difference between each pixel point corresponding to the grid in the surface structure image and its eight neighborhoods:

[0086]

[0087] Among them, C represents the regional roughness feature value corresponding to the grid, M represents the number of pixels corresponding to the grid in the surface structure image, Represents the gray value of the i-th pixel corresponding to the grid in the surface structure image, represents the grayscale value of the pixel in the grid in the eight-neighborhood of the i-th pixel, || represents the absolute value sign, and K represents the number of pixels in the grid in the eight-neighborhood of the i-th pixel.

[0088] It should be noted that Represents the sum of grayscale differences in the neighborhood of the i-th pixel. The larger its value, the stronger the contrast between the i-th pixel and the surrounding adjacent pixels, and the stronger the concave-convex roughness of the corresponding area. Therefore, The larger the value, the greater the overall regional roughness corresponding to the grid.

[0089] (3) For any two adjacent grids in the block area, a first difference absolute value of the material characteristic values ​​and a second difference absolute value of the regional roughness characteristic values ​​between the two adjacent grids are calculated respectively, and the first difference absolute value and the second difference absolute value are normalized respectively to obtain a first normalized value and a second normalized value respectively, and a regional growth standard value between the two adjacent grids is obtained according to the mean of the first normalized value and the second normalized value.

[0090] In one embodiment, the calculation expression of the region growth criterion value between two adjacent grids is:

[0091]

[0092] Among them, S represents the regional growth standard value between two adjacent grids, norm() represents the normalization function, || represents the absolute value symbol, and C represents the regional roughness characteristic value of one grid between two adjacent grids. represents the regional roughness characteristic value of the other grid in two adjacent grids, G represents the material characteristic value corresponding to one grid in two adjacent grids, Indicates the material characteristic value corresponding to the other grid in two adjacent grids.

[0093] It should be noted that Characterizes the sum of the eigenvalue differences between the two grids. The smaller the sum of the eigenvalue differences, the closer the roughness and material characteristics between the two grids are, and the more consistent the correspondence is with the regional growth criterion, then the larger the regional growth standard value between the two grids.

[0094] Based on the above-mentioned regional growth method, multiple growth regions in each block area can be obtained respectively, and then the second-layer running trajectory of the sandblasting robot when performing sandblasting operation is obtained for each block area. For any block area, the center of gravity of each growth area is obtained, and the ant colony algorithm is used to obtain the closest path between all centers of gravity as the second optimal path between all growth areas. Furthermore, for each growth area in the block area, it is also necessary to obtain the third-layer running trajectory of the sandblasting robot when performing sandblasting operation. Specifically: for any growth area, the ant colony algorithm is used to obtain the closest path between all grids in the growth area as the third optimal path of the growth area.

[0095] At this point, the first optimal path between all block areas, the second optimal path between all growth areas in each block area, and the third optimal path of each growth area are obtained, and these are used as the overall optimal path for the sandblasting robot to sandblast large workpieces.

[0096] Step S104, performing sandblasting operation on the large workpiece to be sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each block area.

[0097] After obtaining the overall optimal path, the sandblasting robot can perform sandblasting operations on large workpieces to be sandblasted. The specific process of sandblasting operations is as follows:

[0098] When performing sandblasting operation on the large workpiece to be sandblasted, based on the first optimal path, a first traversed block area is obtained, based on the second optimal path of the first traversed block area, a first traversed growth area in the first traversed block area is obtained, and based on the third optimal path of the first traversed growth area, sandblasting operation is performed on the first traversed growth area;

[0099] After the sandblasting operation of the first traversed growth area is completed, based on the second optimal path of the first traversed block area, a second traversed growth area in the first traversed block area is obtained, and based on the third optimal path of the second traversed growth area, the second traversed growth area is sandblasted, and so on, until the sandblasting operation of the first traversed block area is completed;

[0100] Based on the first optimal path, a second traversed block area is obtained, and sandblasting is performed on the second traversed block area according to the sandblasting method of the first traversed block area until the sandblasting of all the block areas is completed.

[0101] In summary, the embodiment of the present invention obtains a surface structure image of a large workpiece to be sandblasted, marks and distinguishes the sprayable area and the obstacle area in the surface structure image to obtain a target image, and obtains a grid image corresponding to the target image according to the grid method and a preset sandblasting diameter; according to the obstacle area in the grid image, an adaptive block size when dividing the grid image is obtained, and according to the adaptive block size, the grid image is divided into multiple block areas, and the first optimal path between all the block areas is obtained by using an ant colony algorithm; for any block area, the regional growth standard value between each two adjacent grids in the block area is obtained respectively, and the block area is subjected to regional growth according to the regional growth standard value to obtain at least one growth area in the block area, and the second optimal path between all the growth areas and the third optimal path of each of the growth areas are obtained respectively by using the ant colony algorithm; according to the first optimal path and the second optimal path and the third optimal path corresponding to each of the block areas, the large workpiece to be sandblasted is sandblasted. Among them, by combining the distribution characteristics of obstacle areas in the raster image, the sandblasting equipment is divided into three layers of sandblasting path trajectories, namely the first optimal path between all block areas in the raster image, the second optimal path between all growth areas in each block area, and the third optimal path in each growth area. Then, based on these three layers of sandblasting path trajectories, sandblasting operations of large workpieces to be sandblasted are carried out, avoiding the problem of high computational complexity of the ant colony algorithm when facing a large number of grids, so as to obtain a sandblasting path with better sandblasting quality, making local sandblasting operations more delicate and improving the efficiency of sandblasting operations.

[0102] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. An intelligent monitoring system for sandblasting equipment operation trajectory data based on artificial intelligence, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the following steps are implemented: Acquire a surface structure image of a large workpiece to be sandblasted, mark and distinguish the sprayable area and the obstacle area in the surface structure image to obtain a target image, and acquire a grid image corresponding to the target image according to a grid method and a preset sandblasting diameter; According to the obstacle area in the grid image, an adaptive block size when dividing the grid image is obtained, the grid image is divided into a plurality of block areas according to the adaptive block size, and an ant colony algorithm is used to obtain a first optimal path between all the block areas; For any block area, respectively obtain the regional growth standard value between every two adjacent grids in the block area, perform regional growth on the block area according to the regional growth standard value to obtain at least one growth area in the block area, and use the ant colony algorithm to respectively obtain the second optimal path between all the growth areas and the third optimal path of each of the growth areas; Performing sandblasting operation on the large workpiece to be sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each of the block areas; The step of obtaining an adaptive block size when dividing the grid image into blocks according to the obstacle area in the grid image comprises: For any obstacle region in the raster image, count the number of side lengths of the obstacle region, calculate the area ratio between the minimum circumscribed rectangle of the obstacle region and the obstacle region, use the product of the reciprocal of the number of side lengths and the area ratio as a weight, and based on the weight, perform weighted summation on the square of each side length in the obstacle region to obtain a side length characteristic value of the obstacle region; Obtaining a side length characteristic value of each obstacle region in the raster image, accumulating all side length characteristic values ​​to obtain an obstacle distribution characteristic value in the raster image, obtaining a maximum side length and a minimum side length among the side lengths of each obstacle region in the raster image, calculating a side length difference between the maximum side length and the minimum side length, obtaining a product value of a preset multiple of the obstacle distribution characteristic value and the side length difference, performing a rounding operation on a sum of the product value and the minimum side length, and obtaining an adaptive block size accordingly; The step of respectively obtaining the regional growth standard value between every two adjacent grids in the block area comprises: For any grid in the block area, the material feature value corresponding to the grid is obtained according to the grayscale values ​​of all pixels corresponding to the grid in the surface structure image, and the regional roughness feature value corresponding to the grid is obtained according to the grayscale values ​​of the neighboring pixels of each pixel corresponding to the grid in the surface structure image; For any two adjacent grids in the block area, the first absolute difference value of the material characteristic values ​​and the second absolute difference value of the regional roughness characteristic values ​​between the two adjacent grids are calculated respectively, the first absolute difference value and the second absolute difference value are normalized respectively to obtain a first normalized value and a second normalized value respectively, and the regional growth standard value between the two adjacent grids is obtained according to the mean of the first normalized value and the second normalized value.

2. The sandblasting equipment operation trajectory data intelligent monitoring system based on artificial intelligence according to claim 1 is characterized in that: The method of using the ant colony algorithm to obtain the first optimal path between all the block areas includes: The centroid of each of the sub-block areas is obtained respectively, and the shortest path between all the centroids is obtained by using an ant colony algorithm as the first optimal path between all the sub-block areas.

3. The sandblasting equipment operation trajectory data intelligent monitoring system based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining the material feature value corresponding to the grid according to the grayscale values ​​of all pixels corresponding to the grid in the surface structure image includes: The grayscale value average of all pixels corresponding to the grid in the surface structure image is calculated as the material feature value corresponding to the grid.

4. The sandblasting equipment operation trajectory data intelligent monitoring system based on artificial intelligence according to claim 1 is characterized in that: The step of obtaining the regional roughness feature value corresponding to the grid according to the grayscale value of the neighboring pixel points of each pixel point corresponding to the grid in the surface structure image comprises: For any pixel point corresponding to the grid in the surface structure image, among the pixels within a preset neighborhood range of the pixel point, a pixel point existing in the grid is obtained as a target pixel point, and the absolute value of the grayscale difference between the pixel point and each of the target pixel points is obtained to obtain the sum of the absolute values ​​of the differences; The sum of the absolute values ​​of the differences of each pixel point corresponding to the grid in the surface structure image is accumulated to obtain the regional roughness characteristic value corresponding to the grid.

5. The sandblasting equipment operation trajectory data intelligent monitoring system based on artificial intelligence according to claim 1 is characterized in that: The performing region growing on the block region according to the region growing standard value to obtain at least one growing region in the block region comprises: Taking the first grid in the upper left corner of the block area as the initial seed point, and taking the regional growth standard value between two adjacent grids being less than or equal to a preset standard value threshold as the growth criterion, regional growth is performed on the block area to obtain at least one growth area in the block area.

6. The sandblasting equipment operation trajectory data intelligent monitoring system based on artificial intelligence according to claim 1 is characterized in that: The using the ant colony algorithm to respectively obtain the second optimal path between all the growth areas and the third optimal path of each of the growth areas comprises: Obtaining the centroid of each of the growth regions, and using the ant colony algorithm to obtain the shortest path between all centroids as the second optimal path between all the growth regions; For any growth region, the ant colony algorithm is used to obtain the shortest path between all grids in the growth region as the third optimal path of the growth region.

7. The sandblasting equipment operation trajectory data intelligent monitoring system based on artificial intelligence according to claim 1 is characterized in that: The sandblasting operation is performed on the large workpiece to be sandblasted according to the first optimal path and the second optimal path and the third optimal path corresponding to each of the block areas, comprising: When performing sandblasting operation on the large workpiece to be sandblasted, based on the first optimal path, a first traversed block area is obtained, based on the second optimal path of the first traversed block area, a first traversed growth area in the first traversed block area is obtained, and based on the third optimal path of the first traversed growth area, sandblasting operation is performed on the first traversed growth area; After the sandblasting operation of the first traversed growth area is completed, based on the second optimal path of the first traversed block area, a second traversed growth area in the first traversed block area is obtained, and based on the third optimal path of the second traversed growth area, the second traversed growth area is sandblasted, and so on, until the sandblasting operation of the first traversed block area is completed; Based on the first optimal path, a second traversed block area is obtained, and sandblasting is performed on the second traversed block area according to the sandblasting method of the first traversed block area until the sandblasting of all the block areas is completed.

Images