A path planning method and system for laser cleaning

Abstract

The application provides a path planning method and system for laser cleaning, and relates to the technical field of laser cleaning.The path planning method for laser cleaning comprises the following steps: acquiring a binary image of a target area; generating a to-be-cleaned area according to a binary value of the binary image; segmenting the to-be-cleaned area to generate at least one to-be-cleaned sub-area; acquiring a vertex of the to-be-cleaned sub-area; determining a cleaning path starting point of the to-be-cleaned sub-area by using an optimization algorithm according to the vertex, and generating a sub-area cleaning path by using an arc-shaped scanning path method according to the cleaning path starting point; and generating a cleaning path of the target area according to the sub-area cleaning path.Through the path planning method, the unnecessary cleaning area can be effectively avoided, the cleaning cost is reduced, the to-be-cleaned area can be automatically subjected to full-coverage path planning, manual calibration assistance is not needed, the operation and maintenance cycle is reduced, and the efficiency of laser cleaning is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of laser cleaning, in particular to a path planning method and system for laser cleaning. BACKGROUND

[0002] In the field of industrial manufacturing, with the development of laser technology, a series of industrial methods based on laser processing and manufacturing technology have gradually become a research hotspot. Laser cleaning technology, as an emerging branch of laser processing and manufacturing, has the advantages of environmental protection, high flexibility, non-contact, and the ability to achieve high-quality and efficient cleaning. With the continuous in-depth research of researchers on the mechanism of laser cleaning, laser cleaning technology has been able to reliably clean the surfaces of different substrates, becoming a disruptive cleaning technology that is gradually replacing traditional cleaning methods such as machinery and chemistry. Its cleaning objects include steel, aluminum alloy, titanium alloy, and composite materials, and the application industries cover aerospace, aviation, shipbuilding, high-speed rail, automobile, mold, nuclear power, and marine fields.

[0003] The most common phenomenon of the cleaning object substrate is local random dirt and residue. The current existing laser cleaning technology processing method is to clean the work through manual calibration assisted cleaning or full coverage cleaning, but the intervention of too much manpower or full coverage cleaning not only increases the operation and maintenance cycle, but also increases the cleaning cost, affecting the efficiency of laser cleaning. SUMMARY

[0004] The problem solved by the present application is how to improve the efficiency of laser cleaning.

[0005] To solve the above problems, in a first aspect, the present application provides a path planning method for laser cleaning, comprising:

[0006] obtaining a binary image of a target area;

[0007] generating a to-be-cleaned area according to the binary value of the binary image;

[0008] segmenting the to-be-cleaned area to generate at least one to-be-cleaned sub-area;

[0009] obtaining the vertex of the to-be-cleaned sub-area;

[0010] determining the cleaning path starting point of the to-be-cleaned sub-area according to the vertex by using an optimization algorithm, and generating a sub-area cleaning path according to the cleaning path starting point by using an arc-shaped scanning path method;

[0011] generating a cleaning path for the target area according to the sub-area cleaning path.

[0012] Optionally, the segmenting the to-be-cleaned area to generate at least one to-be-cleaned sub-area comprises:

[0013] acquire a sub-configuration space of the region to be cleaned;

[0014] slice scan the sub-configuration space to generate a topology of columns in the sub-configuration space;

[0015] generate a sub-region boundary of the sub-configuration space according to the topology;

[0016] divide the region to be cleaned according to the sub-region boundary to generate a sub-region to be cleaned.

[0017] Optionally, the generating of the sub-region boundary according to the topology comprises:

[0018] determining a first continuous effective region of a current column according to the topology, and judging whether a second continuous effective region exists in a next column;

[0019] when the second continuous effective region does not exist in the next column, generating the sub-region boundary according to the first continuous effective region;

[0020] when the second continuous effective region exists in the next column, extracting the second continuous effective region according to a minimum neighborhood, judging a connected relationship between the first continuous effective region and the second continuous effective region, and generating the sub-region boundary according to the connected relationship.

[0021] Optionally, the extracting of the second continuous effective region according to the minimum neighborhood comprises:

[0022] acquiring a first center point of the current column and a discontinuous effective region of the next column;

[0023] acquiring a second center point in the discontinuous effective region closest to the first center point;

[0024] scanning invalid boundaries of the second center point column towards both ends in combination with the minimum neighborhood;

[0025] acquiring effective points in the invalid boundaries, integrating the effective points, and generating the second continuous effective region.

[0026] Optionally, the generating of the sub-region boundary according to the connected relationship comprises:

[0027] when the connected relationship is connected, iteratively acquiring the second continuous effective region of the next column until the connected relationship is disconnected;

[0028] when the connected relationship is disconnected, generating the sub-region boundary according to the first continuous effective region and the second continuous effective region.

[0029] Optionally, the cleaning path start point of the sub-area to be cleaned is determined according to the vertex by using an optimization algorithm, and a sub-area cleaning path is generated according to the cleaning path start point by using an arc-shaped scanning path method, comprising:

[0030] The cleaning path start point of the sub-area to be cleaned is determined according to the vertex and a preset rule;

[0031] The sub-area cleaning path of the sub-area to be cleaned is generated according to the cleaning path start point by using the arc-shaped scanning path method, and a cleaning path end point of the sub-area cleaning path is obtained;

[0032] The cleaning path start point of the next sub-area to be cleaned is determined according to the cleaning path end point by using the optimization algorithm, and the sub-area cleaning path and the cleaning path end point of the next sub-area to be cleaned are generated by using the arc-shaped scanning path method;

[0033] The sub-area cleaning path and the cleaning path end point are obtained in a loop until the sub-area cleaning path of the last sub-area to be cleaned is obtained.

[0034] Optionally, the sub-area cleaning path of the sub-area to be cleaned is generated according to the cleaning path start point by using the arc-shaped scanning path method, comprising:

[0035] The position of the cleaning path start point is obtained according to the cleaning path start point;

[0036] The path planning direction of the arc-shaped scanning path method is determined according to the position by matching a preset rule base;

[0037] The sub-area cleaning path is generated according to the path planning direction and the arc-shaped scanning path method.

[0038] Optionally, after the sub-area cleaning path of the sub-area to be cleaned is generated according to the cleaning path start point by using the arc-shaped scanning path method, and the cleaning path end point of the sub-area cleaning path is obtained, before the cleaning path start point of the next sub-area to be cleaned is determined according to the cleaning path end point by using the optimization algorithm, it further comprises:

[0039] The serial numbers of all the sub-areas to be cleaned are obtained;

[0040] The sub-area information tabu table is generated according to all the serial numbers and the corresponding vertexes;

[0041] The cleaning path start point of the next sub-area to be cleaned is determined according to the cleaning path end point by using the optimization algorithm, comprising:

[0042] delete the serial number and vertex corresponding to the cleaning path end point in the sub-region information tabu list, and generate an un-planned information tabu list;

[0043] According to the cleaning path end point, the optimization algorithm is used to traverse the un-planned information tabu list to determine the cleaning path start point of the next sub-region to be cleaned in the un-planned information tabu list.

[0044] Optionally, the generating of the region to be cleaned according to the binary value of the binary image comprises:

[0045] An image discrete pixel unit of the binary image is obtained.

[0046] The image discrete pixel unit is converted into an equivalent spot coverage unit, and the binary value is obtained.

[0047] The region to be cleaned is generated according to the equivalent spot coverage unit, the binary value and a region to be cleaned obtaining formula, wherein the region to be cleaned obtaining formula comprises:

[0048] A={(x,y)|x∈[1,row],y∈[1,col]};

[0049] Wherein, A is the region to be cleaned, (x, y) is the horizontal and vertical coordinates of the equivalent spot coverage unit with the binary value of 1, row is the total number of rows of the binary image, and col is the total number of columns of the binary image.

[0050] In a second aspect, the present application provides a path planning system for laser cleaning, comprising:

[0051] An image module is configured to obtain a binary image of a target region.

[0052] A region to be cleaned module is configured to generate a region to be cleaned according to a binary value of the binary image.

[0053] A segmentation module is configured to segment the region to be cleaned to generate at least one sub-region to be cleaned.

[0054] A vertex module is configured to obtain a vertex of the sub-region to be cleaned.

[0055] A sub-region cleaning path module is configured to determine a cleaning path start point of the sub-region to be cleaned according to the vertex by using an optimization algorithm, and generate a sub-region cleaning path according to the cleaning path start point by using an arc-shaped scanning path method.

[0056] A cleaning path module is configured to generate a cleaning path of the target region according to the sub-region cleaning path.

[0057] The path planning method and system for laser cleaning have the following beneficial effects:

[0058] By the binarization value of the target area binarization image, the to-be-cleaned area and the non-cleaned area can be clearly distinguished, the to-be-cleaned area is divided into at least one to-be-cleaned sub-area, full coverage path planning is facilitated, the optimal vertex is selected as the cleaning path starting point in the vertex of the to-be-cleaned sub-area by using an optimization algorithm, the full coverage path planning is performed on the to-be-cleaned sub-area according to the cleaning path starting point by using the arc-shaped scanning path method, the sub-area cleaning path of the to-be-cleaned sub-area is generated, and finally the cleaning path of the target area is generated according to the sub-area cleaning path, which not only effectively avoids the non-cleaned area and reduces the cleaning cost, but also automatically performs full coverage path planning on the to-be-cleaned area, does not need manual calibration assistance, reduces the operation and maintenance cycle, and effectively improves the efficiency of laser cleaning. BRIEF DESCRIPTION OF DRAWINGS

[0059] Figure 1 A flowchart of a path planning method for laser cleaning according to an embodiment of the present application;

[0060] Figure 2 A structure diagram of a path planning system for laser cleaning according to an embodiment of the present application;

[0061] Figure 3 A diagram of a binarization image according to an embodiment of the present application;

[0062] Figure 4 A diagram of a to-be-cleaned sub-area according to an embodiment of the present application;

[0063] Figure 5 A diagram of a to-be-cleaned sub-area vertex according to an embodiment of the present application;

[0064] Figure 6 A diagram of a cleaning path according to an embodiment of the present application. DETAILED DESCRIPTION

[0065] In order to make the above objectives, features and advantages of the present application more obvious and easy to understand, the specific embodiments of the present application will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present application are shown in the drawings, it should be understood that the present application can be implemented in various forms, and should not be interpreted as being limited to the embodiments described herein, on the contrary, these embodiments are provided to make the present application more thorough and complete. It should be understood that the drawings and embodiments of the present application are only for illustrative purposes, and are not intended to limit the scope of protection of the present application.

[0066] It should be understood that each of the steps recited in the method embodiments of the present application can be performed in different orders and / or in parallel. In addition, the method embodiments can include additional steps and / or omit performing the steps shown. The scope of the present application is not limited in this regard.

[0067] The term "comprises" and variations thereof such as "comprising" and "comprises" as used herein are open-ended, that is, "comprising but not limited to." The term "based on" is "based, at least in part, on." The term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; the term "optional" means "optional embodiments." Related definitions are given throughout the description below. It should be noted that the concepts mentioned in the present application are merely illustrative and not limiting, and those skilled in the art should understand that "one", "multiple" modification is illustrative and not limiting, and unless otherwise explicitly indicated in the context, it should be understood as "one or more". The concepts mentioned in the present application are merely illustrative and not limiting, and those skilled in the art should understand that "first", "second" and the like concepts are merely used to distinguish different devices, systems or units, and are not used to limit the order or interdependence of the functions performed by these devices, systems or units.

[0068] It should be noted that the "one", "multiple" modification mentioned in the present application is illustrative and not limiting, and those skilled in the art should understand that unless otherwise explicitly indicated in the context, it should be understood as "one or more". The concepts mentioned in the present application are merely illustrative and not limiting, and those skilled in the art should understand that "first", "second" and the like concepts are merely used to distinguish different devices, systems or units, and are not used to limit the order or interdependence of the functions performed by these devices, systems or units.

[0069] To solve the above problems, in one embodiment, as shown in Figure 1 The present application provides a path planning method for laser cleaning, which comprises

[0070] Step S1, obtaining a binary image of the target area.

[0071] Specifically, a CCD intelligent camera can be used to obtain the binary image of the target area. The CCD intelligent camera can perform image correction and coordinate matching on the captured image in combination with the installation coordinates of the workpiece and itself, and regard the cleaning area as an interest target, extract the target area in the image analysis process, and convert the extraction result into a binary image, wherein the binary image includes binary image display and binary coordinate matrix.

[0072] Step S2, generating a cleaning area according to the binary value of the binary image;

[0073] Specifically, according to the binarization value, the region with the binarization value of 1 is taken as the cleaning region, and the region with the binarization value of 0 is taken as the non-cleaning region. The non-cleaning region can be expressed as: B = {(x, y) ∈ A | (x, y) ∈ background}, where background is the region set with the binarization value of 0. As shown in Figure 3 , the white region of the binarization image is the cleaning region, and the black region is the background region, i.e., the non-cleaning region.

[0074] Step S3: segmenting the cleaning region to generate at least one cleaning sub-region;

[0075] Specifically, the Boustrophedon cell decomposition method is used to segment the cleaning region by referring to the minimum neighborhood technique to generate at least one cleaning sub-region, and each cleaning sub-region is numbered. The cleaning sub-region and its number are as shown in Figure 4 .

[0076] Step S4: obtaining the vertex of the cleaning sub-region;

[0077] Specifically, after obtaining the cleaning sub-region, the sub-region boundary of each cleaning sub-region is extracted, the sub-region vertex set of the column item reference is searched on each sub-region boundary, and all the vertices of each cleaning sub-region are marked. The vertex marking of the cleaning sub-region is as shown in Figure 5 .

[0078] Step S5: determining the cleaning path starting point of the cleaning sub-region by using an optimization algorithm according to the vertex, and generating the sub-region cleaning path by using the arc-shaped scanning path method according to the cleaning path starting point;

[0079] Specifically, according to a preset starting point rule, an initial starting point can be selected from all the vertices, a sub-region to be cleaned where the initial starting point is located can be taken as a starting sub-region, and a starting cleaning path of the starting sub-region can be generated by using an arc-shaped scanning path method for full coverage path planning; according to a terminal point of the starting cleaning path, an optimal vertex can be selected as a starting point for next path planning by using an optimization algorithm to traverse all the vertices of the remaining sub-regions to be cleaned, and a sub-region cleaning path of the sub-region to be cleaned where the vertex is located can be generated by using the arc-shaped scanning path method for full coverage path planning; according to a terminal point of the sub-region cleaning path, an optimal vertex can be selected as a starting point for next path planning by using the optimization algorithm to traverse all the vertices of the remaining sub-regions to be cleaned, and a sub-region cleaning path of the sub-region to be cleaned where the vertex is located can be generated by using the arc-shaped scanning path method for full coverage path planning; the optimization algorithm and the arc-shaped scanning path method can be repeatedly used to select a starting point for next path planning, and full coverage path planning can be performed to generate a sub-region cleaning path of the sub-region to be cleaned, until full coverage path planning of all the sub-regions to be cleaned is completed, and a sub-region cleaning path of the sub-region to be cleaned is generated. As an optimization, the optimization algorithm can be selected from a greedy algorithm, a particle swarm algorithm, an ant colony algorithm, and the like.

[0080] In one embodiment, the preset starting point rule is that a vertex at a left upper corner of a sub-region to be cleaned with a sequence number of 1, that is, a vertex at a left upper corner of a globally valid region, is taken as the initial starting point. It should be noted that the preset starting point rule can be set according to actual conditions.

[0081] In step S6, a cleaning path of the target region is generated according to the sub-region cleaning paths.

[0082] Specifically, a starting point of a sub-region cleaning path is determined according to a terminal point of a previous sub-region cleaning path, so that a final cleaning path of the target region can be generated by concatenating the initial starting point, the starting cleaning path, the terminal point of the starting cleaning path, a starting point of a next sub-region cleaning path, the next sub-region cleaning path, a terminal point of the next sub-region cleaning path, and all the sub-region cleaning paths, as shown in FIG. 8. Figure 6

[0083] Through the above steps, full-automatic path full coverage planning can be achieved without human intervention, and the target region can be cleaned without full coverage, which can greatly reduce the operation and maintenance cycle and effectively improve the efficiency of laser cleaning.

[0084] Optionally, the segmentation of the region to be cleaned to generate at least one sub-region to be cleaned comprises:

[0085] A sub-configuration space of the region to be cleaned is acquired.

[0086] ​slice scanning is performed on the sub-configuration space to generate a topology of columns in the sub-configuration space;

[0087] According to the topology, a sub-region boundary of the sub-configuration space is generated;

[0088] According to the sub-region boundary, the to-be-cleaned region is segmented to generate a to-be-cleaned sub-region.

[0089] Specifically, slice scanning is performed on the sub-configuration space by using a slice function, which is generally defined as a straight line linearly scanning a target region in a bounded two-dimensional plane. In the present application, due to the random island distribution characteristics of the dirt on the base material, the target region contains several sub-configuration spaces that need to be slice scanned. Accordingly, in the movement process of the slice function of the Boustrophedon cell decomposition method, a minimum neighborhood expansion method is introduced to constrain the matching of the slice and the sub-configuration space, and the topological structure after slice scanning is used to judge the region isolation, so as to generate the sub-region boundary of the sub-configuration space, segment the to-be-cleaned region, and generate the to-be-cleaned sub-region.

[0090] Optionally, the generating the sub-region boundary according to the topology comprises:

[0091] According to the topology, a first continuous effective region of the current column is determined, and it is judged whether a second continuous effective region exists in the next column;

[0092] When the second continuous effective region does not exist in the next column, the sub-region boundary is generated according to the first continuous effective region;

[0093] When the second continuous effective region exists in the next column, the second continuous effective region is extracted according to the minimum neighborhood, and the connected relationship between the first continuous effective region and the second continuous effective region is judged, and the sub-region boundary is generated according to the connected relationship.

[0094] Specifically, a column item straight line slice scanning method is used, and the current column item, i.e., the first continuous effective region in the jth column, is obtained by scanning up and down, which is denoted as A(start1:end1,j). Then, it is judged whether a second continuous effective region exists in the next column, i.e., the j+1th column.

[0095] When the effective region in the next column is empty or the number of columns in the next column exceeds the region boundary, it is judged that the second continuous effective region does not exist in the next column, the sub-region boundary of the sub-configuration space is directly generated, the sub-configuration space is numbered and valued, and then the operation of generating the sub-region boundary according to the topology is performed on the next sub-configuration space.

[0096] When the valid region in the next column is not empty or the column number of the next column does not exceed the region boundary, it is judged that the next column has a second continuous valid region, a minimum neighborhood is introduced, the second continuous valid region is extracted according to the minimum neighborhood, and the connected relationship between the first continuous valid region and the second continuous valid region is judged, and according to the connected relationship, the sub-region boundary is generated.

[0097] Optionally, the extracting the second continuous valid region according to the minimum neighborhood comprises:

[0098] The first center point of the current column and the discontinuous valid region of the next column are obtained.

[0099] The second center point closest to the center point in the discontinuous valid region is obtained.

[0100] The invalid boundary of the second center point column to both ends is scanned in combination with the minimum neighborhood.

[0101] The valid points in the invalid boundary are obtained, and the valid points are integrated to generate the second continuous valid region.

[0102] Specifically, according to the minimum neighborhood, the first center point of the current column is extracted, denoted as (mid,j), wherein mid=(start1+end1) / 2, the second center point closest to (mid,j) in the discontinuous valid region of the next column is traversed according to the distance formula, denoted as (scan,j+1), and the second center point is taken as the starting point, the continuous valid region is judged row by row in the upward and downward directions of the second center point in combination with the first continuous valid region, until the first invalid boundary of the second center point column to both ends, i.e. the boundary of the invalid region is scanned, the valid points in the invalid boundary are obtained, and the valid points are integrated to generate the second continuous valid region, denoted as B(start2:end2,j+1). The distance formula comprises:

[0103]

[0104] wherein D ij is the distance between the ith point and the jth point, x i and x j are the horizontal coordinates of the ith point and the jth point, y i and y j are the vertical coordinates of the ith point and the jth point.

[0105] It should be understood that when the second continuous effective area is extracted, each column can be the next column or the current column. For example, when the jth column is the current column and the second continuous effective area of the (j+1)th column is extracted, the (j+1)th column is the next column; when the second continuous effective area of the (j+2)th column is extracted, the (j+1)th column is the current column, and the second continuous effective area of the (j+1)th column is defined as the first continuous effective area.

[0106] Optionally, the generating the sub-region boundary according to the connectivity relationship comprises:

[0107] When the connectivity relationship is connected, the second continuous effective area of the next column is iteratively extracted until the connectivity relationship is disconnected.

[0108] When the connectivity relationship is disconnected, the sub-region boundary is generated according to the first continuous effective area and the second continuous effective area.

[0109] Specifically, the connectivity algorithm can be used to make a bilateral judgment of the topological structure, so as to judge the connectivity relationship. Only when both sides satisfy the condition of the unchanged topological structure, it is determined that the adjacent two columns are connected. When the connectivity relationship is connected, the second continuous effective area of the next column is iteratively extracted, for example, the second continuous effective areas of the (j+2)th column, the (j+3)th column, the (j+4)th column, the (j+5)th column and the (j+n)th column are extracted, until the connectivity relationship is disconnected. When the connectivity relationship is disconnected, it is indicated that the scanning inside the sub-configuration space is completed, the part that needs to be cleaned inside the sub-configuration space is determined, and the boundary of the sub-configuration space is determined. The sub-region boundary can be generated according to the first continuous effective area and the second continuous effective area of the column.

[0110] It should be understood that when the second continuous effective area is extracted, each column can be the next column or the current column. For example, when the jth column is the current column and the second continuous effective area of the (j+1)th column is extracted, the (j+1)th column is the next column; when the second continuous effective area of the (j+2)th column is extracted, the (j+1)th column is the current column, and the second continuous effective area of the (j+1)th column is defined as the first continuous effective area.

[0111] Optionally, the generating the sub-region boundary according to the connectivity relationship comprises:

[0112] The cleaning path starting point of the sub-region to be cleaned is determined according to the vertex and a preset rule.

[0113] According to the cleaning path starting point, the arc-shaped scanning path method is used to generate the sub-area cleaning path of the to-be-cleaned sub-area, and a cleaning path ending point of the sub-area cleaning path is obtained;

[0114] According to the cleaning path ending point, the optimization algorithm is used to determine a cleaning path starting point of a next to-be-cleaned sub-area, and the arc-shaped scanning path method is used to generate a sub-area cleaning path and a cleaning path ending point of the next to-be-cleaned sub-area;

[0115] The sub-area cleaning path and the cleaning path ending point are obtained in a loop until a last sub-area cleaning path of the to-be-cleaned sub-area is obtained.

[0116] Specifically, according to a preset rule, an initial starting point can be selected from all vertices, a to-be-cleaned sub-area where the initial starting point is located can be taken as a starting sub-area, and the arc-shaped scanning path method can be used for full-coverage path planning to generate a starting cleaning path of the starting sub-area; according to an ending point of the starting cleaning path, the optimization algorithm can be used to traverse all vertices of the remaining to-be-cleaned sub-areas, an optimal vertex can be selected as a starting point of next path planning, and the arc-shaped scanning path method can be used for full-coverage path planning on a to-be-cleaned sub-area where the vertex is located to generate a sub-area cleaning path of the to-be-cleaned sub-area; according to an ending point of the sub-area cleaning path, the optimization algorithm can be used to traverse all vertices of the remaining to-be-cleaned sub-areas, an optimal vertex can be selected as a starting point of next path planning, and the arc-shaped scanning path method can be used for full-coverage path planning on a to-be-cleaned sub-area where the vertex is located to generate a sub-area cleaning path of the to-be-cleaned sub-area; the optimization algorithm and the arc-shaped scanning path method can be repeatedly used to select a starting point of next path planning and perform full-coverage path planning to generate a sub-area cleaning path of a to-be-cleaned sub-area until full-coverage path planning is completed on all to-be-cleaned sub-areas to generate a sub-area cleaning path of the to-be-cleaned sub-area. The preset rule can be set according to actual conditions.

[0117] Optionally, the generating, according to the cleaning path starting point, of the sub-area cleaning path of the to-be-cleaned sub-area by using the arc-shaped scanning path method includes:

[0118] According to the cleaning path starting point, a position of the cleaning path starting point is obtained;

[0119] According to the position, a preset rule library is matched to determine a path planning direction of the arc-shaped scanning path method;

[0120] According to the path planning direction and the arc-shaped scanning path method, the sub-area cleaning path is generated.

[0121] Specifically, according to the cleaning path starting point, a position, that is, a coordinate, of the cleaning path starting point is obtained, and the coordinate includes:

[0122] n_a = [min(subarea{n}(:,min(y))), min(y)];

[0123] n_b = [max(subarea{n}(:,min(y))), min(y)];

[0124] n_c = [min(subarea{n}(:,max(y))), max(y)];

[0125] n_d = [max(subarea{n}(:,max(y))), max(y)], n = 1,2,3…N.

[0126] According to the position, matching the preset rule base, the path planning direction of the arc-shaped scanning path method is determined, and the preset rule base includes: when the cleaning path starting point is located at the minimum column boundary, that is, the cleaning path starting point is n_a or n_b, the column direction stepping inertia direction is assigned to right, otherwise, if the cleaning path starting point is n_c or n_d, the column direction stepping inertia direction is assigned to left; if the cleaning path starting point is located at the minimum row of the column boundary, that is, the cleaning path starting point is n_a or n_c, the initial row direction scanning stepping direction is assigned to down, otherwise, if the cleaning path starting point is n_b or n_d, the initial row direction scanning stepping direction is assigned to up. In the dynamic path planning process according to the path planning direction, the scanning path points are recorded one by one, if the effective row boundary of the column where the current scanning point is located is scanned, the row scanning direction is reversed after recording the current path point, and at the same time, the column direction is iteratively added according to the current stepping inertia direction, when the last column is scanned vertically, the path complete coverage of the sub-partition is completed, and the path trajectory and the scanning end point of the sub-partition are obtained.

[0127] Optionally, after the sub-area cleaning path of the to-be-cleaned sub-area is generated by using the arc-shaped scanning path method according to the cleaning path starting point, and the cleaning path end point of the sub-area cleaning path is obtained, before the cleaning path starting point of the next to-be-cleaned sub-area is determined by using the optimization algorithm according to the cleaning path end point, the method further includes:

[0128] obtaining sequence numbers of all the to-be-cleaned sub-areas;

[0129] generating a sub-area information tabu list according to all the sequence numbers and the corresponding vertexes;

[0130] the cleaning path starting point of the next to-be-cleaned sub-area is determined by using the optimization algorithm according to the cleaning path end point, including:

[0131] delete the serial number and vertex corresponding to the cleaning path end point in the sub-region information tabu table, and generate an un-planned information tabu table;

[0132] According to the cleaning path end point, the optimization algorithm is used to traverse the un-planned information tabu table to determine the cleaning path start point of the next sub-region to be cleaned in the un-planned information tabu table.

[0133] Specifically, after the region to be cleaned is segmented, at least one sub-region to be cleaned is generated, and each sub-region to be cleaned is numbered, the serial number of each sub-region to be cleaned is obtained, and the serial number of each sub-region to be cleaned and all vertices thereof are integrated into a sub-region information tabu table. When the path planning of the sub-region to be cleaned is completed, the serial number of the sub-region to be cleaned whose path planning is completed and all vertices thereof are deleted from the sub-region information tabu table, and an un-planned information tabu table is generated. When the path planning of the next sub-region to be cleaned is performed, the optimization algorithm traverses the un-planned information tabu table, and the cleaning path start point of the next sub-region to be cleaned is determined in the un-planned information tabu table, so as to ensure that the same sub-region to be cleaned is not repeatedly planned, and the efficiency of path planning is improved.

[0134] Optionally, the generating the region to be cleaned according to the binary value of the binary image comprises:

[0135] obtaining an image discrete pixel unit of the binary image;

[0136] converting the image discrete pixel unit into an equivalent light spot coverage unit, and obtaining the binary value;

[0137] generating the region to be cleaned according to the equivalent light spot coverage unit, the binary value, and a region to be cleaned obtaining formula, wherein the region to be cleaned obtaining formula comprises:

[0138] A={(x,y)|x∈[1,row],y∈[1,col]};

[0139] wherein A is the region to be cleaned, (x, y) is the horizontal and vertical coordinates of the equivalent light spot coverage unit whose binary value is 1, row is the total number of rows of the binary image, and col is the total number of columns of the binary image.

[0140] Specifically, as shown in Figure 3 the image discrete pixel unit is converted into the equivalent light spot coverage unit, so that each coverage unit of the binary image is a unit equal to the actual light spot size, the image data is better processed, image extraction and segmentation are performed, and the region to be cleaned in the binary image is extracted by using the region to be cleaned obtaining formula, and the region to be cleaned is removed.

[0141] In one embodiment, the path planning method of laser cleaning of the present application is applied to a laser cleaning device, and the workpiece to be cleaned is placed in the laser cleaning device, which comprises a workpiece table, a CCD intelligent camera mounted in a laser cleaning head, an industrial computer, and a cleaning laser controller. The workpiece to be cleaned is a certain size of a planar substrate workpiece with rust or paint dirt, and the material includes but is not limited to steel, aluminum alloy, titanium alloy, composite glass, etc. The CCD intelligent camera is integrated in the laser cleaning head in an off-axis manner. Before and after the laser cleaning process, the CCD intelligent camera realizes digital imaging acquisition, image processing, coordinate matching and other functions of the target workpiece in different cleaning stages by responding to the communication instructions of the industrial computer, and uploads the processed binary image data to the industrial computer.

[0142] The industrial computer comprises a communication module and a selected area planning module. The communication module of the industrial computer uses a communication link to complete the reception of image data and state of the intelligent camera, the reception of cleaning laser parameters, the analysis processing of the above received data, and the transmission of instructions and parameters to each execution unit. The selected area planning module of the industrial computer comprises an improved partition execution unit and a path planning execution unit. The partition execution unit realizes the regularized partition of the random target area by improved optimization, and obtains the coordinate set and sequence index of the sub-area by statistical analysis of the partition result. The path planning execution unit completes the global optimization series of the area and the optimized coverage in each sub-area based on the partition, and the order set of the coverage coordinate points is the planned scanning path of the laser cleaning under the current process. Finally, the obtained scanning path is sent to the cleaning laser controller in the form of a data packet.

[0143] The laser cleaning controller controls the parameters of the laser and the scanning path of the light beam by responding to the instructions of the industrial computer during the cleaning process, realizes the effective cleaning process of the laser on the workpiece, and dynamically returns the state of the laser and the scanning state of the light beam to the industrial computer.

[0144] The working principle of the laser cleaning device for obtaining binary images is as follows: power on each module of the laser cleaning device, and ensure the communication between the industrial computer and each execution unit. Adjust and fix the workpiece to be cleaned. If all the boundaries of the workpiece action section appear in the detection field of view of the CCD intelligent camera, control the CCD intelligent camera to take a picture of the workpiece. At the same time, the CCD intelligent camera can perform image correction and coordinate matching on the photographed image in combination with the installation coordinates of the workpiece and itself, and regard the area to be cleaned as an interesting target. In the image analysis process, the target area is extracted, and the analysis result is provided to the industrial computer in the form of a binary image through a communication link.

[0145] The industrial computer and the CCD intelligent camera are connected in the form of network port communication, the CCD intelligent camera is a server and the industrial computer is a client, the host computer sends instructions in a trigger mode to obtain the real-time state of the CCD intelligent camera and target binary image information, similarly, the industrial computer and the laser controller are connected in the form of network port communication, the laser controller is a server and the industrial computer is a client, the industrial computer sends parameterized instruction data packets to the laser controller at regular intervals and receives laser parameters and working states of the laser controller in real time. The industrial computer integrates and analyzes the received data to realize intuitive display of the binary image and the state of the actuator.

[0146] Exemplarily, in order to make the path planning method of laser cleaning of the application applied to the flow of the laser cleaning equipment more clear, the embodiment of the application provides a path planning step of the laser cleaning equipment, comprising:

[0147] The laser cleaning equipment is fixed and sampled, the position of the workpiece to be cleaned is adjusted to ensure that all boundaries of the action section are in the detection field of view of the CCD intelligent camera, the position of the workpiece to be cleaned at this time is calibrated, the industrial computer sends parameterized instructions to the CCD intelligent camera, including the issued photographing and image analysis instructions and the required spatial coordinate information in the analysis process, and the CCD intelligent camera returns the processed image information to the industrial computer in the form of a data packet after image sampling and analysis of the workpiece in response to the instructions. The application takes the path planning for the rusted area of a steel alloy base in engineering as an example, sets the color gamut offset range to 133-192 and the bad point gray threshold to 99 according to the color gamut characteristics of the rusted section of the rusted area of the steel alloy base, the CCD intelligent camera image processing obtains the corresponding binary image and binary coordinate matrix. The binary image is shown in Figure 3 The white area of the binary image is the area to be cleaned and the black area is the background area, that is, the area that does not need to be cleaned.

[0148] The binary coordinate matrix is segmented by using an improved Boustrophedon cell decomposition method to generate a sub-region to be cleaned, the number of effective comparison columns in a period is fixed with a difference of 1, the iteration step is 1, the topology environment information is buffered from the previous column of the comparison column, the step length is 4, the maximum and minimum row elements of the two columns involved in the comparison are judged for the topology structure, and only when both sides satisfy the topology structure invariance, it is determined that the adjacent two column regions are connected. After judging the next sub-region, the information of the sub-region is recorded in the form of an index data set, and the sub-region information in the original binary coordinate matrix is cleared, that is, the sequence number and the vertex corresponding to the cleaning path endpoint in the sub-region information tabu table are deleted, an un-planned information tabu table is generated, and all sub-regions to be cleaned and the corresponding sub-region index information are recorded. The boundaries of each sub-region are extracted, for example, the background color is set to black, the color gamut saturation is 60%, the partition boundary is colored, the color gamut saturation is 30%, and the inside of the partition is white filled with a saturation of 100%. The sub-region to be cleaned is as shown in Figure 4 .

[0149] The vertices are marked according to the sub-region boundary set. It should be noted that due to the randomness of the edge of the sub-region to be cleaned, the marked vertices may be repeated, but this does not affect the selection of the starting point of each sub-partition and the extraction of the path endpoint. For example, on the basis of the partition diagram, the vertices of each sub-partition are filled with red color with a color gamut saturation of 100%. The vertex marking of the sub-region to be cleaned is as shown in Figure 5 .

[0150] The target region is covered by the path planning in combination with the sub-region to be cleaned and the vertices thereof. According to a preset rule, the initial starting point and the starting sub-region are set, the starting sub-region and the sub-region to be cleaned are covered by the arc-shaped scanning path method, the starting cleaning path and the sub-region cleaning path are generated, and since the scanning step method and the vertex selection method are based on the column item, the final point of the covered path is also one of the sub-partition vertices.

[0151] The endpoints of the starting cleaning path and the sub-region cleaning path are recorded, the nearest vertex and the sub-region to be cleaned where the vertex is located are obtained by using an optimization algorithm to traverse the remaining sub-region to be cleaned, as the starting point and the sub-region to be cleaned of the next path planning, the path between the endpoint and the starting point of the next path planning is recorded, the next region is planned in the same way, the endpoint is extracted, and the subsequent iteration process is performed. For example, on the basis of the partition diagram, the starting point and the endpoint of the internal covering path of each sub-partition are filled with red color with a color gamut saturation of 100%, and the path is depicted by a red straight line. The cleaning path is as shown in Figure 6 .

[0152] In the planning process, the total path is the superposition of the distances of all adjacent path points from the first path point to the last path point, and is expressed as:

[0153] As described above, without excluding the particularity of the to-be-cleaned sub-region, if the effective coverage point of the current to-be-cleaned sub-region is only 1, the path start point and the path end point of the to-be-cleaned sub-region coincide at this time, and the coverage path is the only effective coverage point of the sub-region. According to the distance formula, the distance of the cleaning path of the sub-region is 0, but it does not affect the connection of the global region path. The cleaning path of the target region finally planned is the sequence of scanning coordinates of the laser beam acting on the substrate.

[0154] In another embodiment, as shown in FIG. 1, the present application provides a path planning system for laser cleaning, comprising: Figure 2

[0155] An image module for acquiring a binary image of a target region;

[0156] A to-be-cleaned region module for generating a to-be-cleaned region according to the binary value of the binary image;

[0157] A segmentation module for segmenting the to-be-cleaned region to generate at least one to-be-cleaned sub-region;

[0158] A vertex module for acquiring the vertex of the to-be-cleaned sub-region;

[0159] A sub-region cleaning path module for determining the cleaning path start point of the to-be-cleaned sub-region by using an optimization algorithm according to the vertex, and generating a sub-region cleaning path by using an arc scanning path method according to the cleaning path start point;

[0160] A cleaning path module for generating a cleaning path of the target region according to the sub-region cleaning path.

[0161] Although the present application is disclosed as above, the protection scope of the present application is not limited to this. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present application, and these changes and modifications will fall within the protection scope of the present application.​

Claims

1. A path planning method for laser cleaning, characterized in that, include: Obtain the binarized image of the target region; Based on the binarized values ​​of the binarized image, the area to be cleaned is generated; The area to be cleaned is divided to generate at least one sub-area to be cleaned; Obtain the vertices of the sub-region to be cleaned, extract the sub-region boundary of each sub-region to be cleaned, search the sub-region vertex set of the column reference on the sub-region boundary of each sub-region to be cleaned, and mark all the vertices of each sub-region to be cleaned. Based on the vertices, an optimization algorithm is used to determine the starting point of the cleaning path for the sub-region to be cleaned. Then, based on the starting point of the cleaning path, an arc-shaped scanning path method is used to generate the sub-region cleaning path, including: Based on the vertices and preset rules, an initial starting point is selected from all the vertices, and the sub-region to be cleaned where the initial starting point is located is taken as the starting sub-region. The initial starting point is then determined as the starting point of the cleaning path for the sub-region to be cleaned. Based on the starting point of the cleaning path, the cleaning path of the sub-region to be cleaned is generated using the bow-shaped scanning path method, and the cleaning path ending point of the sub-region cleaning path is obtained. Based on the endpoint of the cleaning path, the optimization algorithm is used to determine the starting point of the cleaning path for the next sub-region to be cleaned, and the bow-shaped scanning path method is used to generate the sub-region cleaning path and the endpoint of the cleaning path for the next sub-region to be cleaned. The process involves iteratively obtaining the cleaning path for the sub-region and the endpoint of the cleaning path until the cleaning path for the last sub-region to be cleaned is obtained; the step of generating the cleaning path for the sub-region to be cleaned using the bow-shaped scanning path method based on the starting point of the cleaning path includes: Based on the starting point of the cleaning path, obtain the position of the starting point of the cleaning path; Based on the location, a preset rule base is matched. During the dynamic path planning process according to the path planning direction, the scanning path points are recorded step by step. If the scanning reaches the valid row boundary of the column where the current scanning point is located, the scanning direction of the row item is reversed after recording the current path point, and the column direction is iteratively added according to the current step inertia direction. When the last column is scanned vertically, the path of the sub-region to be cleaned is completely covered, and the path trajectory and scanning endpoint of the sub-region to be cleaned are obtained. The sub-region cleaning path is generated based on the path planning direction and the bow-shaped scanning path method; Based on the cleaning path of the sub-region, a cleaning path for the target region is generated; The step of segmenting the region to be cleaned to generate at least one sub-region to be cleaned includes: Obtain the sub-configuration space of the area to be cleaned; The subconfiguration space is sliced ​​and scanned to generate the topology of the columns in the subconfiguration space; Based on the topology, generate the sub-region boundaries of the subconfiguration space; The area to be cleaned is divided according to the boundaries of the sub-regions to generate the sub-regions to be cleaned.

2. The laser cleaning path planning method according to claim 1, characterized in that, The step of generating the sub-region boundary based on the topology includes: Based on the topology, determine the first consecutive valid region of the current column, and determine whether the next column has a second consecutive valid region; When the second consecutive valid region does not exist in the next column, the sub-region boundary is generated based on the first consecutive valid region; When the next column has a second continuous valid region, the second continuous valid region is extracted according to the minimum neighborhood, and the connectivity relationship between the first continuous valid region and the second continuous valid region is determined. Based on the connectivity relationship, the sub-region boundary is generated.

3. The laser cleaning path planning method according to claim 2, characterized in that, The step of extracting the second continuous effective region based on the minimum neighborhood includes: Obtain the first center point of the current column, and the discontinuous valid region of the next column; Obtain the second center point that is closest to the center point in the discontinuous valid region; Combining the minimum neighborhood, scan the invalid boundaries at both ends of the column direction of the second center point; Obtain valid points within the invalid boundary, integrate the valid points, and generate the second continuous valid region.

4. The laser cleaning path planning method according to claim 2, characterized in that, The step of generating the sub-region boundary based on the connectivity relationship includes: When the connectivity relationship is connected, the second continuous valid region of the next column is obtained iteratively until the connectivity relationship is not connected; When the connectivity relationship is not connected, the sub-region boundary is generated based on the first continuous valid region and the second continuous valid region.

5. The laser cleaning path planning method according to claim 1, characterized in that, After generating the sub-region cleaning path of the sub-region to be cleaned using the bow-shaped scanning path method based on the starting point of the cleaning path, and obtaining the ending point of the cleaning path of the sub-region cleaning path, before determining the starting point of the next cleaning path of the sub-region to be cleaned using the optimization algorithm based on the ending point of the cleaning path, the method further includes: Obtain the sequence numbers of all the sub-regions to be cleaned; Based on all the sequence numbers and their corresponding vertices, generate a sub-region information taboo table; The step of determining the starting point of the next cleaning path for the sub-region to be cleaned, based on the endpoint of the cleaning path and using the optimization algorithm, includes: Delete the sequence number and vertex corresponding to the endpoint of the cleaning path from the sub-region information taboo table to generate an unplanned information taboo table; Based on the endpoint of the cleaning path, the optimization algorithm is used to traverse the unplanned information taboo table to determine the starting point of the cleaning path for the next sub-region to be cleaned in the unplanned information taboo table.

6. The laser cleaning path planning method according to any one of claims 1-5, characterized in that, The step of generating the region to be cleaned based on the binarized values ​​of the binarized image includes: Obtain the discrete pixel units of the binarized image; The discrete pixel units of the image are converted into equivalent spot coverage units, and the binarized values ​​are obtained. The area to be cleaned is generated based on the equivalent spot coverage unit, the binarized value, and the formula for obtaining the area to be cleaned. The formula for obtaining the area to be cleaned includes: ； Where A is the area to be cleaned, (x, y) are the horizontal and vertical coordinates of the equivalent spot coverage unit with a binarization value of 1, row is the total number of rows in the binarized image, and col is the total number of columns in the binarized image.

7. A path planning system for laser cleaning, characterized in that, include: The image module is used to acquire a binarized image of the target region; The area to be cleaned module is used to generate the area to be cleaned based on the binarized values ​​of the binarized image; A segmentation module is used to segment the region to be cleaned and generate at least one sub-region to be cleaned. The segmentation of the region to be cleaned and the generation of at least one sub-region to be cleaned includes: obtaining a sub-configuration space of the region to be cleaned; performing a slice scan on the sub-configuration space to generate a topological structure of columns in the sub-configuration space; generating sub-region boundaries of the sub-configuration space based on the topological structure; and segmenting the region to be cleaned based on the sub-region boundaries to generate the sub-region to be cleaned. The vertex module is used to obtain the vertices of the sub-region to be cleaned, extract the sub-region boundary of each sub-region to be cleaned, search the sub-region vertex set of the column reference on the sub-region boundary of each sub-region to be cleaned, and mark all the vertices of each sub-region to be cleaned. The sub-region cleaning path module is used to determine the starting point of the cleaning path for the sub-region to be cleaned based on the vertices using an optimization algorithm, and to generate the sub-region cleaning path using a bow-shaped scanning path method based on the starting point of the cleaning path. This includes: selecting an initial starting point from all the vertices according to the vertices and preset rules, taking the sub-region to be cleaned where the initial starting point is located as the starting sub-region, and determining the initial starting point as the starting point of the cleaning path for the sub-region to be cleaned; generating the sub-region cleaning path for the sub-region to be cleaned using the bow-shaped scanning path method based on the starting point of the cleaning path, and obtaining the ending point of the cleaning path; determining the starting point of the cleaning path for the next sub-region to be cleaned using the optimization algorithm based on the ending point of the cleaning path, and generating the sub-region cleaning path and ending point of the next sub-region to be cleaned using the bow-shaped scanning path method; and iteratively obtaining... The process involves obtaining the cleaning path and endpoint of the sub-region, continuing until the cleaning path of the last sub-region to be cleaned is obtained. Specifically, generating the cleaning path of the sub-region to be cleaned using the arc-shaped scanning path method based on the starting point of the cleaning path includes: obtaining the position of the starting point of the cleaning path; matching the position with a preset rule base; recording scanning path points frame by frame during dynamic path planning based on the path planning direction; if the scanning reaches the valid row boundary of the column where the current scanning point is located, then after recording the current path point, reversing the row scanning direction and iteratively adding the column directions according to the current step inertia direction; when the last column is scanned vertically, the path of the sub-region to be cleaned is completely covered, yielding the path trajectory and scanning endpoint of the sub-region to be cleaned; and generating the sub-region cleaning path based on the path planning direction and the arc-shaped scanning path method. The cleaning path module is used to generate a cleaning path for the target area based on the cleaning path for the sub-area.

Images