CN111299815B - Visual detection and laser cutting trajectory planning method for low-gray rubber pad - Google Patents

Visual detection and laser cutting trajectory planning method for low-gray rubber pad Download PDF

Info

Publication number
CN111299815B
CN111299815B CN202010091306.8A CN202010091306A CN111299815B CN 111299815 B CN111299815 B CN 111299815B CN 202010091306 A CN202010091306 A CN 202010091306A CN 111299815 B CN111299815 B CN 111299815B
Authority
CN
China
Prior art keywords
image
rubber pad
function
node
arc
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN202010091306.8A
Other languages
Chinese (zh)
Other versions
CN111299815A (en
Inventor
要义勇
王世超
辜林风
高射
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Xian Jiaotong University
Original Assignee
Xian Jiaotong University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Xian Jiaotong University filed Critical Xian Jiaotong University
Priority to CN202010091306.8A priority Critical patent/CN111299815B/en
Publication of CN111299815A publication Critical patent/CN111299815A/en
Application granted granted Critical
Publication of CN111299815B publication Critical patent/CN111299815B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K26/00—Working by laser beam, e.g. welding, cutting or boring
    • B23K26/02—Positioning or observing the workpiece, e.g. with respect to the point of impact; Aligning, aiming or focusing the laser beam
    • B23K26/03—Observing, e.g. monitoring, the workpiece
    • B23K26/032—Observing, e.g. monitoring, the workpiece using optical means
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K26/00—Working by laser beam, e.g. welding, cutting or boring
    • B23K26/36—Removing material
    • B23K26/38—Removing material by boring or cutting
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1656—Programme controls characterised by programming, planning systems for manipulators
    • B25J9/1664—Programme controls characterised by programming, planning systems for manipulators characterised by motion, path, trajectory planning
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B25—HAND TOOLS; PORTABLE POWER-DRIVEN TOOLS; MANIPULATORS
    • B25J—MANIPULATORS; CHAMBERS PROVIDED WITH MANIPULATION DEVICES
    • B25J9/00—Programme-controlled manipulators
    • B25J9/16—Programme controls
    • B25J9/1694—Programme controls characterised by use of sensors other than normal servo-feedback from position, speed or acceleration sensors, perception control, multi-sensor controlled systems, sensor fusion
    • B25J9/1697—Vision controlled systems
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B23—MACHINE TOOLS; METAL-WORKING NOT OTHERWISE PROVIDED FOR
    • B23K—SOLDERING OR UNSOLDERING; WELDING; CLADDING OR PLATING BY SOLDERING OR WELDING; CUTTING BY APPLYING HEAT LOCALLY, e.g. FLAME CUTTING; WORKING BY LASER BEAM
    • B23K2103/00—Materials to be soldered, welded or cut
    • B23K2103/30—Organic material

Abstract

The invention discloses a method for visual detection and laser cutting trajectory planning of a low-gray rubber pad. The method comprises the steps of image acquisition, global image positioning, local image contour extraction, geometric contour fitting, laser cutting track planning and the like. Through the steps, the method for automatically identifying and processing the workpiece facing the low-gray-scale image is constructed, and the production efficiency is improved.

Description

Visual detection and laser cutting trajectory planning method for low-gray rubber pad
[ technical field ] A method for producing a semiconductor device
The invention belongs to the field of intelligent manufacturing, and relates to a method for visual detection and laser cutting trajectory planning of a low-gray rubber pad.
[ background of the invention ]
The rubber shock pad is a key part for stable operation of the high-speed rail. The rubber pad is made by injection molding process, and features that it is a planar object with square outline and distributed circular lugs and strip grooves. Because the rubber pad produces the overlap because of the injection moulding technology requirement, consequently need carry out secondary operation to the rubber pad and cut off the overlap.
At present, aiming at the problem of cutting off the flash of the rubber pad, a manual cutting off method is basically adopted, the labor cost is high, and the processing performance is poor. The rubber pad is difficult to clamp and fix due to the flexible characteristic. Therefore, it is necessary to construct an intelligent manufacturing method with automatic positioning and non-contact processing. The rubber pad flash is cut by adopting machine vision detection outline characteristics and laser cutting, and the blank of automatic flash cutting of the rubber pad is filled, so that the rubber pad flash cutting method is very necessary.
Machine vision technology is widely applied in recent years, and machine vision means that a computer displays, identifies and cognizes an objective world scene by utilizing a computer information technology and an image acquisition technology. The machine detection technology is a new generation intelligent mapping technology which utilizes a machine vision technology to detect and measure workpieces in an industrial field. In the rubber pad course of working, the necessary appearance of visual detection feature identification is: the rubber pad processing process needs real-time positioning, a reference point position and azimuth angle visual method is researched and determined, and the outline position is determined according to the reference point position, so that the visual detection technology is very necessary.
The laser cutting characteristics are analyzed according to the profile characteristics of visual detection, and a method for constructing a laser cutting track is also very necessary. The laser cutting method is based on the principle of shortest path and the characteristic of multi-dimensional deformation of the rubber pad, and provides a track optimization method of laser cutting according to the principle of high efficiency, flexibility and accuracy, and aims at the cutting characteristic of the rubber pad, and obtains a laser cutting track.
[ summary of the invention ]
The invention aims to solve the problems in the prior art and provides a method for visual detection and laser cutting track planning of a low-gray rubber pad.
In order to achieve the purpose, the invention adopts the following technical scheme to realize the purpose:
a visual detection and laser cutting trajectory planning method for a low-gray rubber pad comprises the following steps:
step 1, carrying out image acquisition on the whole and local parts of a rubber pad;
step 2, realizing coarse positioning of the rubber pad on a station through global image positioning; by analyzing the structural characteristics of the rubber pad, the geometric characteristic positioning is realized by utilizing two ellipses of the rubber pad;
step 3, obtaining segmentation points of the edge of the contour by using an image denoising and image segmentation method, and realizing accurate positioning of the rubber pad flash contour;
step 4, according to the contour extraction result, performing contour fitting on the geometric image to be processed, and performing arc and straight line fitting on the segmentation point set;
and 5, planning the laser cutting track according to the visual detection and geometric fitting results, and planning the cutting track by utilizing a genetic ant colony algorithm according to the shortest path principle.
Compared with the prior art, the invention has the following beneficial effects:
1. the invention relates to a rubber shock pad, which is a technical object directly oriented to and aims to solve the problem of rubber pad flash cutting by applying an automatic technology. The machine vision technology is utilized to carry out vision detection on the flash of the rubber pad to obtain the contour to be cut, then the essence of the cutting process is analyzed, the trajectory planning is completed, and the problem in the automatic cutting process of the rubber pad is solved.
2. The invention applies the machine vision technology to detect the rubber product, the whole idea is firstly whole, then local, firstly rough and then fine, and the idea has inspiration effect on other precise detection. And the integral positioning result is used as a reference, and a numerical control system is used for carrying out high-precision local image acquisition and processing on the key part through the key part on the marking component to obtain a high-precision detection result.
3. Aiming at the problem of image processing with texture noise, the invention provides a thought of combining Gabor transformation and fuzzy level set direction, and effectively solves the problem of image feature extraction with low contrast and high texture noise. Due to the limitation of materials and processing techniques in the machining industry, the texture noise of a cutter often exists on the surface of a processed part, and the visual feature extraction idea provided by the method is not limited to the application of rubber products.
4. The invention explains the track planning problem in the field of machining, summarizes a TSP model of the track planning, and optimizes the process of solving the TSP by utilizing the thought of combining a genetic algorithm and an ant colony algorithm. The invention discusses the mathematical essence of the trajectory planning from the mathematical perspective, compares the existing methods for solving the trajectory planning, utilizes the advantage of global search of the genetic algorithm in the early stage and utilizes the advantage of fast convergence of the ant colony algorithm in the later stage to complete the task of the trajectory planning.
[ description of the drawings ]
FIG. 1 is an overall flow chart of the present invention;
FIG. 2 is a diagram illustrating a camera acquiring a global image according to an embodiment of the present invention;
FIG. 3 is a partial image captured by a camera according to an embodiment of the present invention;
FIG. 4 is a wavelet enhanced image of an embodiment of the present invention;
FIG. 5 is a schematic diagram of the principle of connecting the midpoints of the parallel chords of an ellipse and passing through the center according to an embodiment of the invention;
FIG. 6 is a schematic diagram of a plurality of groups of center connecting lines for finding the center of an ellipse
FIG. 7 is a schematic diagram of clustering of multiple intersections
FIG. 8 is a global image detection result according to an embodiment of the present invention;
FIG. 9 is a local keypoint map of an embodiment of the present invention;
FIG. 10 is a partially acquired image of an embodiment of the invention;
FIG. 11 is a flow chart of local contour extraction according to an embodiment of the present invention;
FIG. 12 is a partial image extraction result graph, in which the left side is the original image of the partial image collection, and the right side is the extraction result graph according to the embodiment of the present invention
FIG. 13 is a graph of the result of fitting the geometric profile according to the embodiment of the present invention, wherein the left side is a graph of the result of fitting the straight line, and the right side is a graph of the result of fitting the circular arc
Fig. 14 is a flowchart of trajectory planning according to an embodiment of the present invention.
[ detailed description ] embodiments
In order to make the technical solutions of the present invention better understood, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, not all of the embodiments, and are not intended to limit the scope of the present disclosure. Moreover, in the following description, descriptions of well-known structures and techniques are omitted so as to not unnecessarily obscure the concepts of the present disclosure. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
Various structural schematics according to the disclosed embodiments of the invention are shown in the drawings. The figures are not drawn to scale, wherein certain details are exaggerated and possibly omitted for clarity of presentation. The shapes of various regions, layers and their relative sizes and positional relationships shown in the drawings are merely exemplary, and deviations may occur in practice due to manufacturing tolerances or technical limitations, and a person skilled in the art may additionally design regions/layers having different shapes, sizes, relative positions, according to actual needs.
In the context of the present disclosure, when a layer/element is referred to as being "on" another layer/element, it can be directly on the other layer/element or intervening layers/elements may be present. In addition, if a layer/element is "on" another layer/element in one orientation, then that layer/element may be "under" the other layer/element when the orientation is reversed.
It should be noted that the terms "first," "second," and the like in the description and claims of the present invention and in the drawings described above are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It is to be understood that the data so used is interchangeable under appropriate circumstances such that the embodiments of the invention described herein are capable of operation in sequences other than those illustrated or described herein. Furthermore, the terms "comprises," "comprising," and "having," and any variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
The invention is described in further detail below with reference to the accompanying drawings:
referring to fig. 1, the method for visual inspection and laser cutting trajectory planning of the low-gray rubber mat comprises the following steps:
step 1, collecting an image; in the embodiment, a CCD industrial camera is adopted, a wide-angle camera is adopted for collecting the global image of the rubber pad, the wide-angle camera is vertically arranged at a position 0.5-1m above the rubber pad, the camera is subjected to plane calibration, a white fluorescent lamp is adopted for uniformly distributing light, and the collected image is shown in a figure 2; and aiming at local image acquisition, according to a global image recognition result, a two-axis motion platform is utilized to realize the motion of a local acquisition camera, the local camera macro-camera is vertically arranged at a position 2cm above the rubber pad, and an acquired image is shown in figure 3.
Step 2, enhancing the image by utilizing wavelet transformation; the image is decomposed by using the "sym" wavelet, the wavelet coefficient is reconstructed according to the threshold quantization, and the reconstructed wavelet is aggregated to obtain an enhanced image, which is shown in fig. 4. The filtering adopts a Gaussian filtering method to remove noise and smooth the image, on a two-dimensional plane, the Gaussian filtering utilizes Gaussian kernel to weight and average the image, and the Gaussian kernel formula is as follows:
Figure BDA0002383814490000061
the Canny operator is used for edge detection, and the basic steps are as follows: performing gradient amplitude image and angle image calculation gradient amplitude image on the smoothed image to perform non-maximum inhibition; edges are detected and connected using dual threshold processing and connection analysis.
The purpose of detecting the concavity and convexity of the edge detection result is to determine whether or not the arc can form an ellipse. Ellipse center estimation takes advantage of the properties of the ellipse: the midpoints of a set of parallel chords necessarily cross the center of the ellipse. And combining the 3 sections of arcs in different quadrants pairwise to obtain the central point of the ellipse, and if the distance between the central points is less than a threshold value, determining that the three sections of arcs possibly belong to the same ellipse. The geometric schematic diagram is shown in fig. 5, fig. 6 and fig. 7, and the formula is obtained as follows:
Figure BDA0002383814490000062
the geometric parameters are calculated by an ellipse geometric formula, the clustering result is calculated to obtain an ellipse detection result, and the size and the roundness of the result are constrained, and the identification result is shown in figure 8.
Step 3, according to the global positioning result, image acquisition is carried out on key positions of the rubber pad, the distribution of the image acquisition is shown in fig. 9, and the acquisition result is shown in fig. 10; the local contour extraction flow shown in fig. 11 is performed on the acquisition result.
First, since analyzing the original image reveals that there is severe texture noise and interference with image segmentation, the image is texture-suppressed using Gabor filtering. The main idea of the Gabor filtering method is as follows: different textures generally have different central frequencies and bandwidths, a group of Gabor filters can be designed according to the frequencies and the bandwidths to filter texture images, each Gabor filter only allows textures corresponding to the frequency of the Gabor filter to pass through smoothly, energy of other textures is restrained, and texture features are analyzed and extracted from output results of the filters and used for subsequent classification or segmentation tasks. The specific formula of the two-dimensional Gabor function is expressed according to the following formula:
Figure BDA0002383814490000071
the main process comprises the following steps: building a Gabor filter bank: selecting 6 scales and 4 directions, thus forming 24 Gabor filters; the second step is that: the Gabor filter bank is convolved with each image block in a space domain, and each image block can obtain 24 filter outputs; the third step: gaussian low-pass filtering compensates for local variations, smoothing the Gabor amplitude information. The fourth step: a mapping of spatial location information is added that allows the classifier to prefer packets that are spatially close together. The fifth step: and (4) performing dimensionality reduction on the data by a Principal Component (PCA) method to obtain Gabor filtering data.
Secondly, FCM clustering is carried out on the filtering data, wherein the FCM clustering is a fuzzy clustering algorithm based on an objective function and is mainly used for clustering analysis of the data. Mature theory and wide application range, is oneExcellent clustering algorithm. Based on this, assuming that the data set is X, if the data are divided into C classes, then there are C class centers C, and each sample j belongs to a certain class i with a degree of membership uijThen, an FCM objective function and its constraints are defined as follows:
Figure BDA0002383814490000072
expression of the attribute function:
Figure BDA0002383814490000081
clustering center formula:
Figure BDA0002383814490000082
and evaluating the clustering result, selecting the clustering result with the minimum spatial frequency to perform level set segmentation, mainly comprising driving force calculation, image evolution curved surface, and determining iteration times to complete segmentation. An edge-based geometric active contour model can be obtained by the level set, as shown in equation 7. Image evolution is realized according to the model, and segmentation results are obtained, and the result pair is shown in fig. 12.
Figure BDA0002383814490000083
And 4, fitting the geometric contour of the straight line and the circular arc according to the local segmentation result, wherein the straight line fitting adopts linear least square, and the equation of the fitted straight line is assumed as follows:
y＝ax+b (8)
then for the data points segmented from two feature points, the least squares distance:
Figure BDA0002383814490000084
and (5) performing arc fitting by adopting an error estimation method. Known circular arc point row (x)1,y1)…(xi,yi)…(xn,yn) Let its center coordinate be (x)c,yc) And the radius is R, the radius of the point row is as follows:
Figure BDA0002383814490000085
the error definition is shown in the formula.
Figure BDA0002383814490000091
Since the function must take a minimum value, the above equation is R, xc、ycMust be equal to 0 and the arc parameters can be obtained. The results of fitting the straight line and the circular arc are obtained by the above two methods, as shown in fig. 13.
Step 5, obtaining a cutting track through local contour integration; and when the machining process is started, the laser cutting head rapidly moves from the origin of the machine tool to the next contour to be cut, one of all vertexes on the contour is used as a cutting starting and stopping point, the ring is machined along the contour track, the ring is moved to the next ring to be cut, and the process is repeated until all rings contained in the workpiece are cut. Through the analysis of the machining process, the method can be uniformly abstracted into the same mathematical problem, namely how to search a series of points on a plurality of closed contours, and then connect the points in a certain sequence within the range of the machining principle, so that the moving path of the cutter between the contours is shorter, the productivity is greatly improved, and the machining cost is saved.
Firstly, determining characteristic points of a cutting contour, selecting the middle point of the circular arc section, and determining the characteristics of the end points of the straight line end. And selecting the characteristic points as contour punching points by adopting a nearest neighbor principle.
The purpose of the trajectory planning is to minimize the idle stroke, and the X, Y axes can be moved at the highest speed allowed by the idle stroke of the laser cutting head. Assuming the programmed origin of the laser cut is, then the mathematical model of the X, Y-axis path optimization problem can be expressed as:
Figure BDA0002383814490000092
therefore, the laser cutting trajectory planning problem translates into the classical combinatorial optimization problem, the traveler TSP problem. A better solution can be obtained by analyzing and adopting a genetic algorithm and an ant colony algorithm to solve the path optimization problem. The flow chart is shown in fig. 14, and the track planning result is obtained through the results of the genetic algorithm and the ant colony algorithm.
The group algorithm carries out high-precision search and realizes a new intelligent heuristic algorithm.
When solving the TSP problem by using a genetic algorithm, the relevant steps of the genetic algorithm need to be combined with the actual problem, which is specifically as follows:
fitness function and coding problem: and combining the TSP problem, adopting decimal real number coding, taking the traversal order of points as coding, and taking the reciprocal of the distance function as an optimized fitness function.
Initial population and chromosome selection: and generating an initial population by using a random function, and selecting and accepting or rejecting a parent chromosome according to the value of the fitness function.
And (3) a crossover operator: and (3) a sequential crossing method is adopted, common double-point crossing is completed firstly, and then the itinerant route is modified.
Mutation operator: an inverse variation method is adopted. And the reversion operator has single direction, the variation result of the improved fitness function after reversion is retained, otherwise, the variation result is discarded.
And after the genetic algorithm is completed, recording the optimization result, and performing initial weighted distribution of pheromones on the relevant paths. The ant colony algorithm has the characteristics of distributed parallel computation and strong feedback, and the solving of the TSP problem by using the ant colony algorithm generally comprises the following steps:
1) at the initial moment, the pheromone on the path is initialized according to the result of the genetic algorithm, and meanwhile, coordinate information and a distance matrix are initialized.
2) Constraints for trajectory selection are set.
3) And (4) path selection. According to the ant colony algorithm principle, ant colony transfer is mainly affected by two conditions: the pheromone concentration and the length of the corresponding path are comprehensively considered, and the method uses
Figure BDA0002383814490000101
Representing the probability of ant k transitioning from node i to node j at time t, is formulated as follows:
Figure BDA0002383814490000111
4) pheromone update function. The pheromone update on path (i, j) at time (t + n) is shown by:
Figure BDA0002383814490000112
5)ηij(t) a distance update function, related to the length of the path (i, j), expressed as follows:
Figure BDA0002383814490000113
the above-mentioned contents are only for illustrating the technical idea of the present invention, and the protection scope of the present invention is not limited thereby, and any modification made on the basis of the technical idea of the present invention falls within the protection scope of the claims of the present invention.

Claims (5)

1. A visual detection and laser cutting trajectory planning method for a low-gray rubber pad is characterized by comprising the following steps of:
step 1, carrying out image acquisition on the whole and local parts of a rubber pad;
step 2, realizing coarse positioning of the rubber pad on a station through global image positioning; by analyzing the structural characteristics of the rubber pad, the geometric characteristic positioning is realized by utilizing two ellipses of the rubber pad; the specific method comprises the following steps:
step 2-1, enhancing the image by utilizing wavelet transformation; decomposing the image by adopting 'sym' wavelet, reconstructing according to the wavelet coefficient of threshold quantization, and polymerizing the reconstructed wavelet to obtain an enhanced image;
step 2-2, preprocessing the enhanced image by adopting Gaussian filtering to realize fuzzy processing and noise reduction to obtain a smooth image, weighting and averaging the image by utilizing Gaussian filtering on a two-dimensional plane, wherein the Gaussian filtering has the following formula:
Figure FDA0002821594390000011
wherein σ represents a dispersion parameter of the gaussian kernel function, x represents an abscissa in the image, y represents an ordinate in the image, and e represents a natural logarithm;
step 2-3, edge detection is carried out on the preprocessed image by using a Canny operator, and the specific method comprises the following steps: performing gradient amplitude calculation and angle amplitude calculation on the smoothed image, and calculating an image result of the gradient amplitude calculation by applying a non-maximum suppression algorithm; detecting and connecting edges with dual-threshold processing;
step 2-4, carrying out convex-concave detection on the edge detection result, and screening the edge according to the concave-convex property of the function and the arc constraint;
2-5, carrying out ellipse parameter estimation by using the screened arc sections, carrying out cross checking calculation to obtain the central coordinate of an ellipse, calculating ellipse fitting quality, and setting a threshold value for screening; using three arcs of different quadrants, combining two arcs to calculate the central point of the ellipse, if the distance between the central points is less than the threshold, considering the three arcs to belong to the same ellipse, and calculating the formula as follows:
Figure FDA0002821594390000021
wherein C represents the center of the target ellipse,
Figure FDA0002821594390000022
representing the center mass point of a first set of parallel chords on an arc,
Figure FDA0002821594390000023
representing the central mass point of another set of parallel chords on the arc, t1Representing the slope of the line connecting the midpoints of a first set of parallel chords on the arc, t2Representing the slope of the line connecting the midpoints of another set of parallel chords on the arc line;
step 2-6, clustering the results by utilizing the calculated circle center parameters, major axes and inclination parameters to obtain ellipse detection results, and constraining the size and roundness of the results to obtain actual coordinates of two positioning ellipses which accord with the structural characteristics of the actual rubber pad so as to complete global positioning;
step 3, obtaining segmentation points of the edge of the contour by using an image denoising and image segmentation method, and realizing accurate positioning of the rubber pad flash contour;
step 4, according to the contour extraction result, performing contour fitting on the geometric image to be processed, and performing arc and straight line fitting on the segmentation point set;
and 5, planning the laser cutting track according to the visual detection and geometric fitting results, and planning the cutting track by utilizing a genetic ant colony algorithm according to the shortest path principle.
2. The method for visual inspection and laser cutting trajectory planning for low-gray rubber mats according to claim 1, wherein a CCD industrial camera is used in step 1 to collect images of the rubber mats globally; the CCD industrial camera is vertically arranged at a position 0.5-1m above the rubber pad, the CD industrial camera is subjected to plane calibration, a white fluorescent lamp is adopted for uniform light distribution, and the global image of the rubber pad is collected;
aiming at adopting a macro camera, locally acquiring an image of the rubber pad; the macro camera is vertically arranged 2cm above the rubber pad; and according to the identification result of the acquired global image, the motion of the macro camera is realized by utilizing the two-axis motion platform.
3. The method for visual inspection and laser cutting trajectory planning for low-grayscale rubber mats according to claim 1, wherein the specific method in step 3 is as follows:
step 3-1, denoising an image;
filtering the image by adopting Gabor filters, wherein each Gabor filter only allows the texture corresponding to the frequency of the Gabor filter to pass smoothly, and the energy of other textures is suppressed, and analyzing and extracting texture features from the output result of each filter for later image segmentation; the specific formula of the two-dimensional Gabor function is as follows:
Figure FDA0002821594390000031
Figure FDA0002821594390000032
wherein λ represents the wavelength of the sine and cosine function, θ represents the angle of the kernel function,
Figure FDA0002821594390000033
the phase angle of a sine and cosine function is represented, x 'represents an abscissa after coordinate conversion, and y' represents an ordinate after coordinate conversion;
the method comprises the following specific steps: the first step is as follows: building a Gabor filter bank: selecting 6 scales and 4 directions, thus forming 24 Gabor filters; the second step is that: the Gabor filter bank is convolved with each image block in a space domain, and each image block obtains 24 filter outputs; the third step: gaussian low-pass filtering compensates local change, and Gabor amplitude information is smoothed; the fourth step: adding a map of spatial location information that allows the classifier to prefer groupings that are spatially close together; the fifth step: performing dimensionality reduction on the data by a Principal Component (PCA) method to obtain Gabor filtering data;
step 3-2, performing FCM clustering on the filtering data;
assuming that the data set is X, the data are divided into C classes, the corresponding C class centers are C, and the membership degree of each sample j belonging to a certain class i is uijThen, an FCM objective function and its constraints are defined as follows:
Figure FDA0002821594390000034
wherein J represents the numerical value of the objective function, c represents the clustering center, n represents the total number of data of a certain class, i represents the sequence of the class, J represents the sequence of single data in a certain class, m represents the power exponent of the membership function, and xjRepresenting the jth individual data, ciRepresenting the ith cluster center;
attribute function uijExpression (c):
Figure FDA0002821594390000041
wherein k denotes the sequence of cluster centers, ckRepresenting the kth cluster center;
clustering center formula:
Figure FDA0002821594390000042
step 3-3, image segmentation;
evaluating the clustering results, and selecting the clustering result with the minimum spatial frequency to perform level set segmentation; calculating a driving force, evolving a curved surface of an image, determining iteration times and finishing segmentation; obtaining a geometric active contour model based on edges through a level set; realizing image evolution according to the geometric active contour model to obtain a segmentation result:
Figure FDA0002821594390000043
wherein phi represents a level set function, u represents a constant coefficient of the level set evolution internal energy, div represents a divergence operation, and dpRepresenting the sign of the differentiation of the energy density function p,
Figure FDA0002821594390000044
representing the gradient of the level set function, deltaτAnd expressing a regularized Dikela function, g expressing an edge detection function, alpha expressing a constant coefficient of the area evolution energy of the level set, and tau expressing a constant threshold of the regularization parameter.
4. The method for visual inspection and laser cutting trajectory planning for low-grayscale rubber mats according to claim 1, wherein the geometric image contour fitting in step 4 comprises fitting of geometric contours of straight lines and circular arcs, and the specific method is as follows:
linear least squares are used for linear fitting, assuming that the fitted linear equation is:
y＝ax+b (8)
then for the data points segmented from two feature points, the least squares distance:
Figure FDA0002821594390000051
wherein i represents a sequence of data points;
performing arc fitting by adopting an error estimation method; known circular arc point row (x)1,y1)…(xi,yi)…(xn,yn) Let its center coordinate be (x)c,yc) And the radius is R, the radius of the point row is as follows:
Figure FDA0002821594390000052
the error definition is shown in the formula;
Figure FDA0002821594390000053
since the function takes a minimum value, the above equation is R, xc、ycThe partial derivative of (d) is equal to 0, the arc parameter can be obtained.
5. The method for low-grayscale rubber pad visual inspection and laser cutting trajectory planning as claimed in claim 1, wherein the specific method of step 5 is as follows:
firstly, determining characteristic points of a cutting contour, selecting a middle point of an arc segment, and determining characteristics of end points of a straight line end; selecting characteristic points as contour punching points by adopting a nearest principle;
let the laser cut have a programmed origin of (X)1,Y1) Then the mathematical model of the X, Y axis path optimization problem is expressed as:
Figure FDA0002821594390000061
converting the laser cutting track planning problem into a TSP problem; solving the path optimization problem by analyzing and adopting a genetic algorithm and an ant colony algorithm to obtain a better solution; obtaining a track planning result through results of a genetic algorithm and an ant colony algorithm; the specific method comprises the following steps:
step 5-1, initializing pheromones on the path at the initial moment according to the result of the genetic algorithm, and simultaneously initializing coordinate information and a distance matrix;
step 5-2, setting constraint conditions for track selection;
step 5-3, selecting a path; according to the ant colony algorithm principle, ant colony transfer is mainly affected by two conditions: the pheromone concentration and the length of the corresponding path are comprehensively considered, and the method uses
Figure FDA0002821594390000062
Denotes the ant k slave node at time tThe probability of the point i transitioning to the node j is formulated as follows:
Figure FDA0002821594390000063
wherein the content of the first and second substances,
Figure FDA0002821594390000064
representing the probability of ant k transitioning from node i to node j at time t,
Figure FDA0002821594390000065
represents the pheromone content from node i to node j to the power of alpha,
Figure FDA0002821594390000066
represents the visibility parameter from node i to node j to the power of beta,
Figure FDA0002821594390000067
represents the pheromone content from node i to node u to the power alpha during the summation,
Figure FDA0002821594390000068
representing the visibility parameter from the node i to the node u in the summation process to the power of beta, wherein beta represents the power exponent value of the visibility parameter and represents the importance degree of the visibility parameter;
step 5-4, updating a function by the pheromone; the pheromone update on path (i, j) at time (t + n) is shown by:
Figure FDA0002821594390000071
where ρ represents the pheromone loss coefficient, Δ τijDenotes the pheromone increment, Q denotes the total pheromone distribution, and is constant, LkIndicating a pheromone distribution diffusion coefficient, related to distance;
in the step 5-5, the step of the method,ηij(t) a distance update function, related to the length of the path (i, j), expressed as follows:
Figure FDA0002821594390000072
wherein d isijIndicating the distance from the ith node to the jth node.
CN202010091306.8A 2020-02-13 2020-02-13 Visual detection and laser cutting trajectory planning method for low-gray rubber pad Active CN111299815B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010091306.8A CN111299815B (en) 2020-02-13 2020-02-13 Visual detection and laser cutting trajectory planning method for low-gray rubber pad

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010091306.8A CN111299815B (en) 2020-02-13 2020-02-13 Visual detection and laser cutting trajectory planning method for low-gray rubber pad

Publications (2)

Publication Number Publication Date
CN111299815A CN111299815A (en) 2020-06-19
CN111299815B true CN111299815B (en) 2021-02-09

Family

ID=71154545

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010091306.8A Active CN111299815B (en) 2020-02-13 2020-02-13 Visual detection and laser cutting trajectory planning method for low-gray rubber pad

Country Status (1)

Country Link
CN (1) CN111299815B (en)

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112634203A (en) * 2020-12-02 2021-04-09 富泰华精密电子（郑州）有限公司 Image detection method, electronic device and computer-readable storage medium
CN113182701A (en) * 2021-03-29 2021-07-30 大族激光科技产业集团股份有限公司 Laser processing method, apparatus, device and storage medium
CN113847906A (en) * 2021-09-10 2021-12-28 交通运输部天津水运工程科学研究所 Laboratory space wave real-time measurement method based on visual edge detection
CN113857698B (en) * 2021-11-04 2023-12-22 安徽省长凌智能装备有限公司 Pulse laser drilling positioning calibration method
CN113953685B (en) * 2021-11-22 2023-09-01 重庆工商大学 Planning method for laser cutting path of planar plate
CN114239958B (en) * 2021-12-15 2022-11-11 广东工业大学 Laser cutting method and system for special-shaped part based on machine vision
CN114799565B (en) * 2022-05-18 2024-04-19 苏州仁和老河口汽车股份有限公司 Intelligent control system and method for laser cutting machine for aluminum casting machining
CN115138527B (en) * 2022-06-22 2023-12-26 深圳市双翌光电科技有限公司 Rapid processing path generation method through visual guidance
CN115170561A (en) * 2022-09-01 2022-10-11 济南鼎点数控设备有限公司 Roundness detection method and system of laser cutting workpiece based on visual processing
CN115246050B (en) * 2022-09-23 2022-12-20 江苏跃格智能装备有限公司 Laser cutting path optimization control method based on machine vision
CN116174941B (en) * 2023-04-26 2023-09-08 弗斯迈智能科技(江苏)有限公司 Laser cutting perovskite film cutting positioning method
CN116303479B (en) * 2023-05-22 2023-08-22 浙大宁波理工学院 Real-time compression method for track data
CN116573218B (en) * 2023-07-07 2023-09-19 深圳市键键通科技有限公司 Auxiliary equipment for watch film lamination
CN117483838B (en) * 2023-12-29 2024-03-12 唐山惠达智能厨卫科技有限公司 Board drilling method and device based on artificial intelligence
CN117655563A (en) * 2024-01-31 2024-03-08 成都沃特塞恩电子技术有限公司 Laser cutting path planning method and device, electronic equipment and storage medium

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101561875A (en) * 2008-07-17 2009-10-21 清华大学 Method for positioning two-dimensional face images
CN101776882A (en) * 2010-01-19 2010-07-14 东莞市大族粤铭激光科技有限公司 Method for generating laser cutting program
CN203636208U (en) * 2013-10-18 2014-06-11 昆山思拓机器有限公司 Processing device for conductive rubber pad
CN105187310A (en) * 2015-09-22 2015-12-23 北京大学深圳研究生院 Path optimization method and system
CN105527965A (en) * 2016-01-04 2016-04-27 江苏理工学院 Route planning method and system based on genetic ant colony algorithm
CN106056648A (en) * 2016-06-14 2016-10-26 深圳市智能机器人研究院 Intelligent robot image drawing method and system
CN107876970A (en) * 2017-12-13 2018-04-06 浙江工业大学 A kind of robot multi-pass welding welding seam three-dimensional values and weld seam inflection point identification method
CN107943045A (en) * 2017-12-08 2018-04-20 江苏商贸职业学院 A kind of method for planning path for mobile robot based on ant colony genetic fusion algorithm
CN110153567A (en) * 2019-04-04 2019-08-23 江苏大学 A kind of laser cutting system based on image recognition

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101561875A (en) * 2008-07-17 2009-10-21 清华大学 Method for positioning two-dimensional face images
CN101776882A (en) * 2010-01-19 2010-07-14 东莞市大族粤铭激光科技有限公司 Method for generating laser cutting program
CN203636208U (en) * 2013-10-18 2014-06-11 昆山思拓机器有限公司 Processing device for conductive rubber pad
CN105187310A (en) * 2015-09-22 2015-12-23 北京大学深圳研究生院 Path optimization method and system
CN105527965A (en) * 2016-01-04 2016-04-27 江苏理工学院 Route planning method and system based on genetic ant colony algorithm
CN106056648A (en) * 2016-06-14 2016-10-26 深圳市智能机器人研究院 Intelligent robot image drawing method and system
CN107943045A (en) * 2017-12-08 2018-04-20 江苏商贸职业学院 A kind of method for planning path for mobile robot based on ant colony genetic fusion algorithm
CN107876970A (en) * 2017-12-13 2018-04-06 浙江工业大学 A kind of robot multi-pass welding welding seam three-dimensional values and weld seam inflection point identification method
CN110153567A (en) * 2019-04-04 2019-08-23 江苏大学 A kind of laser cutting system based on image recognition

Also Published As

Publication number Publication date
CN111299815A (en) 2020-06-19

Similar Documents

Publication Publication Date Title
CN111299815B (en) Visual detection and laser cutting trajectory planning method for low-gray rubber pad
CN109541997B (en) Spraying robot rapid intelligent programming method for plane/approximate plane workpiece
CN111192284B (en) Vehicle-mounted laser point cloud segmentation method and system
CN110322453B (en) 3D point cloud semantic segmentation method based on position attention and auxiliary network
US7995055B1 (en) Classifying objects in a scene
CN107369161A (en) A kind of workpiece point cloud segmentation method at random based on the European cluster of improvement
CN111402227B (en) Bridge crack detection method
CN106373118A (en) A complex curved surface part point cloud reduction method capable of effectively keeping boundary and local features
CN110807781B (en) Point cloud simplifying method for retaining details and boundary characteristics
CN113628263A (en) Point cloud registration method based on local curvature and neighbor characteristics thereof
CN114926699A (en) Indoor three-dimensional point cloud semantic classification method, device, medium and terminal
CN113393426A (en) Method for detecting surface defects of rolled steel plate
CN114663373A (en) Point cloud registration method and device for detecting surface quality of part
CN113450269A (en) Point cloud key point extraction method based on 3D vision
CN110569926B (en) Point cloud classification method based on local edge feature enhancement
CN114359876B (en) Vehicle target identification method and storage medium
CN116258857A (en) Outdoor tree-oriented laser point cloud segmentation and extraction method
CN114898118A (en) Automatic statistical method and system for power transmission line house removal amount based on multi-source point cloud
CN110458938B (en) Real-time three-dimensional reconstruction method and system for bulk material pile
CN112200248A (en) Point cloud semantic segmentation method, system and storage medium under urban road environment based on DBSCAN clustering
CN117292181A (en) Sheet metal part hole group classification and full-size measurement method based on 3D point cloud processing
CN113316080B (en) Indoor positioning method based on Wi-Fi and image fusion fingerprint
Boudani et al. Content-Based Image Retrieval for Surface Defects of Hot Rolled Steel Strip Using Wavelet-Based LBP
Yuan et al. Automatic cables segmentation from a substation device based on 3D point cloud
Xu et al. Identification of street trees’ main nonphotosynthetic components from mobile laser scanning data

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant