A high-precision multi-angle workpiece production detection system based on simulation vision

Abstract

The present invention discloses a high-precision multi-angle workpiece production detection system based on simulation vision, which relates to the field of simulation and is used to solve the problem that when a workpiece is disassembled or assembled, there is a deviation from the lighting characteristics in the real scene during the fusion of multi-angle image data, thereby reducing the detection accuracy. By collecting the lighting characteristic information and material characteristic information in the simulation environment and the real scene, a data analysis model is established, and the deviation evaluation coefficient is compared with the deviation threshold to obtain a comparison result. The deviation evaluation coefficient less than the deviation threshold is defined as a minor deviation, and the deviation evaluation coefficient greater than or equal to the deviation threshold is defined as each deviation angle, and a set of fuzzy rules is formulated with the alarm threshold trigger frequency for fuzzy inference to determine the deviation adjustment scheme. The detection frequency is set according to the production operation of the workpiece, and the deviation of each angle can be effectively adjusted, making the detection decision more flexible and efficient.

Description

Technical Field

[0001] The present invention relates to the field of simulation technology, and more specifically, to a high-precision multi-angle workpiece production detection system based on simulation vision. Background Art

[0002] In modern manufacturing, the production detection of workpieces is a key link to ensure product quality and production efficiency. With the development of automation and intelligent technologies, the use of simulation vision can effectively reduce the contact with workpieces during workpiece production detection operations. At the same time, multi-angle detection can collect workpiece data from different perspectives, improving the detection efficiency and accuracy.

[0003] The prior art has the following deficiencies:

[0004] Currently, based on simulation vision, a simulation scene is generated according to real scene data for multi-angle generation detection of workpieces. However, in the modern workpiece production process, operations such as disassembling or assembling workpieces are often required. And the simulation vision technology is always constructed based on the light source and material data information of the real scene, resulting in a deviation from the lighting characteristics in the real scene during the fusion of multi-angle image data, thereby reducing the detection accuracy. Therefore, a high-precision multi-angle workpiece production detection system based on simulation vision is proposed.

[0005] The above information disclosed in the background art section is only used to enhance the understanding of the background of the present disclosure, and thus it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a high-precision multi-angle workpiece production detection system based on simulation vision, which solves the problems raised in the above background art by using different product inspection methods.

[0007] To achieve the above object, the present invention provides the following technical solution: A high-precision multi-angle workpiece production detection system based on simulation vision, including a data acquisition module, a data processing module, a deviation analysis module, and a deviation adjustment module; the modules are signal-connected to each other;

[0008] The data acquisition module is used to collect the lighting characteristic information and material characteristic information in the simulation environment and the real scene, obtain the lighting reflectance difference, bidirectional reflectance distribution difference, background light intensity difference, and lighting uniform distribution difference through data processing, and send the lighting characteristic information and material characteristic information in the simulation environment and the real scene to the data processing module;

[0009] The data processing module is used to obtain the lighting characteristic information and material characteristic information in the simulation environment and the real scene, establish a data analysis model, obtain a deviation evaluation coefficient, and send it to the deviation analysis module;

[0010] The deviation analysis module is used to obtain the deviation evaluation coefficients from multiple angles, compare them with the deviation threshold, obtain a comparison result, define the deviation evaluation coefficients less than the deviation threshold as minor deviations, define the deviation evaluation coefficients greater than or equal to the deviation threshold as the respective deviation angles, send them to the deviation adjustment module, and perform statistical calculations on the maximum deviation evaluation coefficient and the deviation evaluation coefficients greater than or equal to the deviation threshold around it, and compare with the alarm threshold to evaluate and obtain the maximum deviation angle;

[0011] The deviation adjustment module is used to obtain the respective deviation angles, obtain the respective deviation values, collect the alarm threshold trigger frequency, and use fuzzy logic to determine the deviation adjustment scheme for the respective deviation values and the alarm threshold trigger frequency.

[0012] In a preferred embodiment, the material characteristic information includes the difference in light reflectance and the difference in bidirectional reflectance distribution; the lighting characteristic information includes the difference in background light intensity and the difference in uniform light distribution;

[0013] The difference in light reflectance Rd is obtained by measuring the light flux reflected by the material properties and light source conditions set in the real scene workpiece surface and the simulation environment and the light flux incident on the workpiece surface, respectively performing ratio calculations, and then performing difference calculations. i where i is the i-th angle;

[0014] The difference in bidirectional reflectance distribution Bd is obtained by substituting the bidirectional reflectance distribution values in the real scene and the bidirectional reflectance distribution values in the simulation environment into the root mean square error calculation. i ;

[0015] The difference in background light intensity Bl is obtained by using a light sensor and setting virtual light sources and scene parameters to measure the overall ambient light intensity of the simulation environment and the real scene, and performing a difference calculation on them. i ;

[0016] The difference in uniform light distribution Ud is obtained by taking the absolute difference between the ratio of the difference between the maximum and minimum light intensities in the real scene and the simulation environment and their corresponding average light intensities. i .

[0017] In a preferred embodiment, obtaining the lighting characteristic information and material characteristic information in the simulation environment and the real scene includes the difference in light reflectance Rd i , the difference in bidirectional reflectance distribution Bd i , the difference in background light intensity Bl i and the difference in uniform light distribution Udi , establish a data analysis model to generate a deviation evaluation coefficient De i , and the specific formula is:

[0018]

[0019] In the formula, De i is the deviation evaluation coefficient, and are the preset proportionality coefficients of the light reflection rate difference, bidirectional reflection distribution difference, background light intensity difference, and light illumination uniform distribution difference, and and are all greater than 0.

[0020] In a preferred embodiment, after obtaining the deviation evaluation coefficients from multiple angles, compare and analyze the deviation evaluation coefficients with the continuously iterated deviation thresholds;

[0021] If the deviation evaluation coefficient is greater than or equal to the deviation threshold, mark the light illumination characteristic information obtained at the angle corresponding to the deviation evaluation coefficient as an obvious deviation, and generate an obvious signal;

[0022] If the deviation evaluation coefficient is less than the deviation threshold, mark the light illumination characteristic information obtained at the angle corresponding to the deviation evaluation coefficient as a minor deviation, and generate an end signal.

[0023] In a preferred embodiment, count the deviation evaluation coefficients of the light illumination characteristics from multiple angles that are greater than or equal to the deviation threshold, sort the deviation evaluation coefficients, obtain each deviation angle, and send it to the deviation adjustment module to identify the maximum deviation evaluation coefficient De max , statistically calculate the values of the deviation evaluation coefficients greater than or equal to the deviation threshold around it to obtain the maximum possible deviation;

[0024] Collect the maximum possible deviation and compare it with the alarm threshold. If it is greater than the alarm threshold, issue an alarm and mark the angle of the maximum possible deviation as the maximum deviation angle.

[0025] In a preferred embodiment, obtain each deviation angle, obtain each deviation value, and collect the alarm threshold trigger frequency;

[0026] Obtain each deviation value by calculating the parameter differences between the light reflection rate, bidirectional reflection distribution, background light intensity, and light illumination uniform distribution of each deviation angle and the corresponding real - world scene angle.

[0027] In a preferred embodiment, by statistically calculating the number of times of light illumination characteristic deviation detection between the simulation environment and the real - world scene and the number of times exceeding the alarm threshold at different detection angles and calculating the ratio, obtain the alarm threshold trigger frequency.

[0028] In a preferred embodiment, each deviation value and the alarm threshold trigger frequency are defined as input variables and are respectively divided into different fuzzy sets;

[0029] The deviation adjustment scheme is defined as an output variable and is divided into a fuzzy set;

[0030] Fuzzy rules are formulated to describe the influence of each deviation value and the alarm threshold trigger frequency on the deviation adjustment scheme;

[0031] Fuzzy reasoning is performed according to the fuzzy rules to determine the deviation adjustment scheme.

[0032] Technical effects and advantages of the present invention:

[0033] 1. By collecting the illumination characteristic information and material characteristic information in the simulation environment and the real scenario, the present invention establishes a data analysis model, obtains the comparison between the deviation evaluation coefficient and the deviation threshold, obtains the comparison result, defines the deviation evaluation coefficient less than the deviation threshold as a minor deviation, and defines the deviation evaluation coefficient greater than or equal to the deviation threshold as each deviation angle. The maximum deviation evaluation coefficient and the deviation evaluation coefficients greater than or equal to the deviation threshold around it are collected for statistical calculation and compared with the alarm threshold to evaluate and obtain the maximum deviation angle, simplifying the traditional simulation environment that changes in real time according to the real scenario, setting the detection frequency according to the production operation of the workpiece, giving an early warning and accurately finding the maximum deviation angle, and saving the detection time.

[0034] 2. Based on each deviation angle, the present invention obtains each deviation value, and formulates a set of fuzzy rules with the alarm threshold trigger frequency for fuzzy reasoning to determine the deviation adjustment scheme, improving the automation of the system, reducing human intervention, enabling the deviations at each angle to be effectively adjusted, further optimizing the existing model, and making the detection decision more flexible and efficient. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 It is a schematic diagram of the modules of a high-precision multi-angle workpiece production detection system based on simulation vision according to the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0036] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0037] The present invention conducts multi-angle detection of workpiece production through a simulation scenario generated based on simulation vision, extracts the three-dimensional model of the workpiece through three-dimensional scanning technology, including the geometric shape and surface material information of the workpiece, then sets multi-angle virtual light sources consistent with the direction, color, and intensity of the light source in the real scenario, simulates the real lighting conditions of the workpiece production environment, and after rendering using the GGX model, captures images of the workpiece from multiple angles, obtains the lighting characteristics of the workpiece based on the images, and generates a detection report to ensure the quality of the workpiece.

[0038] Among them, the surface lighting characteristics are extracted from the fused image and compared with the surface lighting characteristics obtained in the real scenario to analyze the deviation of the lighting characteristics to verify the accuracy of the detection report;

[0039] Embodiment 1

[0040] The present invention discloses a high-precision multi-angle workpiece production detection system based on simulation vision, as Figure 1 shown, including a data acquisition module, a data processing module, a deviation analysis module, and a deviation adjustment module; the modules are signal-connected to each other.

[0041] The data acquisition module is used to collect the lighting characteristic information and material feature information in the simulation environment and the real scenario, obtain the lighting reflectance difference, bidirectional reflection distribution difference, background light intensity difference, and lighting uniform distribution difference through data processing, and send the lighting characteristic information and material feature information in the simulation environment and the real scenario to the data processing module.

[0042] Among them, the material feature information includes the lighting reflectance difference and the bidirectional reflection distribution difference; the lighting characteristic information includes the background light intensity difference and the lighting uniform distribution difference.

[0043] It should be noted that the lighting reflectance difference can reflect the light reflection ability of the workpiece surface, including specular reflection and diffuse reflection. Among them, there are processes of grinding, assembling, and disassembling during the workpiece production and processing. Therefore, the acquisition module collects according to the number of operations during the workpiece production and processing to evaluate the deviation between the lighting characteristics in the simulation environment and the real scenario in real time;

[0044] The lighting reflectance difference refers to the difference in the light reflection ability of the workpiece surface between the simulation environment and the real scenario. Specifically, the reflectance represents the proportion of the surface material reflecting the incident light;

[0045] Its acquisition logic is to use a photometer or illuminometer to measure the light flux reflected by the workpiece surface in the real scenario and the light flux incident on the workpiece surface and perform a ratio calculation to obtain the reflectance Rr in the real scenario i, according to the material properties and light source conditions set in the simulation environment, the light flux reflected from the workpiece surface and the light flux of the workpiece surface are used to obtain the reflectivity Rs in the simulation environment i , based on the formula: Rd i = |Rs i -Rr i | to obtain the difference in light reflection rate Rd i ; where i is the i-th angle;

[0046] The bidirectional reflection distribution difference refers to the difference in the distribution characteristics of the light incident and reflected on the workpiece surface between the simulation environment and the real scene. Specifically, the bidirectional reflection distribution (BREF) describes the intensity distribution of the light reflected at different angles after entering the surface from one incident direction; the corresponding bidirectional reflection distribution is calculated for the real scene and the simulation environment through an optical measurement sensor and GGX model rendering using a functional formula; the specific formula is as follows:

[0047]

[0048] In the formula, ω i is the incident direction of light, ω o is the reflection direction, dL o (ω o ) is the radiance reflected in the direction ω o , dE i (ω i ) is the irradiance incident from the direction ω i ;

[0049] Substitute the BRDF values in the real scene and the BRDF values in the simulation environment into the root mean square error calculation, and its formula expression is:

[0050]

[0051] In the formula, Bd i is the bidirectional reflection distribution difference, y i is the BRDF value in the i-th real scene, is the BRDF value in the i-th simulation environment;

[0052] The background light intensity difference refers to the diffused light from all directions in the simulation environment and the real scene, which does not depend on the direct illumination of a specific light source (such as ambient light, reflected light, etc.); its acquisition logic is to use a light sensor and set virtual light source and scene parameters, measure the overall ambient light intensity of the simulation environment and the real scene, and perform a difference calculation on them to obtain the background light intensity difference Bl i ;

[0053] Among them, if the difference in background light intensity is greater, the deviation of the corresponding light characteristics is greater, resulting in the inconsistency between the performance of the workpiece in the simulation environment and the real scene, and reducing the accuracy of the detection report;

[0054] The difference in uniform light distribution refers to the difference in the areas where light is evenly distributed or unevenly bright and dark on the entire surface of the workpiece in the simulation environment and the real scene; it can be understood that in the real scene, due to factors such as the position, quantity of light sources and the geometric characteristics of the scene, the light is not necessarily evenly distributed, that is, the distribution is random, while in the simulation environment, the light uniformity is determined by the configuration of virtual light sources, light models and rendering techniques, that is, the distribution is controllable. Although the distribution is controlled by adjusting the light source parameters, there are still slight deviations from the light distribution in the real scene during the complex process of workpiece production;

[0055] Its acquisition logic is to obtain the difference in uniform light distribution Ud through the absolute difference between the ratio of the difference between the maximum and minimum light intensities in the real scene and the simulation environment and their corresponding average light intensities i ;

[0056] It should be noted that the above light characteristic information and material characteristic information in the simulation environment and the real scene are all obtained from multi-angle images taken by multi-angle cameras set in the real scene and multi-angle virtual cameras set in the virtual environment corresponding to the position of the real scene; among them, the specific image detection methods (such as edge detection, texture analysis, drawing color and brightness histograms, etc.) are specifically implemented by the experimenters according to the specific actual situation and are not limited here;

[0057] The data processing module is used to obtain the light characteristic information and material characteristic information in the simulation environment and the real scene, establish a data analysis model, obtain the deviation evaluation coefficient and send it to the deviation analysis module.

[0058] Obtain the light characteristic information and material characteristic information in the simulation environment and the real scene, including the difference in light reflectance Rd i , the difference in bidirectional reflection distribution Bd i , the difference in background light intensity Bl i and the difference in uniform light distribution Ud i , establish a data analysis model, generate a deviation evaluation coefficient De i , and the specific formula is:

[0059]

[0060] In the formula, De i is the deviation evaluation coefficient, and is the preset proportional coefficient of the difference in illumination reflectance, the difference in bidirectional reflectance distribution, the difference in background light intensity, and the difference in illumination uniformity distribution, and as well as Both are greater than 0.

[0061] Among them, the difference in illumination reflectance, the difference in bidirectional reflection distribution, the difference in background light intensity, and the difference in illumination uniformity distribution all directly express the differences in illumination characteristics between the current angle simulation environment and the real scene.

[0062] Specifically, from the above formula, it can be seen that the greater the difference in illumination reflectance, bidirectional reflection distribution, background light intensity and illumination uniform distribution, the greater the deviation evaluation coefficient, and the greater the difference between the illumination characteristics in the image captured at a single angle and the material characteristics it reflects, the lower the degree of consistency between the built-in data of the simulation environment as the real scene changes. Conversely, the smaller the deviation evaluation coefficient.

[0063] The deviation analysis module is used to obtain the deviation assessment coefficients of multiple angles and compare them with the deviation threshold to obtain the comparison results. The deviation assessment coefficients less than the deviation threshold are defined as slight deviations, and the deviation assessment coefficients greater than or equal to the deviation threshold are defined as individual deviation angles and sent to the deviation adjustment module. The maximum deviation assessment coefficient and the deviation assessment coefficients around it that are greater than or equal to the deviation threshold are statistically calculated and compared with the alarm threshold to evaluate the maximum deviation angle.

[0064] The logic for obtaining the deviation threshold is to collect a set of distributed samples from the historical deviation evaluation information library, divide the data set into a training set and a test set, set the evaluation indicators and clustering algorithm, and in each round of cross-validation, train the model on the training set and evaluate the model performance on the test set. Then, adjust the deviation threshold based on the performance of the validation set. Therefore, the deviation threshold is constantly iteratively updated.

[0065] In the present invention, clustering algorithm is a type of unsupervised learning algorithm, which is used to divide the data points in the data set into groups or clusters with similarities; a common one is K-means clustering, which divides the data points in the data set into K clusters, so that the distance between the data point of each curve and the center point (centroid) of the cluster to which it belongs is minimized, and finally the effect of the adjusted deviation threshold is measured by the Euclidean distance, so as to set the deviation threshold.

[0066] After obtaining the deviation assessment coefficients from multiple angles, the deviation assessment coefficients are compared and analyzed with the continuously iterated deviation thresholds.

[0067] If the deviation evaluation coefficient is greater than or equal to the deviation threshold, the illumination characteristic information acquired at the angle corresponding to the deviation evaluation coefficient is marked as an obvious deviation, and an obvious signal is generated.

[0068] If the deviation evaluation coefficient is less than the deviation threshold, the illumination characteristic information obtained at the angle corresponding to the deviation evaluation coefficient is marked as a minor deviation, and an end signal is generated.

[0069] Statistically calculate the deviation evaluation coefficients of the multi-angle illumination characteristics greater than or equal to the deviation threshold, sort the deviation evaluation coefficients, obtain each deviation angle, and send them to the deviation adjustment module to identify the maximum deviation evaluation coefficient De max Statistically calculate the values of the deviation evaluation coefficients greater than or equal to the deviation threshold around it to obtain the maximum possible deviation;

[0070] Among them, the maximum possible deviation means that when the deviation evaluation coefficient of the illumination characteristic at a certain angle is the maximum deviation value in this detection, the values of the deviation evaluation coefficients greater than or equal to the deviation threshold around it are statistically calculated; for example, nine virtual cameras corresponding to the simulation environment and the real scene are set. It is detected that the deviation evaluation coefficient calculated by comprehensively calculating the image data captured by the virtual camera at the 30-degree position on the left side of the workpiece and the image data captured by the camera corresponding to the real scene is the largest. Then, the deviation evaluation coefficients greater than or equal to the deviation threshold at multiple angles such as 60 degrees to the left and 90 degrees to the left around it are brought into the statistical calculation to evaluate the maximum deviation of this deviation; among them, the experimenter can set different ranges according to the actual situation to finely adjust the value of the maximum possible deviation, which is convenient for better accurate adjustment according to different information.

[0071] Collect the maximum possible deviation and compare it with the alarm threshold. If it is greater than the alarm threshold, an alarm is issued and the angle of the maximum possible deviation is marked as the maximum deviation angle.

[0072] Among them, the maximum deviation angle refers to the angle of the illumination characteristic with the largest deviation from the real scene in the simulation environment, and the alarm threshold is obtained through the current deviation evaluation coefficient and the historical deviation evaluation information database.

[0073] The present invention simplifies the traditional simulation environment that changes in real time according to the real scene by collecting the illumination characteristic information and material characteristic information in the simulation environment and the real scene, establishing a data analysis model, obtaining the comparison result by comparing the deviation evaluation coefficient with the deviation threshold, defining the deviation evaluation coefficient less than the deviation threshold as a minor deviation, defining the deviation evaluation coefficient greater than or equal to the deviation threshold as each deviation angle, collecting the maximum deviation evaluation coefficient and the deviation evaluation coefficients greater than or equal to the deviation threshold around it for statistical calculation, and comparing it with the alarm threshold to evaluate and obtain the maximum deviation angle, sets the detection frequency according to the production operation of the workpiece, gives early warnings and accurately discovers the maximum deviation angle, saving detection time.

[0074] Embodiment 2

[0075] In Embodiment 1 of the present invention, it is mainly illustrated by way of example that through the illumination characteristic information and material characteristic information in the simulation environment and the real scenario, a data analysis model is established, and the deviation evaluation coefficient is compared with the deviation threshold to obtain a comparison result. The deviation evaluation coefficient less than the deviation threshold is defined as a minor deviation. The maximum deviation evaluation coefficient and the deviation evaluation coefficients greater than or equal to the deviation threshold around it are collected for statistical calculation and compared with the alarm threshold to evaluate the operation strategy of the maximum deviation angle; however, in Embodiment 1, only starting from finding the maximum deviation angle, how to adjust the deviation is not considered. Obviously, although manual adjustment can be performed after the alarm is issued to reduce the deviation, there is a possibility that multiple workpieces are polished, assembled or disassembled simultaneously in the workpiece production environment, which will lead to insufficient system automation and excessive consumption of manpower; for the above problems, Embodiment 2 of the present invention is further refined;

[0076] The deviation adjustment module is used to obtain each deviation angle, obtain each deviation value and collect the alarm threshold trigger frequency, and use fuzzy logic for each deviation value and the alarm threshold trigger frequency to determine the deviation adjustment scheme;

[0077] Among them, each deviation value refers to the specific deviation value of each parameter in the illumination characteristics of the image data obtained from each deviation angle and the illumination characteristics of the image data obtained from the corresponding real scenario. Specifically, each parameter has been exemplified in Embodiment 1 and will not be elaborated here.

[0078] Its acquisition logic is to calculate the parameter differences of the light reflectivity, bidirectional reflection distribution, background light intensity, and light uniformity distribution of each deviation angle and the corresponding real scenario angle. Specifically, if it is a positive difference, it means that the parameter is set too low in the simulation environment; conversely, if it is a negative difference, it means that the parameter is set too high in the simulation environment;

[0079] The alarm threshold trigger frequency refers to the frequency of triggering an alarm when the system monitors that the illumination characteristic difference between the simulation environment and the real scenario reaches or exceeds the preset alarm threshold during the workpiece detection process;

[0080] Its acquisition logic is to calculate the ratio by counting the number of times of detecting the illumination characteristic deviation between the simulation environment and the real scenario and exceeding the alarm threshold at different detection angles to obtain the alarm threshold trigger frequency;

[0081] Among them, the value of the alarm threshold trigger frequency is further limited. It can be understood that if the alarm threshold trigger frequency is continuously counted, there will be a situation of data redundancy. Therefore, the experimenter can limit the acquisition duration of the alarm threshold trigger frequency based on information such as the adjustment frequency of this angle to reduce the disadvantages of excessive data or too high recent trigger frequency, resulting in too many detection times to detect the recent abnormality of this angle in time;

[0082] Determine the deviation adjustment plan using fuzzy logic based on each deviation value and the alarm threshold trigger frequency;

[0083] For example, "High", "Low", "Medium" for each deviation value, and "Many", "Few", "Moderate" for the alarm threshold trigger frequency;

[0084] Formulate a set of fuzzy rules to describe the influence of different input variables on the output variable. The definition of the rules can be based on professional knowledge or obtained through data analysis and experiments. For example:

[0085] Mark each deviation value as X, the alarm threshold trigger frequency as U, and the deviation adjustment plan as C_Public;

[0086] Then it can be defined as:

[0087] Rule 1: IF (X is High) AND (U is Many) THEN (C_Public is High)

[0088] Rule 2: IF (U is Low) AND (U is Few) THEN (C_Public is Low) ...

[0090] Conduct fuzzy reasoning according to the fuzzy rules to determine the deviation adjustment plan;

[0091] It should be noted that the division of the fuzzy set can be adjusted according to the actual situation. For example, although three fuzzy sets are used as an example in this embodiment, in fact, each deviation value and the alarm threshold trigger frequency can be divided into more than three sets to facilitate more accurate adjustment according to different public information.

[0092] Furthermore, for the judgment of high, medium, and low of each deviation value and the alarm threshold trigger frequency, the threshold can be set according to the actual situation for judgment. For example, when each deviation value exceeds 68% of the real - world scenario value, it is labeled as "High", and when the alarm threshold trigger frequency is higher than 75%, it is labeled as "Many", etc., which will not be elaborated here.

[0093] Among them, the specific steps of GGX model rendering are to calculate the microscopic normal distribution of the surface through the GGX micro-surface distribution function (NDF) to describe how light scatters on rough or shiny surfaces. The GGX model allows light to be reflected in different directions by defining a distribution of microscopic normals. Combining with the Fresnel reflection model, the reflection intensity of the incident light is calculated based on the refractive index of the material and the incident angle, thereby accurately simulating light. Finally, the Smith G geometry occlusion-shadow model is adopted. By determining the microscopic normal distribution, calculating the Fresnel reflection effect, and combining with geometry occlusion, a visual effect consistent with the actual workpiece surface lighting conditions is finally generated.

[0094] The present invention obtains respective deviation values through respective deviation angles, and formulates a set of fuzzy rules with the alarm threshold triggering frequency for fuzzy inference to determine the deviation adjustment scheme, improving the automation of the system, reducing human intervention, enabling the deviations at each angle to be effectively adjusted, further optimizing the existing model, and making the detection decision more flexible and efficient.

[0095] The above formulas are all dimensionless and take their numerical values for calculation. The formula is a formula obtained by collecting a large amount of data for software simulation to be closest to the real situation. The preset parameters in the formula are set by those skilled in the art according to the actual situation.

[0096] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center in a wired or wireless (such as infrared, wireless, microwave, etc.) manner. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0097] It should be understood that in various embodiments of the present application, the sequence numbers of the above processes do not imply the order of execution, and the order of execution of each process should be determined by its function and internal logic, and should not constitute any limitation to the implementation process of the embodiments of the present application.

[0098] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0099] Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working processes of the systems, devices, and units described above can refer to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0100] In several embodiments provided by the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling or direct coupling or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the devices or units can be in an electrical, mechanical, or other form.

[0101] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they can be located in one place, or distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0102] In addition, the functional units in various embodiments of the present application can be integrated into one processing unit, or each unit can exist physically alone, or two or more units can be integrated into one unit.

[0103] When the above-mentioned functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of this technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application. The aforementioned storage medium includes: various media that can store program codes, such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs.

[0104] As described above, the above are only specific implementation manners of this application, but the protection scope of this application is not limited thereto. Any person skilled in the art within the technical scope disclosed by this application can easily think of changes or substitutions, which should all be covered within the protection scope of this application. Therefore, the protection scope of this application should be subject to the protection scope of the claims.

Claims

1. A high-precision multi-angle workpiece production inspection system based on simulated vision, characterized by: It includes data acquisition module, data processing module, deviation analysis module and deviation adjustment module; signal connection between each module; The data acquisition module is used to collect the illumination characteristic information and material characteristic information in the simulation environment and the real scene, obtain the illumination reflectivity difference, the bidirectional reflection distribution difference, the background light intensity difference and the illumination uniform distribution difference through data processing, and send the illumination characteristic information and material characteristic information in the simulation environment and the real scene to the data processing module; The data processing module is used to obtain the illumination characteristic information and material characteristic information in the simulation environment and the real scene, establish a data analysis model, obtain the deviation evaluation coefficient and send it to the deviation analysis module; The deviation analysis module is used to obtain the deviation evaluation coefficients of multiple angles, and compare them with the deviation threshold to obtain the comparison results. The deviation evaluation coefficients less than the deviation threshold are defined as slight deviations, and the deviation evaluation coefficients greater than or equal to the deviation threshold are defined as individual deviation angles and sent to the deviation adjustment module. The maximum deviation evaluation coefficient and the surrounding deviation evaluation coefficients greater than or equal to the deviation threshold are statistically calculated and compared with the alarm threshold to evaluate the maximum deviation angle. The deviation adjustment module is used to obtain each deviation angle, obtain each deviation value and collect the alarm threshold trigger frequency, and use fuzzy logic to determine the deviation adjustment plan for each deviation value and alarm threshold trigger frequency; Material feature information includes the difference in illumination reflectance and the difference in bidirectional reflection distribution; illumination feature information includes the difference in background light intensity and the difference in illumination uniformity distribution; The light reflectance difference Rd is obtained by measuring the reflected luminous flux of the workpiece surface in the real scene and the material properties and light source conditions set in the simulation environment, as well as the luminous flux incident on the workpiece surface, and calculating the ratio and difference respectively. i ; Where i is the i-th angle; The bidirectional reflection distribution value in the real scene and the bidirectional reflection distribution value in the simulation environment are brought into the root mean square error calculation to obtain the bidirectional reflection distribution difference Bd i ; Use the light sensor and set the virtual light source and scene parameters to measure the overall ambient light intensity of the simulated environment and the real scene, and calculate the difference between them to obtain the background light intensity difference Bl i ; The illumination uniform distribution difference Ud is obtained by the absolute difference between the difference between the maximum and minimum illumination intensities in the real scene and the simulation environment and the ratio of the corresponding average illumination intensity. i .

2. The high-precision multi-angle workpiece production inspection system based on simulated vision according to claim 1, characterized in that: Obtain lighting characteristic information and material feature information in the simulation environment and real scene, including the lighting reflectivity difference Rd i , Bidirectional reflection distribution difference Bd i 、Background light intensity difference Bl i And the uniform distribution difference of illumination Ud i , establish a data analysis model and generate the deviation evaluation coefficient De i , the specific formula is: In the formula, De i is the deviation assessment coefficient, as well as is the preset proportional coefficient of the difference in illumination reflectance, the difference in bidirectional reflectance distribution, the difference in background light intensity, and the difference in illumination uniformity distribution, and as well as Both are greater than 0.

3. The high-precision multi-angle workpiece production inspection system based on simulated vision according to claim 1, characterized in that: After obtaining the deviation assessment coefficients from multiple angles, the deviation assessment coefficients are compared and analyzed with the continuously iterated deviation thresholds; If the deviation evaluation coefficient is greater than or equal to the deviation threshold, the illumination characteristic information obtained at the angle corresponding to the deviation evaluation coefficient is marked as an obvious deviation, and an obvious signal is generated; If the deviation evaluation coefficient is less than the deviation threshold, the illumination characteristic information acquired at the angle corresponding to the deviation evaluation coefficient is marked as a slight deviation, and an end signal is generated.

4. The high-precision multi-angle workpiece production inspection system based on simulated vision according to claim 3 is characterized by: The deviation evaluation coefficients of multi-angle illumination characteristics that are greater than or equal to the deviation threshold are counted, and the deviation evaluation coefficients are sorted to obtain each deviation angle and send them to the deviation adjustment module to identify the maximum deviation evaluation coefficient De max , perform statistical calculations on the values ​​of the deviation assessment coefficients around it that are greater than or equal to the deviation threshold to obtain the maximum possible deviation; The maximum possible deviation is collected and compared with the alarm threshold. If it is greater than the alarm threshold, an alarm is issued and the angle of the maximum possible deviation is marked as the maximum deviation angle.

5. The high-precision multi-angle workpiece production inspection system based on simulated vision according to claim 1, characterized in that: Obtain each deviation angle, obtain each deviation value and collect the alarm threshold trigger frequency; Each deviation value is obtained by calculating the light reflectivity, bidirectional reflection distribution, background light intensity, and uniform light distribution at each deviation angle and the parameter difference between them and the corresponding real scene angle.

6. The high-precision multi-angle workpiece production inspection system based on simulated vision according to claim 5, characterized in that: By counting the number of times the illumination characteristics of the simulation environment and the real scene deviate from each other and exceed the alarm threshold at different detection angles and performing ratio calculation, the alarm threshold triggering frequency is obtained.

7. The high-precision multi-angle workpiece production inspection system based on simulated vision according to claim 5, characterized in that: Each deviation value and alarm threshold trigger frequency are defined as input variables and divided into different fuzzy sets respectively; The deviation adjustment scheme is defined as the output variable and divided into fuzzy sets; Formulate fuzzy rules to describe the impact of each deviation value and alarm threshold triggering frequency on the deviation adjustment plan; Perform fuzzy reasoning based on fuzzy rules and determine the deviation adjustment plan.

Images