Precision feeding methods and related equipment based on industrial miniature camera vision perception

Abstract

This application relates to the field of industrial automation control technology, and in particular to a precise material placement method and related equipment using industrial miniature camera visual perception. The method includes: a robot grasping material and moving it above a fixture; simultaneously acquiring real-time images of the material's penetration features and the fixture's positioning markers using an industrial miniature camera; identifying the center position and angle of the two images based on a visual algorithm, and calculating the deviation value; adjusting the robot's trajectory and posture in real time according to the deviation value to form a closed-loop control; repeating the above steps until the deviation value is less than a preset threshold, thus completing the precise placement of the material. This application utilizes the characteristics of industrial miniature cameras—small size, large depth of field, easy sealing, and integrated light source—to solve the problems of traditional imaging systems being difficult to deploy in the confined space of machine tools and susceptible to corrosion, significantly improving material placement accuracy and reliability. Simultaneously, the use of visual perception technology reduces manual teaching costs, adapting to the high precision and flexibility requirements of automated production lines.

Description

Technical Field

[0001] This application relates to the field of industrial automation control technology, and in particular to a precise material feeding method and related equipment based on industrial micro-camera visual perception. Background Technology

[0002] Machine vision technology, by simulating the working principle of the human visual system, enables computers to perceive, understand, and process image information. This allows for the rapid and accurate identification and positioning of workpieces in industrial settings, providing robots with precise grasping and placement instructions. By monitoring the position and status of workpieces in real time, as well as their positional relationships with other equipment, the robot's trajectory and speed can be adjusted to achieve efficient loading and unloading operations. Simultaneously, machine vision technology can optimize the production process, improving the overall efficiency of the production line. By introducing advanced technologies and equipment such as machine vision and industrial embodied intelligent robots, enterprises can achieve automated and intelligent upgrades to their production lines.

[0003] Traditional machine vision-based material handling methods primarily involve identifying features of the primary fixture, comparing them with those shown during the initial teaching phase, calculating the visual correction amount, and then having the robot place the material into the machine tool based on this correction. This method not only requires photographic teaching of each machine tool, which is time-consuming, but it is also an open-loop control approach with poor accuracy. This can potentially scratch the material during the material handling process or even cause material to fall off.

[0004] In traditional vision-based material handling applications, cameras with wider fields of view are used to improve the accuracy of camera positioning and material handling, often resulting in a large imaging system (a combination of camera, lens, and light source). However, the space for unloading on machine tools is limited due to the presence of cutting tools and cutter heads, making it difficult to accommodate a large robot end effector. Furthermore, cutting fluid and shavings generated during machining severely corrode camera chips and lens elements, and the large size of the imaging system also presents sealing challenges. This creates a difficult balance between the design of the robot end effector and the need for a high-precision imaging system.

[0005] Therefore, based on the above problems, the existing technology still needs to be improved in terms of the accuracy and stability of material feeding. Summary of the Invention

[0006] The purpose of this application is to provide a precise material placement method and related equipment using industrial miniature camera vision perception, aiming to solve the problems of traditional imaging systems being difficult to deploy and susceptible to corrosion, as well as the high cost of manual teaching in traditional vision-guided material placement in high-precision material placement scenarios in confined spaces such as machine tools.

[0007] The purpose of this application is to provide a precise material feeding method based on industrial miniature camera vision perception, including:

[0008] The robot grasps material with at least two through-features and moves it above a machine tool fixture with at least two positioning marks;

[0009] Real-time images of material penetration features and fixture positioning marks are simultaneously acquired using an industrial miniature camera installed on the robot's end effector.

[0010] Based on visual algorithms, the center position and angle of material penetration features and the center position and angle of fixture positioning marks are identified;

[0011] Calculate the positional and angular deviations between the material penetration feature and the fixture positioning mark;

[0012] The robot's trajectory and posture are adjusted in real time based on the deviation value to form a closed-loop control.

[0013] Repeat the steps of image acquisition, visual recognition, deviation calculation and trajectory adjustment until the position deviation value and angle deviation value are less than the preset threshold, and complete the precise placement of materials.

[0014] By adopting the above technical solution, and utilizing the characteristics of industrial miniature cameras—small size, large depth of field, easy sealing, and integrated light source—the problem of traditional imaging systems being difficult to deploy in the confined space of machine tools and susceptible to corrosion from cutting fluid and flying debris is solved. This enables the robot end effector to stably acquire real-time images of material penetration features and fixture positioning marks in complex environments. By synchronously identifying the position and angle information of the material and fixture and forming closed-loop control, the traditional open-loop control mode, which relies on repeated teaching and correction, is changed. During the material feeding process, the robot trajectory is dynamically adjusted according to the real-time deviation, effectively avoiding problems such as material scratches and incomplete feeding caused by the accumulation of deviations. This significantly improves feeding accuracy and reliability, while reducing manual teaching costs and meeting the high precision and high flexibility requirements of automated production lines.

[0015] In one possible implementation of this application, the step of simultaneously acquiring real-time images of the material penetration feature and the fixture positioning mark includes:

[0016] The industrial miniature camera is positioned directly on the penetrating feature of the corresponding fixture positioning mark on the material, so that the material penetrating feature and the fixture positioning mark are imaged in the central area of ​​the real-time image.

[0017] By analyzing the brightness distribution and edge sharpness of the location marker area in the real-time image, it is determined whether the imaging of the location marker meets the visual recognition requirements.

[0018] When the positioning mark is unclear, the relative distance between the industrial miniature camera and the material penetration feature is automatically adjusted by the telescopic device;

[0019] The process continues until the image brightness and edge sharpness of the real-time image meet the preset recognition conditions, or until the maximum extension stroke of the industrial miniature camera is reached.

[0020] By adopting the above technical solution, the industrial miniature camera, by directly facing the penetrating feature, images the material penetrating feature and the fixture positioning mark in the central area of ​​the real-time image, which can effectively reduce image distortion and improve the accuracy of feature recognition. By analyzing the brightness distribution and edge sharpness of the positioning mark area in real time, it can adaptively determine whether the imaging meets the visual recognition requirements and avoid feature extraction failure due to changes in lighting. Combined with the telescopic device, the relative distance between the industrial miniature camera and the material penetrating feature is automatically adjusted to form a dynamic feedback adjustment mechanism, which can fully automatically optimize the imaging quality and ensure that images that meet the recognition requirements can still be stably acquired under complex working conditions such as changes in ambient lighting. This improves the environmental adaptability and automation level of the visual perception feeding process and ensures the reliability of feeding accuracy.

[0021] In one possible implementation of this application, the step of identifying the center position and angle of the material penetration feature and the center position and angle of the fixture positioning mark based on a visual algorithm includes:

[0022] The real-time image is converted from the RGB color space to the HSV color space, and grayscale binarization is performed on the three dimensions of H, S, and V respectively. The region of interest for material penetration features and fixture positioning marks is initially determined by grayscale value filtering.

[0023] Based on the roundness, area, and edge gradient features extracted from the region of interest, interference from machining waste and cutting fluid residue is eliminated, and the target area with material penetration features and fixture positioning marks is locked.

[0024] One-dimensional grayscale projection is performed on the target area to generate grayscale contour lines and Gaussian smoothing is applied. The pixel-level coordinates of the contour points are extracted by the first derivative.

[0025] Within the neighborhood of pixel-level contour points, a weighted average of pixel values ​​is calculated using a bilinear interpolation algorithm to optimize and obtain sub-pixel precision point coordinates.

[0026] Based on the optimized sub-pixel level point coordinates, the contour shape of the material penetration feature and the fixture positioning mark is fitted.

[0027] By adopting the above technical solutions, real-time images are converted to the HSV color space and subjected to multi-dimensional grayscale binarization processing, which can effectively separate color and brightness information and improve the image feature extraction capability under complex lighting conditions. Based on features such as roundness, area, and edge gradient, interference from machining waste and cutting fluid residue can be eliminated, and the target area of ​​material penetration features and fixture positioning marks can be accurately located, reducing misidentification. Through one-dimensional grayscale projection, Gaussian smoothing, and first-order derivative processing, image noise can be suppressed and pixel-level coordinates of contour points can be accurately extracted. Combined with bilinear interpolation algorithm, it can be further optimized to sub-pixel accuracy, significantly improving positioning accuracy. Finally, by contour fitting and principal axis direction analysis to determine the angle deviation, high-precision visual calculation of the position and attitude of material and fixture is achieved, providing reliable deviation data for robot closed-loop control, ensuring that the material penetration features and fixture positioning marks can still be accurately aligned in complex industrial scenarios, effectively improving the anti-interference capability and positioning accuracy of the unloading process.

[0028] In one possible implementation of this application, the method further includes:

[0029] When switching between producing different types of materials, the industrial miniature camera identifies the QR code on the surface of the material and automatically matches the preset positioning model of the material penetration feature with the fixture positioning mark.

[0030] By adopting the above technical solution, when switching between different material models, the system can automatically recognize the QR codes on the surface of the fixture using an industrial miniature camera, quickly matching the material penetration features with the preset positioning model of the fixture positioning marks. This eliminates the need for manual re-teaching, significantly improving the efficiency of multi-product changeover and reducing manual debugging costs. Based on the material model, the system retrieves the corresponding visual template parameters, enabling rapid and accurate adaptation of different fixture positioning benchmarks and ensuring the accuracy of the material release trajectory. This solution, through the synergistic effect of visual recognition, data matching, and intelligent verification, constructs a fully automated material adaptation system for flexible production scenarios, significantly shortening changeover time, improving the production line's responsiveness to multi-product production, and providing key technical support for efficient and flexible mixed-flow production modes in intelligent manufacturing.

[0031] In one possible implementation of this application, the step of determining whether the brightness of the positioning marker meets the visual recognition requirements by analyzing the brightness distribution and edge sharpness of the positioning marker area in a real-time image includes:

[0032] When the brightness and contrast of the location marker area in the real-time image are lower than the preset threshold or the edges are blurred, it is determined that there is surface contamination.

[0033] The robot is controlled to use an air blowing device at its end to clean the surface of the positioning markers and the lens of the miniature camera;

[0034] After cleaning, real-time images are captured again. If the brightness and contrast increases meet the preset standards, the feeding process continues.

[0035] By adopting the above technical solution, the system can automatically determine the surface contamination status of positioning markers based on the brightness and contrast values ​​of real-time images. This triggers a closed-loop processing flow of "contamination detection-cleaning-verification" without manual intervention: when the detected brightness and contrast are below a preset threshold, the robot uses its end-effector's air-blowing device to clean the surface of the positioning markers and the miniature camera lens, effectively removing contaminants such as oil and cutting fluid residue, thus improving the image clarity of the positioning markers. After cleaning, images are collected again to verify the effect, preventing visual recognition failure due to contamination. This solution deeply integrates visual detection and physical cleaning, forming an intelligent preprocessing mechanism that adapts to environmental interference. This significantly improves the reliability of visual perception in complex processing scenarios, reduces the material feeding failure rate caused by surface contamination, reduces manual cleaning and debugging costs, and ensures the continuous and stable operation of the production line.

[0036] The second objective of this application is to provide a precise material feeding system based on industrial miniature camera vision perception, the system comprising:

[0037] Robot control module: The robot grasps material with at least two through-features and moves it above a machine tool fixture with at least two positioning marks;

[0038] Real-time image acquisition module: Through an industrial miniature camera installed on the robot's end effector, it synchronously acquires real-time images of material penetration features and fixture positioning marks;

[0039] Center position and angle recognition module: Based on visual algorithms, it identifies the center position and angle of material penetration features and the center position and angle of fixture positioning marks;

[0040] Deviation calculation module: Calculates the positional and angular deviations between the material penetration feature and the fixture positioning mark;

[0041] Motion trajectory and attitude adjustment module: Adjusts the robot's motion trajectory and attitude in real time based on the deviation value to form closed-loop control;

[0042] Material placement module: Repeatedly execute image acquisition, visual recognition, deviation calculation and trajectory adjustment steps until the position deviation value and angle deviation value are less than the preset threshold, thus completing the precise placement of materials.

[0043] By adopting the above technical solution, and utilizing the characteristics of industrial miniature cameras—small size, large depth of field, easy sealing, and integrated light source—the problems of traditional imaging systems being difficult to deploy in the confined space of machine tools and susceptible to corrosion by cutting fluid are solved. This enables the robot end effector to stably acquire real-time images of material penetration features and fixture positioning marks in complex environments. By synchronously identifying the position and angle information of the material and fixture and forming closed-loop control, the traditional open-loop control mode, which relies solely on one-time teaching and correction, is changed. During the material feeding process, the robot trajectory can be dynamically adjusted according to real-time deviations, effectively avoiding problems such as material scratches and incomplete feeding caused by accumulated deviations. This significantly improves feeding accuracy and reliability, while reducing manual teaching costs and meeting the high precision and high flexibility requirements of automated production lines.

[0044] The third objective of this application is to provide a precise material feeding device based on industrial miniature camera vision perception, the device comprising:

[0045] The memory and processor, wherein the memory stores a computer program that can be loaded by the processor and executed to perform the aforementioned precise material feeding method based on visual perception of an industrial miniature camera.

[0046] The fourth objective of this application is to provide a storage medium.

[0047] The fourth objective of this application is achieved through the following technical solution:

[0048] A storage medium storing a computer program capable of being loaded by a processor and executed by the aforementioned precise material feeding method based on visual perception using an industrial miniature camera.

[0049] In summary, this application includes at least one of the following beneficial technical effects:

[0050] 1. By utilizing the characteristics of industrial miniature cameras—small size, large depth of field, easy sealing, and integrated light source—the problem of traditional imaging systems being difficult to deploy in the confined space of machine tools and susceptible to corrosion from cutting fluid and flying debris is solved. This enables the robot end effector to stably acquire real-time images of material penetration features and fixture positioning marks in complex environments. By simultaneously identifying the position and angle information of the material and fixture during the unloading process and adjusting the robot's motion trajectory and posture in real time to form a closed-loop control, the traditional open-loop control mode that relies solely on teaching and correction is changed. This effectively avoids problems such as material scratches and incomplete unloading caused by accumulated deviations, significantly improving unloading accuracy and reliability, while reducing manual teaching costs and adapting to the high precision and high flexibility requirements of automated production lines.

[0051] 2. The industrial miniature camera, by directly facing the penetrating feature, images the material penetrating feature and the fixture positioning mark in the central area of ​​the real-time image. This effectively reduces image distortion and improves the accuracy of feature recognition. By analyzing the brightness distribution and edge sharpness of the positioning mark area in real time, it can adaptively determine whether the imaging meets the visual recognition requirements, avoiding feature extraction failure due to changes in lighting. Combined with the telescopic device, the relative distance between the industrial miniature camera and the material penetrating feature is automatically adjusted, forming a dynamic feedback adjustment mechanism. This optimizes the imaging quality without manual intervention, ensuring that images meeting recognition requirements can be stably acquired even under complex working conditions such as changes in lighting. This improves the environmental adaptability and automation level of the visual perception feeding process, and ensures the reliability of feeding accuracy. Attached Figure Description

[0052] Figure 1 This is a flowchart illustrating the precise material feeding method using industrial micro-camera visual perception provided in this application embodiment;

[0053] Figure 2 This is a virtual structural diagram of the precision feeding system for industrial micro-camera visual perception provided in the embodiments of this application. Detailed Implementation

[0054] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0055] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0056] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0057] The precise material placement method based on industrial micro-camera vision perception provided in this application embodiment is applicable to material loading scenarios using fixtures with positioning markers, including but not limited to the machining and assembly processes of components for 3C products such as mobile phones, tablets, and laptops. The system achieves high-precision material placement through the collaborative operation of an industrial embodied intelligent robot, an industrial micro-camera, and a machine tool fixture. The end effector of the industrial robot integrates an industrial micro-camera. This compact camera combines imaging and light source functions, can be installed at the robot's end effector, and can be used in confined spaces. Its observation direction is directly aligned with the penetration features on the material, ensuring image acquisition of both the material penetration features and the fixture positioning markers.

[0058] This application provides a precise material feeding method based on industrial micro-camera vision perception, referring to... Figure 1 The main process of the method is described as follows:

[0059] S1: The robot picks up the material with at least two through-features and moves it over the fixture with at least two positioning marks;

[0060] The robot uses its end effector to grasp materials with at least two through-features, such as mobile phone frames or tablet casings, and moves them along a preset path to a fixture. The fixture has at least two positioning marks that form a positioning relationship with the through-features of the material. The robot needs to move to a safe height directly above the fixture to provide a stable operating reference for subsequent visual perception.

[0061] S2: Real-time images of material penetration features and fixture positioning marks are simultaneously acquired by an industrial miniature camera installed on the robot's end effector;

[0062] The industrial miniature camera installed on the robot's end effector begins operation, simultaneously acquiring real-time images of the material penetration feature and the fixture positioning mark. The observation direction of the industrial miniature camera is directly facing the material penetration feature, ensuring that the material penetration feature and the fixture positioning mark are clearly imaged in the picture, providing a foundation for subsequent visual recognition.

[0063] S3: Based on visual algorithms, identify the center position and angle of material penetration features and the center position and angle of fixture positioning marks;

[0064] Based on the acquired real-time images, the system processes and analyzes the images using built-in visual algorithms to identify the center position and angle of the material penetration features and the center position and angle of the fixture positioning marks. This visual algorithm can accurately extract feature information from the images, distinguish the key positioning elements of the material and the fixture, and provide reliable data for deviation calculation.

[0065] S4: Calculate the positional and angular deviations between the material penetration feature and the fixture positioning mark;

[0066] After acquiring the pose information of the material and the fixture, the system further calculates the positional and angular deviation values ​​between the material penetration feature and the fixture positioning mark, quantifies the alignment error between the two, and clarifies the direction and magnitude of the robot adjustment.

[0067] S5: Adjust the robot's trajectory and posture in real time based on the deviation value to form a closed-loop control;

[0068] In this process, the robot adjusts its trajectory and posture in real time based on the calculated deviation value, forming a closed-loop control. The robot continuously fine-tunes itself based on deviation feedback, ensuring that each adjustment is directed towards reducing the deviation, gradually approaching the precisely aligned target.

[0069] S6: Repeat the steps of image acquisition, visual recognition, deviation calculation and trajectory adjustment until the position deviation value and angle deviation value are less than the preset threshold, and complete the precise placement of materials.

[0070] The process involves repeatedly performing image acquisition, visual recognition, deviation calculation, and trajectory adjustment steps until the positional and angular deviation values ​​between the material penetration feature and the fixture positioning mark are less than preset thresholds. At this point, it is determined that the material has reached a precise alignment state, and the robot completes the material placement. The entire process achieves high-precision material placement through dynamic closed-loop control, avoiding the error accumulation problem of traditional open-loop control.

[0071] By combining the visual perception of industrial miniature cameras with the closed-loop control of robots, precise control of the entire process from material grasping to placement is achieved, effectively improving the material placement accuracy and reliability in industrial automation scenarios, and is especially suitable for precision manufacturing fields with stringent positioning accuracy requirements.

[0072] Specifically, in some possible embodiments, the step of simultaneously acquiring real-time images of material penetration features and fixture positioning markers includes:

[0073] The industrial miniature camera is positioned directly on the penetrating feature of the corresponding fixture positioning mark on the material, so that the material penetrating feature and the fixture positioning mark are imaged in the central area of ​​the real-time image.

[0074] By analyzing the brightness distribution or edge sharpness of the location marker area in the real-time image, it can be determined whether the location marker imaging meets the visual recognition requirements.

[0075] When the positioning mark is unclear, the relative distance between the industrial miniature camera and the material penetration feature is automatically adjusted by the telescopic device;

[0076] The process continues until the image brightness and edge sharpness of the real-time image meet the preset recognition conditions, or until the maximum extension stroke of the industrial miniature camera is reached.

[0077] When simultaneously acquiring real-time images of the material penetration feature and the fixture positioning mark, the industrial miniature camera must first be adjusted to face the corresponding fixture positioning mark penetration feature on the material. Through fine-tuning by the robot's end effector, the material penetration feature and fixture positioning mark are fully imaged in the central area of ​​the real-time image, thus reducing the impact of edge distortion on subsequent visual recognition. The system analyzes the uniformity of brightness distribution and edge contour sharpness of the positioning mark area in the real-time image using built-in algorithms. For example, it determines whether the edge gradient value of the positioning mark reaches a preset threshold or whether the average brightness of the area is within the optimal recognition range, thereby determining whether the positioning mark imaging meets the visual recognition requirements. When a change in illumination causes feature blurring, the telescopic device installed at the tail of the industrial miniature camera automatically activates, adjusting the relative distance between the industrial miniature camera and the material penetration feature by a preset step size, while continuously monitoring changes in image quality. This process continues until the brightness, contrast, or edge sharpness of the real-time image reaches the preset recognition conditions.

[0078] Specifically, in some possible embodiments, the steps of identifying the center position and angle of the material penetration feature and the center position and angle of the fixture positioning mark based on the visual algorithm include:

[0079] The real-time image is converted from the RGB color space to the HSV color space, and grayscale binarization is performed on the three dimensions of H, S, and V respectively. The region of interest for material penetration features and fixture positioning marks is initially determined by grayscale value filtering.

[0080] Based on the roundness, area, and edge gradient features extracted from the region of interest, interference from machining waste and cutting fluid residue is eliminated, and the target area with material penetration features and fixture positioning marks is locked.

[0081] One-dimensional grayscale projection is performed on the target area to generate grayscale contour lines and Gaussian smoothing is applied. The pixel-level coordinates of the contour points are extracted by the first derivative.

[0082] Within the neighborhood of pixel-level contour points, a weighted average of pixel values ​​is calculated using a bilinear interpolation algorithm to optimize and obtain sub-pixel precision point coordinates.

[0083] Based on the optimized sub-pixel level point coordinates, the contour shape of the material penetration feature and the fixture positioning mark is fitted.

[0084] In some possible embodiments, when identifying the position and angle of material penetration features and fixture positioning marks based on visual algorithms, the real-time images captured by the industrial miniature camera are first converted from the RGB color space to the HSV color space. By separating the information of the three dimensions of hue (H), saturation (S), and lightness (V), the impact of illumination changes on color recognition is reduced. Grayscale binarization processing is performed on each dimension. For example, non-target colors are filtered out by setting a hue range, or bright or dark areas are segmented by using a lightness threshold. This initially filters out regions of interest that may contain material penetration features and fixture positioning marks, reducing the amount of data in subsequent processing.

[0085] Subsequently, the system performs interference elimination based on the geometric features of the region of interest. By calculating indicators such as roundness (measures the degree to which the shape is close to a circle), area (number of pixels), and edge gradient (contour clarity) of each region, regions with roundness below the threshold (such as irregular shapes of machining debris) and regions with too small an area (such as cutting fluid spots) are excluded, and only target regions that meet the characteristics of penetration and positioning markers are retained.

[0086] The locked target area is subjected to one-dimensional grayscale projection, that is, the two-dimensional image is projected into a one-dimensional grayscale contour line along the horizontal or vertical direction. The contour line is then denoised using a Gaussian smoothing algorithm to suppress the interference of random noise on edge detection. Next, the gradient change of the contour line is calculated using the first derivative, and the locations of abrupt changes in grayscale value are extracted as contour points to determine the pixel-level edge coordinates of material penetration features and positioning markers, providing a foundation for subsequent sub-pixel optimization.

[0087] After obtaining pixel-level contour points, a weighted average of the pixel values ​​in the neighborhood of each contour point is calculated using a bilinear interpolation algorithm to optimize and obtain sub-pixel point coordinates with higher accuracy than pixel-level coordinates, meeting the requirements for high-precision material placement. Based on the sub-pixel-level coordinates, the contour shape of the material penetration feature and the fixture positioning mark is fitted using the least squares method or an ellipse fitting algorithm. By calculating the direction angle of the contour principal axis, the angular deviation between the material and the fixture is determined, providing accurate data for robot posture adjustment.

[0088] Specifically, in some possible embodiments, the method further includes:

[0089] When switching between producing different types of materials, the industrial miniature camera identifies the QR code on the surface of the material and automatically matches the preset positioning model of the material penetration feature with the fixture positioning mark.

[0090] Based on the identified material model, the corresponding visual template parameters are retrieved.

[0091] Specifically, in some possible embodiments, the step of determining whether the reflectivity of the positioning marker meets the visual recognition requirements by analyzing the brightness distribution or edge sharpness of the positioning marker area in the real-time image includes:

[0092] When the brightness and contrast of the location marker area in the real-time image are lower than a preset threshold, it is determined that there is surface contamination.

[0093] The robot is controlled to use an air blowing device at its end to clean the surface of the positioning markers and the lens of the miniature camera;

[0094] After cleaning, real-time images are captured again. If the brightness and contrast increases meet the preset standards, the feeding process continues.

[0095] When analyzing the brightness distribution or edge sharpness of the location marker area in a real-time image, the system first extracts the grayscale histogram of the location marker area and calculates the average brightness value and contrast.

[0096] Another embodiment of this application provides a precise material feeding system based on industrial miniature camera vision perception, wherein, see reference Figure 2 A precision feeding system based on industrial miniature camera vision perception includes:

[0097] Robot control module 100: The robot grasps material with at least two through features and moves it above a machine tool fixture with at least two positioning marks;

[0098] Real-time image acquisition module 200: Simultaneously acquires real-time images of material penetration features and fixture positioning marks through an industrial miniature camera installed on the robot's end effector;

[0099] Center position and angle recognition module 300: Based on visual algorithms, it identifies the center position and angle of material penetration features and the center position and angle of fixture positioning marks;

[0100] Deviation Calculation Module 400: Calculates the positional and angular deviation values ​​between the material penetration feature and the fixture positioning mark;

[0101] Motion trajectory and attitude adjustment module 500: Adjusts the robot's motion trajectory and attitude in real time according to the deviation value to form closed-loop control;

[0102] Material placement module 600: Repeatedly executes image acquisition, visual recognition, deviation calculation and trajectory adjustment steps until the position deviation value and angle deviation value are less than the preset threshold, thus completing the precise placement of materials.

[0103] The industrial miniature camera vision perception precision feeding system provided in this embodiment can achieve the steps of the aforementioned embodiment due to the functions of each module and the logical connections between them. Therefore, it can achieve the same technical effect as the aforementioned embodiment. For the principle analysis, please refer to the relevant description of the steps of the aforementioned industrial miniature camera vision perception precision feeding method, which will not be repeated here.

[0104] This application embodiment also provides a precision feeding device for industrial micro-camera vision perception, including a memory and a processor. The memory stores a computer program that can be loaded by the processor and executed by the above-described precision feeding method for industrial micro-camera vision perception.

[0105] This application embodiment also provides a storage medium storing a computer program that can be loaded by a processor and executed by the above-described industrial miniature camera visual perception precision feeding method.

[0106] The storage medium provided in this embodiment can achieve the same technical effect as the aforementioned embodiments because the computer program therein, after being loaded and run on the processor, will implement the various steps of the aforementioned embodiments. For the principle analysis, please refer to the relevant description of the aforementioned method steps, which will not be repeated here.

[0107] The storage medium includes, for example, various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0108] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0109] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the present invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of different embodiments or examples.

[0110] Furthermore, features defined by the terms "first" and "second" may explicitly or implicitly include at least one of those features. In the description of this invention, "a plurality of" means at least two, such as two, three, etc., and unless otherwise explicitly specified, is used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features.

[0111] Therefore, any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0112] The embodiments described in this specific implementation are preferred embodiments of this application and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. A precision discharge method for industrial micro-camera vision perception, characterized in that, The method comprises the following steps: A robot picks up a material with at least two through features, and moves above a machine tool fixture with at least two positioning marks; A real-time image of the material through feature and the fixture positioning mark is synchronously captured by an industrial micro camera installed at the end effector of the robot, the industrial micro camera is opposite to the through feature of the corresponding fixture positioning mark on the material, so that the material through feature and the fixture positioning mark are imaged in the central region of the real-time image; whether the imaging of the positioning mark meets the visual recognition requirement is determined by analyzing the brightness distribution and edge sharpness of the positioning mark region in the real-time image; when the positioning mark is not clear, the relative distance between the industrial micro camera and the material through feature is automatically adjusted by a telescopic device until the image brightness and edge sharpness of the real-time image meet the preset recognition condition, or the maximum telescopic stroke of the industrial micro camera is reached; The step of determining whether the imaging of the positioning mark meets the visual recognition requirement comprises: when the brightness value and contrast of the positioning mark region in the real-time image are lower than the preset threshold or the edge is blurred, it is determined that there is surface contamination; The robot uses the blowing device at the end to clean the surface of the positioning mark and the lens of the micro camera; after cleaning, the real-time image is captured again, and if the brightness value and contrast improvement meet the preset standard, the material placing process is continued; Based on the visual algorithm, the center position and angle of the material through feature and the center position and angle of the fixture positioning mark are recognized, the real-time image is converted from the RGB color space to the HSV color space, the H, S and V three dimensions are respectively subjected to gray scale binary processing, the gray scale value is screened, and the regions of interest of the material through feature and the fixture positioning mark are preliminarily determined; based on the circularity, area and edge gradient features extracted from the regions of interest, the machining waste and cutting fluid residue interference are excluded, and the target regions of the material through feature and the fixture positioning mark are locked; one-dimensional gray scale projection is performed on the target region to generate a gray scale contour line and perform Gaussian smoothing processing, and the pixel-level coordinates of the contour points are extracted through the first-order derivative; in the neighborhood of the pixel-level contour points, the pixel values are weighted and averaged by the bilinear interpolation algorithm, and the point coordinates with sub-pixel accuracy are obtained through optimization; Based on the optimized sub-pixel level point coordinates, the contour shapes of the material through feature and the fixture positioning mark are fitted, and the angle deviation is determined by the difference between the center point connection angles of the two material through features and the center point connection angles of the fixture positioning marks; The position deviation value and the angle deviation value of the material through feature and the fixture positioning mark are calculated; The motion trajectory and posture of the robot are adjusted in real time according to the deviation values to form a closed-loop control; The image capturing, visual recognition, deviation calculation and trajectory adjustment steps are repeatedly performed until the position deviation value and the angle deviation value are less than the preset threshold, and the accurate placement of the material is completed; When different models of materials are switched, the two-dimensional code on the surface of the material is recognized by the industrial micro camera, the preset positioning model of the material through feature and the fixture positioning mark is automatically matched, and corresponding visual template parameters are called based on the recognized material model.

2. An accurate discharge system of industrial micro-camera visual perception, applied to the method of claim 1, capable of realizing each step of the method of claim 1, characterized in that, The method comprises the following steps: The robot control module: the robot grabs the material with at least two through features, moves to above the machine tool fixture with at least two positioning marks; the real-time image acquisition module: through the industrial micro camera installed at the end effector of the robot, synchronously acquires the real-time image of the material through feature and the fixture positioning mark; the center position and angle identification module: based on the visual algorithm, identifies the center position and angle of the material through feature and the center position and angle of the fixture positioning mark; the deviation value calculation module: calculates the position deviation value and the angle deviation value of the material through feature and the fixture positioning mark; the motion trajectory and posture adjustment module: according to the deviation value, adjusts the robot motion trajectory and posture in real time, forms a closed loop control; the material placing module: repeatedly executes the image acquisition, visual identification, deviation calculation and trajectory adjustment steps, until the position deviation value and the angle deviation value are less than the preset threshold, completes the accurate placement of the material.

3. Precise discharge equipment for industrial micro-camera vision perception, characterized by, Comprise: A memory and a processor, the memory has stored thereon a computer program capable of being loaded and executed by the processor to perform the accurate material placing method of the industrial micro camera visual perception of claim 1.

4. A storage medium, characterized by A computer program capable of being loaded and executed by the processor to perform the accurate material placing method of the industrial micro camera visual perception of claim 1 is stored.

Images