Object positioning and automatic grabbing method and system based on dot-matrix inductor

Abstract

The invention discloses an object positioning and automatic grabbing method and system based on a dot-matrix sensor, and is suitable for an automatic material carrying scene. The method comprises the following steps of: uniformly arranging a photoelectric dot matrix sensor network below a working table, scanning and detecting a trigger signal of light intensity shielded by an object in real time, collecting two-dimensional coordinate data of a trigger sensor, calculating a geometric center position of the object by adopting a weighting algorithm, and combining contour fitting and standard contour template matching to obtain a geometric center position of the object. The system further ensures that the object state is static and stable through multi-cycle stability judgment, then robot path planning and pre-positioning are conducted on the basis of positioning information, finally, a robot tail end vision system conducts secondary confirmation on the object, accurate grabbing is achieved, and the robot can be accurately grabbed. The method improves the robustness of object recognition and the accuracy of grabbing, effectively reduces the false grabbing rate, is suitable for a high-speed assembly line and a complex operation environment, and has good practicability and safety.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of intelligent manufacturing and industrial automation, in particular to an object positioning and automatic grabbing method and system based on dot matrix sensors. BACKGROUND

[0002] In application scenarios such as automated production, logistics transportation and precision assembly, industrial robots have been widely used to complete repetitive and high-precision object grabbing and transportation tasks. In order to ensure the accuracy and reliability of grabbing, the robot system usually relies on a visual recognition system or a small number of sensors to detect the position and state of the target object. However, in the existing robot grabbing system, the judgment of the object state usually only depends on a single detection result, lacking a multi-cycle verification and redundant confirmation mechanism. When the object slides, tilts or is placed unstably, it is easy to cause problems such as misgrabbing, deviation and even equipment collision, affecting the stability and safety of the system. Therefore, an intelligent grabbing positioning method that integrates high-density perception, geometric calculation, state judgment and control execution is urgently needed, which can accurately identify the position and attitude of the object and analyze and confirm the stability of the object state, so as to improve the success rate of robot operation and the robustness of system operation. SUMMARY

[0003] The purpose of the present application is to provide an object positioning and automatic grabbing method and system based on dot matrix sensors to solve the problems raised in the background.

[0004] In order to solve the above technical problems, the present application provides the following technical scheme: an object positioning and automatic grabbing method based on dot matrix sensors, comprising the following steps: S1, uniformly arranging a plurality of photoelectric dot matrix sensors under the workbench to form a two-dimensional sensor array, setting a fixed interval and identifying their coordinates through coding; S2, the sensor network periodically scans the workbench at a set scanning frequency, when the target object blocks the sensor, triggering a light blocking signal and transmitting the coordinate and state information to the control system; S3, the control system collects the coordinate data of all triggered sensors, performs weighted calculation based on the triggering intensity and proximity, and determines the geometric center position of the object; S4, the control system performs geometric fitting on the triggered point set, constructs the object contour boundary, and performs contour matching judgment with the preset standard template; S5, judging the geometric center change, the number of triggered points and the contour matching degree in a plurality of continuous scanning periods, and starting the grabbing process when the stability condition is met; S6, the control system transmits the object position, contour orientation and grabbing point offset information to the robot controller, performs path planning and end effector pre-positioning; S7, the end vision system collects images, identifies the object boundary and compares with the sensing system data to confirm, then performs the grabbing action and completes the object carrying.

[0005] According to the above technical solution, the control system collects the coordinate data of all trigger sensors, and the steps of determining the geometric center position of the object based on the trigger intensity and proximity include: Performing trigger intensity calculation; Estimating the geometric center and calculating the proximity; The trigger intensity and proximity are integrated by a weighting formula to obtain the final weight of each sensor.

[0006] According to the above technical solution, the steps of estimating the geometric center and calculating the proximity include: The preliminary geometric center calculation formula is as follows: Wherein, And is the coordinate of the i-th trigger sensor, and n is the total number of trigger sensors. This preliminary estimated geometric center provides a rough reference position for subsequent calculation, and the proximity represents the distance between the sensor and the geometric center of the object. The closer the sensor to the object, the greater the contribution to the estimation of the object position. The control system measures the proximity by calculating the Euclidean distance between the sensor and the geometric center of the object: Wherein, ( , ) is the coordinate of the i-th sensor. The control system calculates the proximity according to the distance , usually using an inverse distance or exponential decay function to represent: Wherein, represents the proximity of the i-th sensor, and the proximity of the sensor closer to the object is greater.

[0007] According to the above technical solution, the steps of integrating the trigger intensity and proximity by a weighting formula to obtain the final weight of each sensor include: Wherein, and β are weight coefficients for balancing the influence of trigger intensity and proximity on weight. Through weighting, the system can ensure that the sensor closer to the object and with greater trigger intensity has higher weight in estimating the position of the object. Finally, the control system calculates the geometric center position of the object by weighted average algorithm ( , ), Wherein, ( , ) is the coordinate of the i-th trigger sensor, and the weight of the sensor is the weight of the sensor.

[0008] According to the above technical solution, the step of judging the geometric center change, the trigger point number fluctuation and the contour matching degree in the continuous multiple scanning periods, and starting the grabbing process after meeting the stability condition, comprises: In each period, the control system records the trigger sensor coordinate data collected in the current period, the calculated object geometric center position, the total number of trigger points and the contour matching result. After three consecutive scans are completed, the system performs the following stability determination: First, compare the geometric center coordinate values calculated in the three periods. When the change amplitude between any two periods is less than ±1 mm, it meets the position stability requirement. The specific judgment condition is:

[0009] : represents the X coordinate of the object center calculated in the i-th scanning period; : represents the Y coordinate of the object center calculated in the i-th scanning period; i+1: represents the next continuous period; Secondly, the system counts the number of triggered sensors in each period , and calculates the change rate of the number of trigger points. When the number fluctuation between any two periods is within ±5%, it is considered that the trigger point distribution is stable. The condition to be met is:

[0010] Finally, the system analyzes the matching degree change of the contour model constructed in each period and the standard template. When the matching degree of the three times remains above the set threshold (such as 85%), and the matching degree fluctuation between adjacent periods does not exceed the set value (such as ±2%), it is determined that the contour matching is stable.

[0011] When the above three indicators meet the requirements, the system determines that the object is in a stable and grabbable state at the current table position.

[0012] According to the above technical solution, the step of the control system transmitting the object position, contour orientation and grabbing point offset information to the robot controller for path planning and end effector positioning, comprises: The control system packages the geometric center coordinates of the current object, the contour graph and the recommended grabbing point information into control instructions, and transmits them to the robot control unit through the industrial communication bus (such as EtherCAT). The transmitted information includes but is not limited to: object center coordinates (X, Y), contour orientation (θ), recommended grabbing point (Xg, Yg). , , ), a contour orientation angle θ, offset coordinates (Δx, Δy) of a grabbing point relative to a center of the object, and a current system timestamp.

[0013] An object positioning and automatic grabbing system based on a dot matrix inductor, which is implemented by any of the above methods, characterized in that the system comprises: a perception and preprocessing module for preliminary perception of the object position, data acquisition and center identification, providing a data basis for subsequent path planning; a state judgment and decision module for stability verification and grabbing condition determination of the object state, ensuring the accuracy and reliability of the grabbing action; an execution control and action module for realizing the planning, confirmation and execution of the grabbing action, and completing system reset after the action, preparing for the next operation.

[0014] According to the above technical solution, the perception and preprocessing module comprises: a trigger identification module for identifying the existence and trigger state of the object on the table based on the dot matrix photoelectric inductor network, and collecting preliminary position data; a coordinate calculation module for collecting trigger inductor coordinates and calculating the geometric center of the object as the center estimation result; a contour fitting module for constructing a contour fitting graph, matching with a standard template, and confirming the object type and posture.

[0015] According to the above technical solution, the state judgment and decision module comprises: a periodic monitoring module for starting multi-period continuous data monitoring to monitor the motion trend and contact state of the object; a triple threshold judgment module for threshold determination of the center coordinate change, trigger point number fluctuation and contour matching degree to ensure static stability; a grabbing permission confirmation module for the system to determine that the object can be grabbed after meeting all stability conditions, entering the control execution phase.

[0016] According to the above technical solution, the execution control and action module comprises: a path planning and pre-positioning module for the controller to receive perception information and generate an optimal path, and the robot to move to a pre-positioning point above the object; an end visual comparison module for the end visual module to perform secondary confirmation of the object state, ensuring identification consistency and position accuracy; a grabbing and reset module for executing the clamping action, lifting the object after confirming the successful grabbing, and resetting the sensing system, preparing for the next identification process.

[0017] Compared with the prior art, the present application has the beneficial effects that: the present application can realize high-precision recognition and dynamic monitoring of objects on the workbench by setting a two-dimensional coordinate acquisition network based on a dot matrix photoelectric sensor, combining a multi-parameter weighted object geometric center estimation algorithm, and a contour matching and multi-cycle stability judgment mechanism; compared with the traditional positioning method relying on single visual recognition, the present application can effectively avoid recognition errors caused by factors such as shielding, reflection, and weak texture. By formulating the calculation of trigger strength, proximity, and geometric center, more robust target estimation is realized; combined with end visual secondary confirmation and real-time monitoring of clamping force, the robot grasping action has higher success rate and environmental adaptability, especially suitable for intelligent material handling in high-speed assembly lines and multi-target dynamic scenes. The overall scheme has the advantages of simple structure, high recognition accuracy, strong stability, and rapid system response, which can significantly improve the reliability and work efficiency of the automatic grasping system. BRIEF DESCRIPTION OF DRAWINGS

[0018] The accompanying drawings are included to provide a further understanding of the present application, and constitute a part of the specification, illustrate the present application, and are used to explain the present application, and do not constitute a limitation on the present application. In the drawings: Figure 1 An implementation step diagram of an object positioning and automatic grasping method based on a dot matrix sensor provided for the first embodiment of the present application; Figure 2 A module composition schematic diagram of an object positioning and automatic grasping system based on a dot matrix sensor provided for the second embodiment of the present application. DETAILED DESCRIPTION

[0019] The technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0020] Embodiment one: Figure 1 A step diagram of an object positioning and automatic grasping method based on a dot matrix sensor provided for the first embodiment of the present application, the present embodiment can be applied to the scene of automatic material handling, and the method can be executed by an object positioning and automatic grasping system based on a dot matrix sensor provided by the present embodiment, as shown in Figure 1 The method specifically includes the following steps: Step one: target detection trigger; In the embodiment of the present application, a plurality of photoelectric sensors are uniformly arranged under the workbench to form a dot matrix sensor network. The spacing between each sensor is set to a fixed value (such as 10 mm x 10 mm), and each sensor is identified by a specific coding method. The sensor type is a photoelectric sensor, which is used to detect the blocking of the light intensity of the workbench by the object. When the system starts working, the sensor network periodically scans the workbench at a set scan frequency (such as 10 Hz), ensuring that 10 detections are made per second. During each scan, each sensor in turn determines whether a change in light intensity has been detected. When an object is placed above the workbench, the surface of the object will block some of the sensors below, causing a change in light intensity for these sensors, thereby triggering the light blocking state of the sensors. Each triggered sensor transmits its position coordinates and state information (light blocking trigger signal) to the control system. The control system determines the position of the sensors covered by the object according to the received trigger signal and records the coordinate information of all triggered sensors. These data will be used for subsequent geometric center calculation and object position estimation, providing accurate input for subsequent path planning and action execution.

[0021] Step two: coordinate data acquisition and center calculation; In the embodiment of the present application, after the target object is placed and triggers several sensors in the dot matrix sensor network, the control system begins to collect the coordinate data of all triggered sensors. The coordinates of each triggered sensor are transmitted to the control system through an industrial bus (such as EtherCAT, CAN bus, etc.), and the control system identifies the position of each sensor according to the coding method of the sensor and records the coordinate information of the triggered sensors. The coordinate value of each sensor is usually represented in a two-dimensional coordinate system, with the X-axis representing the horizontal direction and the Y-axis representing the vertical direction. Subsequently, the control system performs a weighted calculation on all triggered sensors to determine the geometric center position of the object. The weighted algorithm takes into account the trigger strength of the sensor and the distance from the geometric center of the object, and the specific calculation steps are as follows: For example, first, the trigger strength is calculated, which reflects the degree of blocking of the sensor by the object and is measured by the change in light intensity of the sensor. Assuming that the light intensity of the sensor without object blocking is and the light intensity when the object blocks is then the trigger strength is: Next, the control system normalizes the trigger strength of each sensor to calculate the normalized trigger strength : where is the maximum trigger strength value. The normalized trigger strength allows for a uniform scale between different sensors. Subsequently, the geometric center is estimated and the proximity is calculated. Since the exact geometric center cannot be immediately determined, it is assumed that the geometric center of the object is near the average position of all trigger sensors. The preliminary geometric center calculation is as follows: wherein, and is the coordinate of the i-th trigger sensor, and n is the total number of trigger sensors. This preliminary estimate of the geometric center provides a rough reference position for subsequent calculations. The proximity represents the distance between the sensor and the geometric center of the object, and the closer the sensor, the greater its contribution to the estimation of the object's position. The control system measures the proximity by calculating the Euclidean distance between the sensor and the geometric center of the object: wherein, , is the coordinate of the i-th sensor. The control system calculates the proximity based on the distance , typically using an inverse or exponential decay function: wherein, represents the proximity of the i-th sensor, and the closer the sensor, the greater its proximity.

[0022] By way of example, the control system combines the trigger strength and proximity to calculate the final weight of each sensor by a weighted formula, : wherein, and β are weight coefficients used to balance the influence of trigger strength and proximity on the weight. Through weighting, the system ensures that the closer and stronger sensors have a higher weight in estimating the position of the object. Finally, the control system calculates the geometric center position of the object by a weighted average algorithm, , ): wherein, , is the coordinate of the i-th trigger sensor, and the weight of the sensor. Through this weighted average method, the system can calculate the geometric center position of the object as an estimate of the center of the object. This is transmitted to the subsequent path planning system and grasping control system for calculating the optimal grasping path and executing the grasping action.

[0023] Step three: contour recognition and matching judgment; In the embodiment of the present application, after the estimation of the geometric center of the object is completed, the control system further identifies and models the object contour based on all the triggered sensor coordinate points. First, the control system inputs the coordinate information of all the triggered points as a point set, adopts a geometric fitting algorithm to construct the contour boundary of the point set, and specifically can select the minimum circumscribed rectangle method or the convex hull algorithm for boundary fitting. The minimum circumscribed rectangle method calculates the minimum rectangle that can enclose all the triggered points by rotating the point set, and the result can be used to quickly estimate the length, width and orientation of the object; while the convex hull algorithm forms an approximate contour that fits the boundary of the point set by constructing a non-concave minimum closed polygon. After the contour modeling is completed, the system takes the approximate contour graph as the object boundary model detected at present.

[0024] Subsequently, the control system calls the standard contour template of the corresponding object from the local database and performs matching comparison between the approximate contour graph and the template contour. The matching process can adopt a contour shape similarity comparison algorithm, for example, a comprehensive evaluation is performed by using parameters such as the length-width ratio of the contour shape, the number of corner points, the contour area, the boundary length, etc. The matching degree can be expressed as a percentage value, reflecting the similarity between the current contour and the standard model. When the system judges that the contour matching degree is higher than a preset threshold (such as 85%), it is considered that the object type recognition is successful, and the next step of grasping preparation stage is entered; if the matching degree is insufficient, the system determines that the object is placed abnormally or there is foreign matter interference, and the grasping process is terminated and an abnormal prompt information is output.

[0025] This step can make the system effectively identify the approximate shape and type of the object on the current table surface through boundary fitting and template comparison of the dot matrix sensor trigger data, so as to ensure that the subsequent grasping action has the basis of object recognition, and improve the identification robustness and error tolerance rate of grasping.

[0026] Step four: multi-period stability judgment; In the embodiment of the present application, in order to ensure that the target object is in a static and stable state before grasping, the control system continues to monitor the state of the target region for a plurality of continuous periods after the object is identified. Specifically, the system sets a fixed scanning period, for example, completes a complete scan every 0.1 seconds, acquires data at a frequency of 10 Hz, and compares and analyzes whether the state of the target object changes significantly within three continuous scanning periods.

[0027] For example, in each period, the control system records the triggered sensor coordinate data collected in the current period, the calculated geometric center position of the object, the total number of triggered points, and the contour matching result. After three continuous scans are completed, the system performs the following stability judgment: First, compare the geometric center coordinate values calculated in the three periods, and when the change amplitude between any two periods is less than ±1 mm, the position stability requirement is met. The specific judgment condition is:

[0028] Xi: represents the X coordinate of the object center calculated in the i-th scanning cycle; Yi: represents the Y coordinate of the object center calculated in the i-th scanning cycle; i+1: represents the next consecutive cycle; Secondly, the system counts the number of sensors triggered in each cycle , and calculates the rate of change of the number of trigger points. When the number fluctuates within ±5% between any two cycles, it is considered that the distribution of trigger points is stable. The condition to be met is:

[0029] Finally, the system analyzes the change of the matching degree of the contour model constructed in each cycle with the standard template. When the matching degree of three times remains above the set threshold (such as 85%), and the fluctuation of the matching degree between adjacent cycles does not exceed the set value (such as ±2%), it is determined that the contour matching is stable.

[0030] When the above three indicators meet the requirements, the system determines that the object is in a stable and graspable state at the current table position, allowing the robot control system to enter the next path planning and end position preparation stage. If any of the indicators does not meet the requirements, the system considers that the target state is unstable, and keeps waiting or triggers the re-identification process to avoid the risk of mis-grabbing caused by object shaking, sliding or position deviation.

[0031] By setting a multi-cycle stability judgment mechanism, the reliability of the object recognition and grasping process can be effectively improved, which is especially suitable for high-speed assembly lines or scenes with human intervention, ensuring accuracy while enhancing the robustness and safety of the system.

[0032] Step five: robot path planning and pre-positioning.

[0033] In the embodiments of the present application, after confirming that the object state meets the stability judgment condition, the control system packages the geometric center coordinates of the current object, the contour graph and the recommended grasping point information into control instructions, and transmits them to the robot control unit through an industrial communication bus (such as EtherCAT). The transmitted information includes but is not limited to: object center coordinates (X, Y), , , ), contour orientation angle θ, grasping point offset coordinates relative to the object center (Δx, Δy), and the current system timestamp.

[0034] Exemplarily, after receiving the above information, the robot controller first calls the path planning module to solve the trajectory according to the current position state of the robot arm and the position of the target grasping point, using a preset kinematic model. During the trajectory planning process, multiple factors such as the degrees of freedom of the robot arm, joint limitations, end posture requirements, obstacle avoidance paths, safety heights, and the like are comprehensively considered to ensure that the grasping path is efficient and does not produce interference. After the path is solved, the robot system controls the end effector (i.e., the robot hand) to move to the predetermined position state according to the planned trajectory, which is usually a hovering height of about 50 mm above the target object. This hovering height can ensure that the end vision system has a good field of view, and also avoids the robot hand from accidentally touching the object or other structures before confirming the grasping. The system can automatically adjust the pre-positioning height according to the size of the object to adapt to the grasping requirements of different types of objects. After completing the pre-positioning, the robot remains stationary to wait for the end vision system to perform the next image confirmation operation, so as to verify and trigger the final grasping instruction. After the entire pre-positioning process is completed, the system enters the execution phase.

[0035] Step six: end vision confirmation and grasping execution.

[0036] In the embodiment of the present application, after the robot end effector completes the pre-positioning operation, the system starts the end vision unit to perform secondary identification and confirmation on the target object. The vision unit is installed at the end of the robot hand, usually a compact industrial camera or depth camera module, which has high-definition image acquisition and edge detection capabilities. The control system controls the vision unit to acquire the current image of the object, and uses image processing algorithms to extract the edge profile of the object. Exemplarily, after the image acquisition is completed, the control system matches and compares the identified profile in the image with the profile model constructed by the dot matrix sensing system in the early stage. The matching process can calculate the matching degree based on the profile barycenter position, boundary shape features, area ratio, length-width ratio, corner point number, boundary curvature, and the like. If the matching degree is higher than a set threshold (for example, 90%), and the profile attitude direction and the grasping point coordinates remain consistent, the system confirms that the target has not been displaced or changed in attitude, and the grasping condition is established.

[0037] After confirmation, the control system immediately issues a grasping execution instruction to drive the robot hand to press down to the surface of the target object along the predetermined path and perform the grasping action. The grasping process can be completed by a gripper, a vacuum suction cup, or a magnetic suction end effector, and the system selects the most suitable grasping method according to the material and structure of the object. At the same time of executing the grasping, the clamping force sensor monitors the grasping state in real time to determine whether the object is successfully clamped. If the clamping force is within a specified threshold range, the system determines that the grasping is successful.

[0038] Subsequently, the mechanical arm lifts the object to a preset safety height and exits the current work area along the original path or a planned path. Meanwhile, the dot matrix sensor network continuously performs state monitoring, and when all the sensing points return to the untriggered state, the system determines that the object has completely moved out of the tabletop sensing area, the task flow ends, and the system resets to enter the next round of work preparation state.

[0039] This step improves the accuracy verification capability before the grabbing action through the fusion comparison of end vision and sensing data, avoids the misgrabbing caused by object micro-motion, interference shielding or identification deviation, and effectively guarantees the stable operation and action accuracy of the robot system in a multi-target and high-density work scene.

[0040] Embodiment two: The embodiment two of the present application provides an object positioning and automatic grabbing system based on a dot matrix sensor, Figure 2 A module composition schematic diagram of the object positioning and automatic grabbing system based on the dot matrix sensor is shown in the embodiment two of the present application, and the system comprises: Figure 2 a perception and preprocessing module, which is used for the preliminary perception of the object position, data acquisition and center identification, and provides a data basis for subsequent path planning; a state judgment and decision module, which is used for stability verification and grabbing condition judgment of the object state, and ensures the accuracy and reliability of the grabbing action; an execution control and action module, which is used for realizing the planning, confirmation and execution of the grabbing action, and resets the system after the action to prepare for the next work; In some embodiments of the present application, the perception and preprocessing module comprises: a trigger identification module, which is used for identifying the existence and triggered state of the object on the tabletop based on the dot matrix photoelectric sensor network, and collecting preliminary position data; a coordinate calculation module, which is used for collecting the coordinates of the triggered sensor and calculating the geometric center of the object as a center estimation result; a contour fitting module, which is used for constructing a contour fitting graph, matching with a standard template, confirming the object type and posture, and the like; In some embodiments of the present application, the state judgment and decision module comprises: a period monitoring module, which is used for starting multi-period continuous data monitoring, monitoring the motion trend and contact state of the object; a triple threshold judgment module, which is used for threshold judgment on the center coordinate change, the number fluctuation of the triggered points and the contour matching degree, and ensures the static stability; a grabbing permission confirmation module, which is used for confirming that the object can be grabbed after all the stability conditions are met, and enters the control execution stage.

[0041] ​In some embodiments of the present application, the execution control and action module comprises: A path planning and pre-positioning module, configured to control the controller to receive the perception information and generate an optimal path, and move the robot arm to a pre-positioning point above the object; An end vision comparison module, configured to perform secondary confirmation on the object state by the end vision module, to ensure consistency of recognition and accuracy of position; A grasping and resetting module, configured to perform a clamping action, to lift the object after confirming the success of the grasping, and to reset the sensing system, to prepare for the next identification process.

[0042] It should be noted that, in this document, relational terms such as first and second and the like can only be used to distinguish one entity or action from another entity or action, without necessarily requiring or implying any actual such relationship or order between or among the entities or actions. Moreover, the terms "comprises", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus.

[0043] Finally, it should be noted that the above only describes the preferred embodiments of the present application, and is not intended to limit the present application. Although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions recorded in the foregoing embodiments, or replace some of the technical features with equivalent ones. Any modifications, equivalent replacements, improvements, and the like made within the spirit and principles of the present application shall be included in the protection scope of the present application.

Claims

1. A method for object positioning and automatic grasping based on dot matrix sensor, characterized in that, Comprise the following steps: S1, uniformly arrange a plurality of photoelectric dot array sensors under the workbench to form a two-dimensional sensor array, set a fixed interval and identify the respective coordinates through coding; S2, the sensor network periodically scans the workbench at a set scanning frequency, when the target object blocks the sensor, triggers the light shielding signal and transmits the coordinate and state information to the control system; S3, the control system collects the coordinate data of all triggered sensors, performs weighted calculation based on the triggering strength and proximity, and determines the geometric center position of the object; S4, the control system geometrically fits the trigger point set, constructs the object contour boundary, and performs contour matching judgment with the preset standard template; S5, judge the geometric center change, trigger point number fluctuation and contour matching degree respectively in continuous multiple scanning periods, and start the grabbing process after meeting the stability condition; S6, the control system transmits the object position, contour orientation and grabbing point offset information to the robot controller, performs path planning and end effector pre-positioning; S7, the end visual system collects images, identifies the object boundary and compares and confirms with the data of the sensing system, then performs grabbing action and completes object carrying.

2. The object positioning and automatic grasping method based on dot matrix sensor according to claim 1, wherein, The step of the control system collecting coordinate data of all triggered sensors, performing weighted calculation based on triggering strength and proximity, and determining the geometric center position of the object, comprises: Perform triggering strength calculation; Estimate the geometric center and calculate the proximity; Synthetically calculate the triggering strength and proximity by the weighted formula to obtain the final weight of each sensor.

3. The method of claim 2, wherein the method further comprises: The step of estimating the geometric center and calculating the proximity, comprises: The preliminary geometric center calculation formula is as follows: wherein, and is the coordinate of the i-th trigger sensor, n is the total number of trigger sensors, the preliminary estimated geometric center provides a rough reference position for subsequent calculation, the proximity represents the distance between the sensor and the geometric center of the object, the closer the distance, the greater the contribution of the sensor to the estimation of the object position, the control system measures the proximity by calculating the Euclidean distance between the sensor and the geometric center of the object wherein, ( , ) is the coordinate of the i-th sensor, the control system calculates the proximity according to the distance , that is, the reciprocal or exponential decay function of the distance is used to represent: wherein, the proximity of the i-th sensor represents the proximity, the closer the distance, the greater the proximity.​​ 4. The method of claim 2, wherein the method further comprises: The step of synthetically calculating the triggering strength and proximity by the weighted formula to obtain the final weight of each sensor, comprises: wherein, and β are weight coefficients used to balance the influence of trigger strength and proximity on the weight, by weighting, the system can ensure that the sensors closer and with larger trigger strength have higher weight on the estimation of the object position, finally, the control system calculates the geometric center position of the object by weighted average algorithm , ), wherein, ( , ) is the coordinate of the i-th trigger sensor, is the weight of the sensor.

5. The method of claim 1, wherein the method further comprises: The step of judging the geometric center change, trigger point number fluctuation and contour matching degree respectively in continuous multiple scanning periods, and starting the grabbing process after meeting the stability condition, comprises: In each period, the control system records the trigger sensor coordinate data collected in the current period, the object geometric center position calculated, the total number of trigger points and the contour matching result, after three consecutive scans, the system performs the following stability judgment: Firstly, compare the geometric center coordinate values calculated in three periods, when the change amplitude between any two periods is less than ±1mm, that is, it meets the position stability requirement, the specific judgment condition is: ； Xi: represents the X coordinate of the center of the object calculated in the i-th scanning cycle; : indicates the Y coordinate of the center of the object calculated in the i-th scanning cycle; i+1: represents the next continuous period; Secondly, the system counts the number of sensors triggered in each cycle and calculates the rate of change of the number of trigger points. If the number fluctuates within ±5% between any two cycles, it is considered that the distribution of trigger points is stable, and the condition is met: ； Finally, the system analyzes the matching degree change of the contour model constructed in each period and the standard template, when the matching degree of three times remains above the set threshold value, and the matching degree fluctuation between adjacent periods does not exceed the set value, that is, the contour matching is stable, When the above three indicators meet the requirements, the system determines that the object is in a stable and grabbable state at the current table position.

6. The method of claim 1, wherein the method further comprises: The step of the control system transmitting the object position, contour orientation and grabbing point offset information to the robot controller, performing path planning and end effector pre-positioning, comprises: The control system packs the current object's geometric center coordinates, the proposed contour graph, and the recommended grasp point information into a control command, which is transmitted to the robot control unit through an industrial communication bus. The transmitted information includes but is not limited to: object center coordinates (x, y, z) , , ), contour orientation angle θ, grasp point offset coordinates relative to the object center (Δx, Δy), and the current system timestamp.

7. A dot-matrix sensor-based object positioning and automatic grasping system, which is implemented by the method of any one of claims 1 to 6, characterized in that, The system comprises: A perception and preprocessing module for preliminary perception of object position, data acquisition and center identification to provide data basis for subsequent path planning; The state judgment and decision module is used for stability verification and grasping condition judgment of the object state, and ensures the accuracy and reliability of the grasping action. The execution control and action module is used for planning, confirming and executing the grasping action, and resets the system after the action, and prepares for the next operation.

8. The dot-matrix inductor-based object positioning and automatic grasping system according to claim 7, wherein, The perception and preprocessing module comprises: The trigger identification module is used for identifying the existence and trigger state of the object on the table based on the dot matrix photoelectric sensor network, and collecting preliminary position data. The coordinate calculation module is used for collecting trigger sensor coordinates and calculating the geometric center of the object as the center estimation result. The contour fitting module is used for constructing a contour fitting graph, matching with a standard template, confirming the object type and posture.

9. The dot-matrix inductor-based object positioning and automatic grasping system according to claim 7, wherein, The state judgment and decision module comprises: The cycle monitoring module is used for starting multi-cycle continuous data monitoring, monitoring the motion trend and contact state of the object. The triple threshold judgment module is used for threshold judgment of the center coordinate change, trigger point number fluctuation and contour matching degree, and ensures the static stability. The grasping permission confirmation module is used for confirming that the object can be grasped after satisfying all stability conditions, and entering the control execution stage.

10. The dot-matrix inductor-based object positioning and automatic grasping system according to claim 7, wherein, The execution control and action module comprises: The path planning and pre-positioning module is used for the controller to receive perception information and generate an optimal path, and the manipulator moves to a pre-positioning point above the object. The end visual comparison module is used for the end visual module to perform secondary confirmation on the object state, and ensures the identification consistency and position accuracy. The grasping and resetting module is used for executing the clamping action, lifting the object after confirming the grasping success, and resetting the sensing system, and preparing for the next identification process.

Images