Robot collaborative operation control method and system in edible mushroom planting process

Abstract

The invention belongs to the field of edible mushroom planting control, and particularly relates to a robot collaborative operation control method and system in the edible mushroom planting process, and the method comprises the steps: an inspection robot scans a mushroom bed, judges the maturity of edible mushrooms in combination with an image recognition algorithm, calculates the regional to-be-picked density, and generates a to-be-picked task list; according to the list, a preset parameter table and a curve, dynamically generating a coordinated picking task control instruction including the target coordinate, the number of the robots and the specific clamping posture, force and lifting speed of each robot; after the picking robot responds to the instruction to execute operation, a signal is fed back to the nearest inspection robot, secondary imaging confirmation is triggered, if residues exist, a supplementary picking list is generated according to the residue density, execution is repeated, and therefore efficient, accurate and lossless automatic cooperative harvesting of the edible mushrooms is achieved.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of edible mushroom cultivation control, and particularly relates to a robot collaborative operation control method and system in the process of edible mushroom cultivation. BACKGROUND

[0002] In the process of modern edible mushroom cultivation, using robots for automatic inspection and picking has become a key technology to improve production efficiency. Existing systems usually rely on inspection robots to collect mushroom bed images and environmental data along a preset initial path, and based on this, the maturity of edible mushrooms is preliminarily judged, and then a picking task list is generated to dispatch picking robots to perform the operation. However, such systems still have significant deficiencies in actual application; first, the maturity determination of the inspection link relies on two-dimensional image features of a single perspective, lacking deep integration with three-dimensional spatial position, regional density and growth prediction model, resulting in insufficient accuracy of maturity evaluation, making it difficult to accurately identify individuals and areas to be picked. Secondly, the task generation and scheduling strategy does not fully consider real-time load balancing, dynamic path planning among multiple robots, and fine picking parameters matched with maturity, which easily leads to high damage rate, low efficiency and uneven resource allocation in the picking process. In addition, the existing system lacks effective closed-loop verification mechanism, and cannot quickly confirm the effect and make supplementary picking decisions after picking is completed, which may cause residual problems and affect the overall harvest quality and output. These technical bottlenecks restrict the reliability, efficiency and automation level of the robot collaborative operation system in complex planting environments. SUMMARY

[0003] In view of the deficiencies of the prior art, the present application provides a robot collaborative operation control method and system in the process of edible mushroom cultivation, which comprises: an inspection robot scans the mushroom bed, determines the maturity of edible mushrooms in combination with an image recognition algorithm and calculates the area to be picked, and generates a to-be-picked task list; a central control unit dynamically generates a coordinated picking task control instruction containing target coordinates, the number of robots, and the specific clamping posture, force and pulling speed of each robot according to the list and preset parameter table and curve; the picking robot feeds back a signal to the nearest inspection robot after executing the operation in response to the instruction and triggers a secondary imaging confirmation, and if there are residues, a supplementary picking list is generated according to the residue density and repeated execution, thereby realizing efficient, accurate and lossless automatic collaborative harvesting of edible mushrooms.

[0004] To achieve the above-mentioned purpose, the present application provides the following technical solutions:

[0005] The robot collaborative operation control method in the process of edible mushroom cultivation comprises:

[0006] The inspection robot scans the mushroom bed along a preset initial path to collect environmental perception data of edible mushrooms;

[0007] The environmental perception data of the edible fungi is combined with an image recognition algorithm to determine the maturity of at least one edible fungus and calculate the area density of the edible fungi to be picked, and a picking task list containing the coordinates of the area to be picked, the maturity level and the area density of the edible fungi to be picked is generated; the environmental perception data of the edible fungi at least contains the image and three-dimensional space information of the edible fungi;

[0008] Based on the picking task list, a coordinated picking task control instruction is generated by combining a load balancing strategy and a preset maturity-gripping posture parameter table, a maturity-gripping force curve and a force-pulling speed curve; the coordinated picking task control instruction at least contains target picking coordinates, the number of picking robots and the corresponding gripping posture, gripping force and pulling speed control parameters of each picking robot;

[0009] In response to the coordinated picking task control instruction, the end gripper of the picking robot performs the picking operation with the gripping posture, gripping force and pulling speed, and after the picking is completed, a picking completion signal is fed back to the inspection robot closest to the picking area, and a secondary imaging confirmation of the picking area is triggered, if there is picking residue, a supplementary picking task list is generated according to the picking density of the residual area and the corresponding coordinated picking task control instruction is repeated.

[0010] Specifically, the image and three-dimensional space information of the edible fungi are collected, including:

[0011] S101, the first inspection robot continuously collects first environmental perception data within the preset initial path coverage range and second environmental perception data within the preset initial path coverage range and within the collection angle range during the travel of the first inspection robot along the preset initial path in the preset world coordinate system;

[0012] S102, based on the second environmental perception data, the position of the second to-be-evaluated mushroom area in the second environmental perception data is identified in real time through a lightweight convolutional neural network, and based on the position of the second to-be-evaluated mushroom area, a second inspection robot corresponding to the preset initial path to which the position of the second to-be-evaluated mushroom area belongs is queried through an information sharing algorithm.

[0013] Specifically, the image and three-dimensional space information of the edible fungi are collected, including:

[0014] S103, based on the second environmental perception data, the initial maturity discrimination score of the second to-be-evaluated mushroom area is calculated; the initial maturity discrimination score is obtained based on the estimated density of the mushroom in the second to-be-evaluated mushroom area, the Euclidean distance between the second to-be-evaluated mushroom area and the current position of the inspection robot, and the growth state of the edible fungi predicted based on a preset maturity growth prediction model in combination with an expert evaluation system;

[0015] S104. When the initial maturity discrimination score of at least one second mushroom body region to be evaluated exceeds the set threshold, the first inspection robot shares the second environmental perception data and the initial maturity discrimination score with the second inspection robot through the information sharing algorithm, based on the query obtained by the second inspection robot, and constructs the third environmental perception data of the second inspection robot by combining the first environmental perception data collected by the second inspection robot in real time.

[0016] Specifically, the process of obtaining the list of tasks to be picked includes:

[0017] Based on the third environmental perception data of any inspection robot, the images are arranged in the order of collection timestamps to obtain the sequence of images to be identified, wherein each frame of the image is associated with the three-dimensional spatial information marker of the corresponding timestamp.

[0018] Each frame in the image sequence to be identified is subjected to coordinate mapping, interference suppression, feature repair, and inter-frame redundancy removal to obtain a complete inter-frame incremental region image sequence for any inspection robot; the length of the complete inter-frame incremental region image sequence is the same as the preset initial path length of the corresponding inspection robot.

[0019] Based on the complete inter-frame incremental region image sequence of any inspection robot, the standardized mushroom body region image is obtained through image size normalization processing.

[0020] Based on the standardized mushroom body region image, the improved YOLOv8 algorithm is used to detect mushroom targets and obtain the bounding box coordinates of each edible fungus.

[0021] Based on the bounding box coordinates of each edible fungus, a region image of a single fungus is obtained through a region segmentation algorithm, and the edge contour data of a single fungus is obtained through an edge detection algorithm based on the single fungus region image.

[0022] Based on the edge contour data of a single mushroom, the contour data of the cap and the stem are obtained through a contour segmentation algorithm.

[0023] Specifically, the process of obtaining the list of tasks to be picked also includes:

[0024] Based on the cap outline data, we obtained the cap diameter parameter, cap edge curvature parameter, cap color RGB mean parameter, and gill clarity parameter.

[0025] Based on the stipe outline data, the stipe height parameter is obtained through a height calculation algorithm;

[0026] Based on the bounding box coordinates of all edible fungi, the total number of edible fungi detected within the preset initial path is obtained through quantity statistics, and is recorded as the total count of mushroom bodies in the region.

[0027] Specifically, the obtaining process of the to-be-picked task list further includes:

[0028] Based on the initial maturity discrimination score in the current preset initial path and the single-mushroom feature parameter set extracted from the corresponding area, the maturity score and the corresponding maturity level of the edible fungi at each position in the preset initial path of any inspection robot are obtained through the maturity growth prediction model and the pre-trained expert evaluation system; the single-mushroom feature parameter includes a cap diameter parameter, a cap edge curvature parameter, a cap color RGB mean value parameter, a lobe definition parameter, and a stem height parameter.

[0029] Based on the maturity level of the edible fungi at each position in the preset initial path of any inspection robot and the corresponding single-mushroom feature parameter, the single-mushroom feature parameter similarity and the maturity score deviation ratio are taken as the clustering density parameter, and the adaptive clustering algorithm is used to perform regional clustering division on the preset initial path to obtain the edible fungi evaluation clustering cluster set on the current preset initial path.

[0030] Based on the edible fungi evaluation clustering cluster set on the current preset initial path and the preset to-be-picked evaluation threshold, the to-be-picked edible fungi evaluation clustering cluster set and the to-be-mature edible fungi evaluation clustering cluster set are obtained.

[0031] Specifically, the obtaining process of the to-be-picked task list further includes:

[0032] Based on the ratio of the corresponding maturity score, the to-be-picked edible fungi evaluation clustering cluster number, and the total number of area mushroom bodies in each to-be-picked edible fungi evaluation clustering cluster, the preset edible fungi growth curve, and the expert evaluation algorithm, the to-be-picked priority of each to-be-picked edible fungi evaluation clustering cluster is obtained.

[0033] Meanwhile, according to the number of edible fungi in each to-be-picked edible fungi evaluation clustering cluster, the regional to-be-picked density of each to-be-picked edible fungi evaluation clustering cluster is calculated.

[0034] The to-be-picked task list in each preset initial path is constructed by using the to-be-picked priority of the to-be-picked edible fungi evaluation clustering cluster in each preset initial path, the corresponding regional to-be-picked density, and the position coordinate information of the to-be-picked edible fungi evaluation clustering cluster.

[0035] Specifically, the generated coordination picking task control instruction includes:

[0036] The number, position coordinates and remaining working time of the picking robots in the working state in the current moment are acquired, and the number and corresponding coordinates of the instantaneously scheduled picking robots and the number and corresponding coordinates of the scheduled waiting picking robots are acquired by using a statistical algorithm, wherein the remaining working time is predicted according to the historical average picking rate of the picking robots, the remaining number of edible fungi in the corresponding edible fungi evaluation clustering group and the number of the picking robots currently working in the corresponding edible fungi evaluation clustering group in combination with an LSTM algorithm;

[0037] Based on the number and corresponding coordinates of the instantaneously scheduled picking robots and the number and corresponding coordinates of the scheduled waiting picking robots in combination with the list of the picking tasks in each preset initial path, the picking robot distribution label list and the corresponding travel path parameters of each edible fungi evaluation clustering group are acquired by using an optimal task distribution algorithm and a load balancing algorithm, with the shortest travel time and picking task completion time as the target;

[0038] Based on the picking robot distribution label list and the corresponding travel path parameters of each edible fungi evaluation clustering group in combination with a fuzzy control algorithm, the optimal path instructions of the picking robots corresponding to each edible fungi evaluation clustering group are generated, wherein the optimal path instructions of the picking robots corresponding to each edible fungi evaluation clustering group include the number of the distributed picking robots, the travel starting point and ending point, the travel starting time point and the travel time length of each picking robot.

[0039] Specifically, generating the coordinated picking task control instructions further includes:

[0040] Based on the maturity score and the corresponding maturity level of each edible fungus in the edible fungi evaluation clustering group in combination with a maturity-gripping posture parameter table, the gripping posture parameters of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering group are obtained;

[0041] Meanwhile, based on the maturity score and the corresponding maturity level of each edible fungus in the edible fungi evaluation clustering group in combination with a maturity-gripping force curve, the gripping force parameters of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering group are obtained;

[0042] Based on the gripping force parameters of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering group in combination with a force-lifting speed curve, the picking lifting speed control parameters of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering group are obtained;

[0043] The gripping posture parameter, the force parameter and the pulling speed control parameter of each picking robot in the cluster of the single edible mushroom for picking the current edible mushroom to be picked are evaluated based on a PID control algorithm to generate the picking task control instruction of each picking robot.

[0044] The picking task control instruction of all picking robots is combined with a simulation algorithm to optimize training with the minimum picking damage ratio as the target to generate the coordinated picking task control instruction in each cluster of the edible mushroom to be picked for real-time picking control.

[0045] The robot cooperative operation control system in the edible mushroom planting process comprises a data acquisition module, an identification module, an instruction module, a response module and a feedback confirmation module.

[0046] The data acquisition module is used for the scanning of the mushroom bed by the inspection robot along the preset initial path to collect the environment perception data of the edible mushroom.

[0047] The identification module determines the maturity of at least one edible mushroom and calculates the regional picking density based on the environment perception data of the edible mushroom combined with an image recognition algorithm to generate a picking task list containing the coordinates of the picking area, the maturity level and the regional picking density; the environment perception data of the edible mushroom at least contains the image and three-dimensional space information of one edible mushroom.

[0048] The instruction module generates the coordinated picking task control instruction based on the picking task list combined with the load balancing strategy and the preset maturity-gripping posture parameter table, maturity-gripping force curve and force-pulling speed curve; the coordinated picking task control instruction at least contains the target picking coordinates, the number of picking robots and the corresponding gripping posture, gripping force and pulling speed control parameters of each picking robot.

[0049] The response module responds to the coordinated picking task control instruction to execute the picking operation through the end gripper of the picking robot with the gripping posture, gripping force and pulling speed, and feeds back the picking completion signal to the inspection robot closest to the picking area after picking.

[0050] The feedback confirmation module is used to trigger the secondary imaging confirmation of the picking area, and if there is picking residue, a supplementary picking task list is generated according to the residual area picking density and the corresponding coordinated picking task control instruction is repeatedly responded.

[0051] Compared with the prior art, the beneficial effects of the present application are:

[0052] The present application aims at the deficiencies of the prior art, and realizes accurate determination and quantitative calculation of maturity and regional picking density of edible fungi by multi-dimensional environment perception data collected by the inspection robot, generates a refined task list containing spatial coordinates, maturity level and distribution density; then based on the load balancing strategy and the mature picking parameter mapping model, dynamically generates a cooperative control instruction to accurately guide multiple picking robots to perform the work with the optimal clamping posture, force and speed, significantly improves the picking efficiency and greatly reduces the mechanical damage; finally, through the immediate feedback and secondary imaging confirmation mechanism after picking, a closed-loop control loop of execution, verification and secondary picking is constructed, effectively eliminating the missed picking residual problem, and overall realizing the whole process automation, precision and high reliability of the edible fungi picking operation from perception, decision-making to execution. BRIEF DESCRIPTION OF DRAWINGS

[0053] Figure 1 A flow chart of the robot cooperative operation control method in the edible fungi planting process of the present application;

[0054] Figure 2 A module diagram of the robot cooperative operation control system in the edible fungi planting process of the present application. DETAILED DESCRIPTION

[0055] Example 1

[0056] Please refer to Figure 1 The present application provides an embodiment: a robot cooperative operation control method in the edible fungi planting process, comprising the following steps:

[0057] S1, the inspection robot scans the mushroom bed along the preset initial path and collects the environment perception data of edible fungi; it needs to be further explained that the preset initial path is obtained by dividing according to the current number of inspection robots and the space size of the area to be inspected combined with grid algorithm and simulation algorithm by the person skilled in the art, taking the maximum picking efficiency as the goal; wherein the picking efficiency is constructed by the ratio of the number of real-time picking, the number of secondary confirmation omissions and the time length of picking task completion;

[0058] S2, based on the environment perception data of edible fungi combined with image recognition algorithm, the maturity of at least one edible fungus is determined and the regional picking density is calculated, and a picking task list containing the coordinates of the picking area, the maturity level and the regional picking density is generated; the environment perception data of edible fungi at least contains the image and three-dimensional space information of one edible fungus;

[0059] S3, generating a coordinated picking task control instruction based on the to-be-picked task list, a load balancing strategy, and a preset maturity-gripping posture parameter table, a maturity-gripping force curve, and a force-pulling speed curve; the coordinated picking task control instruction at least includes a target picking coordinate, a number of picking robots, and a corresponding gripping posture, gripping force, and pulling speed control parameter of each picking robot;

[0060] S4, in response to the coordinated picking task control instruction, performing a picking operation by an end gripper of a picking robot at the gripping posture, gripping force, and pulling speed, feeding back a picking completion signal to a nearest inspection robot from a picking area after picking is completed, and triggering secondary imaging confirmation of the picking area. If there is picking residue, a supplementary picking task list is generated according to the to-be-picked density of the residue area, and a corresponding coordinated picking task control instruction is repeatedly responded.

[0061] It needs to be further explained that the image and three-dimensional spatial information of the collected edible fungi in the embodiment include:

[0062] S101, continuously collecting first environment perception data in a preset initial path coverage range and second environment perception data outside the preset initial path coverage range and within a collection visual angle range by using a configured first inspection robot during travel in a preset world coordinate system along a preset initial path; the core reason for collecting the second environment perception data in the embodiment is that the preset initial path can only cover part of the mushroom house area, and there is a mushroom body area outside the path but within the visual angle of the inspection robot. If only the first environment perception data within the path is collected, a perception blind area will be formed, the mushroom body growth information in these areas will be missed, and the overall distribution and growth state of the mushroom bodies in the mushroom house cannot be fully grasped; its role is to expand the environmental perception range, include the areas within the visual angle outside the path into data collection, provide data support for subsequent identification of the second to-be-evaluated mushroom body area outside the path through a lightweight convolutional neural network, and at the same time, the data can be shared with the second inspection robot corresponding to the preset initial path of the area, supplement the perception data of the second inspection robot, and perfect the third environment perception data of the second inspection robot; the final effect is to avoid missing the mushroom body information outside the path, ensure that the mushroom bodies within the visual angle range in the mushroom house can be perceived and evaluated, reduce the problems of mature mushroom body missed picking and delayed picking caused by incomplete perception, at the same time, lay a data foundation for cross-robot collaborative perception, improve the comprehensiveness and timeliness of overall mushroom body monitoring in the mushroom house, and ensure the integrity and accuracy of subsequent to-be-picked task list construction.

[0063] It needs to be further explained that the process of constructing the world coordinate system in the embodiment includes:

[0064] The origin is selected as the intersection of the lines connecting the mushroom house floor and two vertical walls. A three-dimensional right-handed coordinate system is formed, with the vertical line pointing to the top of the mushroom house as the Z-axis, the parallel line along the long side of the mushroom house as the X-axis, and the Y-axis determined using the right-hand screw rule. At least three non-collinear static reference objects are deployed within the mushroom house to cover the entire work area. The geometric center coordinates of each reference object are measured multiple times using a total station or high-precision differential GPS, and the average value is calculated and stored in the central control unit. During robot power-on initialization, at least two reference objects are identified using a visual sensor or LiDAR. The pose relationship between the sensor and the world coordinate system is calculated using the PnP algorithm or triangulation method. Combined with the pre-calibrated installation coordinates of the sensor on the robot body, the initial pose of the robot is obtained. During operation, the robot calculates the pose increment by scanning a fixed feature matching map using laser SLAM or visual inertial odometry, accumulating the increments to obtain the real-time pose. Reference objects are periodically re-identified for correction, and adjustments are made when deviations exceed limits to maintain consistency with the world coordinate system. After the sensor senses the mushroom body, it first calculates the coordinates of the mushroom body in the robot's coordinate system based on the sensor's installation coordinates, and then converts them into three-dimensional coordinates in the world coordinate system based on the real-time pose and uploads them. The central control unit integrates all robot poses, mushroom body world coordinates and fixed environmental features of the mushroom house to build and update a global map containing the mushroom house environment, robot and mushroom body distribution in real time.

[0065] S102. Based on the second environmental perception data, the location of the second mushroom body region to be evaluated in the second environmental perception data is identified in real time through a lightweight convolutional neural network. At the same time, based on the location of the second mushroom body region to be evaluated and combined with an information sharing algorithm, the second inspection robot corresponding to the preset initial path to which the location of the second mushroom body region to be evaluated belongs is queried. It should be further noted that this embodiment includes at least a first inspection robot and a second inspection robot.

[0066] It should be further explained that, in this embodiment, the second inspection robot corresponding to the preset initial path to which the location of the second mushroom body to be evaluated belongs includes:

[0067] S1021. Based on the location of the second mushroom body region to be evaluated and the established world coordinate system, obtain the three-dimensional coordinates of the second mushroom body region to be evaluated in the world coordinate system, and denot them as the target coordinates P.

[0068] S1022. Retrieve the stored global path information in the world coordinate system. The global path information includes at least the preset initial path identifiers of all inspection robots and their path point sequences in the world coordinate system. The path data includes the inspection robot identifiers corresponding to each preset initial path and the three-dimensional coordinate set of that path.

[0069] S1023. For each preset initial path, calculate the coverage area of ​​the path in three-dimensional space based on the corresponding path point sequence; the coverage area is a three-dimensional strip-shaped area formed by extending a preset working width to both sides of the path centerline as a reference; the preset working width is set according to half of the effective detection width of the sensor carried by the corresponding inspection robot.

[0070] S1024. The spatial position relationship between the target coordinate P and all calculated path coverage areas is determined, and the path coverage areas whose spatial range includes the target coordinate P are selected. It should be further explained that the spatial inclusion relationship matching step in this embodiment is achieved by calculating the vertical distance from the target coordinate P to the path centerline and comparing it with the preset working width: if the vertical distance is less than or equal to the working width, it is determined that the coordinate P is located within the coverage area of ​​the path.

[0071] S1025. Based on the association between the preset initial path and the inspection robot in the global map, retrieve the inspection robot identifier corresponding to the filtered preset initial path. The inspection robot corresponding to this identifier is the second inspection robot corresponding to the preset initial path to which the second mushroom body area to be evaluated belongs.

[0072] S103. Based on the second environmental perception data, calculate the initial maturity discrimination score of the second mushroom body region to be evaluated; the initial maturity discrimination score is obtained based on the estimated density of mushrooms in the second mushroom body region to be evaluated, the Euclidean distance between the second mushroom body region to be evaluated and the current position of the inspection robot, and the edible fungus growth status predicted by the preset maturity growth prediction model combined with the evaluation by the expert evaluation system; it should be further explained that the construction and training process of the maturity growth prediction model in this embodiment includes:

[0073] S1031. Based on the historical database of the central control unit, the historical monitoring data set of a specific mushroom bed area within multiple consecutive inspection cycles is obtained through time-series data extraction operations; the historical monitoring data set includes at least the average maturity level of the mushrooms at each inspection time point and the corresponding environmental parameters, including environmental temperature, environmental humidity and carbon dioxide concentration.

[0074] S1032. Based on the historical monitoring data set, a standardized time-series dataset is obtained through a data preprocessing process. The data preprocessing process includes: filling in missing data using time-series linear interpolation, converting maturity levels into numerical values ​​using one-hot encoding, and aligning and normalizing based on timestamps to achieve a unified numerical scale between environmental data and maturity data.

[0075] S1033. Based on the standardized time series dataset, obtain the model training sample set through the supervised learning sample construction method; the supervised learning sample construction method refers to: taking the complete data sequence of the previous one or more consecutive inspection cycles as input features, taking the actual maturity level of the future specified prediction cycle as prediction label, and constructing a set of input-output sample pairs with temporal correlation.

[0076] S1034. Based on the training sample set of the model, a pre-trained maturity growth prediction model is obtained through the long short-term memory network training process; the training process includes: dividing the sample set into a training set and a validation set in chronological order, using mean squared error as the loss function to optimize the model parameters, monitoring the training process through the validation set, and terminating the training when the model loss converges.

[0077] S1035. Based on the pre-trained maturity growth prediction model, a continuously optimized prediction system is obtained through an online update mechanism; the online update mechanism refers to: periodically collecting new inspection data during system operation, and incrementally training the model based on the new data within the most recent time window, so that the model continuously adapts to the dynamic changes in the growth characteristics of edible fungi.

[0078] S104: When the initial maturity discrimination score of at least one second mushroom body region to be evaluated exceeds a set threshold, the first inspection robot, based on the query results from the second inspection robot, shares the second environmental perception data and the initial maturity discrimination score with the second inspection robot through an information sharing algorithm. Combined with the first environmental perception data collected in real time by the second inspection robot, the first inspection robot constructs the third environmental perception data for the second inspection robot. It should be further noted that the set threshold in this embodiment is specifically set by those skilled in the art based on actual scenario parameters; the information sharing algorithm in this embodiment is preferably the Raft algorithm.

[0079] This process achieves a unified spatial reference for multi-robot collaborative operations by constructing a precise world coordinate system and global map. During inspection, the system not only collects environmental data within the preset initial path range but also intelligently identifies potential areas to be evaluated outside the path. It then uses a lightweight convolutional neural network for real-time localization and matches the corresponding responsible robot based on spatial relationships, enabling dynamic optimization of inspection resource allocation. In particular, when high-maturity areas are detected, key data is transmitted to the corresponding inspection robot through an information sharing mechanism, forming complete third-party environmental perception data. This cross-robot data fusion and task collaboration mechanism effectively expands the perception range of a single path, significantly improving the system's global perception capability and response efficiency regarding the maturity status of edible fungi.

[0080] It should be further explained that the process of obtaining the list of tasks to be picked in this embodiment includes:

[0081] S201, arranging in time stamp order based on the third environment perception data of any inspection robot, obtaining a sequence of images to be identified, wherein each frame of image is associated with three-dimensional space information mark of corresponding time stamp;

[0082] S202, performing coordinate mapping, interference suppression, feature repair and inter-frame redundancy removal processing on each frame of image in the sequence of images to be identified, obtaining a sequence of complete inter-frame incremental area images of any inspection robot; the length of the sequence of complete inter-frame incremental area images is the same as the preset initial path length of the corresponding inspection robot; the coordinate mapping, interference suppression, feature repair and inter-frame redundancy removal processing in this embodiment includes:

[0083] S2021, based on the pose data of the inspection robot corresponding to each frame of image in the sequence of images to be identified, the image shooting focal length and the lens distortion parameters, constructing a coordinate mapping matrix to obtain the mapping relationship between the pixel coordinate system and the world coordinate system; based on the mapping relationship, the three-dimensional space information of the edible fungus is converted into the corresponding pixel coordinates through coordinate inversion calculation, and the pixel coordinates obtained by inversion are compared and analyzed with the mushroom pixel position in the actual image to obtain the image sequence after coordinate mapping deviation correction;

[0084] It needs to be further explained that the process of comparing and analyzing the pixel coordinates obtained by inversion with the mushroom pixel position in the actual image in this embodiment includes:

[0085] Step 1, based on the time stamp information of each frame of image in the sequence of images to be identified, the six-degree-of-freedom pose data of the inspection robot in the world coordinate system at the image acquisition time is obtained by querying the historical record data of the laser SLAM system or visual inertial odometry system carried by the inspection robot; the pose data includes X-axis position parameter, Y-axis position parameter, Z-axis position parameter and rotation angle around X-axis, rotation angle around Y-axis, rotation angle around Z-axis;

[0086] Step 2, based on the obtained pose data, image shooting focal length parameter and lens distortion parameter, the initial mapping matrix between the pixel coordinate system and the world coordinate system of each frame of image is obtained through coordinate transformation matrix construction algorithm; the shooting focal length parameter is obtained from the wide-angle camera parameter library of the inspection robot; the coordinate transformation matrix construction algorithm specifically includes: obtaining the camera intrinsic matrix and lens distortion parameter through Zhang Zhengyou camera calibration method, obtaining the robot extrinsic matrix through pose data calculation, and performing matrix multiplication operation on the intrinsic matrix and extrinsic matrix, lens distortion parameter to obtain the initial mapping matrix. The lens distortion parameter includes radial distortion coefficient and tangential distortion coefficient, and the lens distortion parameter is obtained from the camera calibration data; the robot extrinsic matrix includes rotation matrix and translation matrix;

[0087] Step 3, based on the initial mapping matrix and the three-dimensional space coordinates of the edible fungi in the world coordinate system, the inverse pixel coordinates of the edible fungi in the image are calculated through the perspective projection transformation algorithm; the specific implementation of the perspective projection transformation algorithm includes: converting the three-dimensional coordinates in the world coordinate system to the camera coordinate system through rigid body transformation, and then converting to the image coordinate system through perspective projection transformation, and finally obtaining the final pixel coordinates through distortion correction processing.

[0088] Step 4, based on the inverse pixel coordinates and the actual pixel coordinates of the mushroom body obtained through the image recognition algorithm, the coordinate deviation value of the edible fungi in each frame of image is obtained through the Euclidean distance calculation; the actual pixel position data of the mushroom body is obtained by recognizing the mushroom body edge through the image target detection algorithm; the deviation calculation of the embodiment is used to calculate the Euclidean distance between the inverse pixel coordinates and the actual pixel coordinates to construct; the image recognition algorithm of the embodiment uses Canny edge detection algorithm combined with Hough transformation to accurately identify the actual pixel position of the mushroom body in the image.

[0089] Step 5, based on the comparison result of the coordinate deviation value and the preset deviation threshold, the pose parameters and distortion coefficients are iteratively adjusted through the Levenberg-Marquardt optimization algorithm until the coordinate deviation value meets the preset requirement, and the optimized mapping matrix is obtained; the Levenberg-Marquardt optimization algorithm of the embodiment balances the advantages of gradient descent method and Gauss-Newton method by dynamically adjusting the damping factor, so as to realize the rapid convergence of the mapping matrix parameters.

[0090] Step 6, based on the optimized mapping matrix, the three-dimensional space information of the edible fungi in each frame of image is re-marked through the coordinate re-projection algorithm, and the single frame image after coordinate mapping deviation correction is obtained; the coordinate re-marking algorithm is used to mark the three-dimensional space information of the edible fungi in the world coordinate system to the corresponding pixel position according to the corrected mapping matrix;

[0091] Step 7, repeat the operations of steps 1 to 6 for each frame of image to obtain the complete image sequence after coordinate mapping deviation correction.

[0092] S2022, based on the image sequence after coordinate mapping deviation correction, the feature enhancement processing is performed on the mushroom body region through the super-resolution reconstruction algorithm, and the color parameters of the distortion region are adjusted by using the color features of the same mushroom body in the normal frame as the reference; based on the image sequence after feature repair, the cross-frame feature similarity calculation and verification are performed to obtain the to-be-identified image sequence which meets the preset similarity threshold after local feature repair is completed;

[0093] It needs to be further explained that in the embodiment, the process of adjusting the color of the distorted region by taking the color of the same mushroom body normal frame as the reference by using the color migration algorithm, and checking the cross-frame feature similarity to the preset similarity threshold value, includes:

[0094] Step 1, based on each frame of image in the corrected to-be-identified image sequence based on coordinate mapping deviation, the mushroom body region image data in each frame of image is obtained by a semantic segmentation algorithm; the semantic segmentation algorithm uses a DeepLabv3+ model to identify and extract the mushroom body region in the image, and excludes the background region;

[0095] Step 2, based on the obtained mushroom body region image data, the image data of the mushroom body region after fine feature enhancement is obtained by a super-resolution reconstruction algorithm; the super-resolution reconstruction algorithm uses an SRGAN model to generate a high-resolution mushroom body image according to the input mushroom body region image;

[0096] Step 3, based on the obtained fine feature enhanced image data and the original frame image, a single frame image after fine feature enhancement is obtained by an image fusion algorithm; the image fusion algorithm uses a Laplacian pyramid fusion method to fuse the enhanced mushroom body features and the original frame image;

[0097] Step 4, based on the obtained single frame image after fine feature enhancement and the normal frame image of the same mushroom body in the to-be-identified image sequence, the mushroom body reference color distribution parameters are obtained by a color histogram statistical algorithm; the color histogram statistical algorithm calculates the RGB three-channel color distribution histogram of the same mushroom body region in the normal frame;

[0098] Step 5, based on the obtained mushroom body reference color distribution parameters and the single frame image after fine feature enhancement, the mushroom body color distortion region in the single frame image is obtained by a color difference detection algorithm; the color difference detection algorithm uses a CIEDE2000 color difference formula to calculate the color deviation;

[0099] Step 6, based on the obtained mushroom body color distortion region and the mushroom body reference color distribution parameters, the adjusted RGB color value data of the color distortion region is obtained by a Reinhard color migration algorithm; the Reinhard color migration algorithm realizes color migration through LAB color space conversion and statistical feature matching;

[0100] Step 7, based on the obtained adjusted RGB color value data, the single frame image after color repair is obtained by a pixel replacement algorithm; the pixel replacement algorithm replaces the original pixel value of the color distortion region with the migrated pixel value;

[0101] Step 8, based on the obtained single frame image after color repair and other frame images of the same mushroom body in the image sequence to be recognized, the cross-frame feature similarity value of the same mushroom body is obtained through a feature similarity calculation algorithm; the feature similarity calculation algorithm adopts SIFT feature extraction and cosine similarity calculation;

[0102] Step 9, based on the obtained cross-frame feature similarity value and the preset feature similarity threshold, if the similarity value is lower than the preset feature similarity threshold, steps 2 to 8 are re-executed, the hyperparameters of SRGAN and the matching parameters of Reinhard algorithm are adjusted, until the similarity value reaches the preset threshold; the above steps are repeatedly executed for all frames in the image sequence to obtain the image sequence after local feature repair.

[0103] S2023, based on the screened interference frame, the adjacent normal frames before and after the interference frame are used as reference frames, a weighted average interpolation algorithm is used to calculate the pixel value of the blurred area in the interference frame, the blurred pixels in the interference frame are repaired, and the repaired interference frame is processed through color feature normalization, the RGB channel value of the interference frame is adjusted based on the color mean value of the reference frame, and an image sequence after dynamic interference suppression is obtained;

[0104] It needs to be further explained that in the embodiment, the interference frame is screened by calculating the gray level fluctuation value between adjacent frames, the adjacent normal frames before and after the interference frame are used as reference frames, the blurred pixels in the interference frame are repaired by using a weighted average interpolation, and the color feature is normalized to be consistent with the reference frame. The process includes:

[0105] Step 1, based on each frame image in the image sequence to be recognized after local feature repair, the gray value data of each frame image is obtained through a gray conversion algorithm, and the RGB three-channel image is converted into a single-channel gray image;

[0106] Step 2, based on the gray value data of the adjacent two frame images, the inter-frame gray level fluctuation value of the adjacent two frame images is obtained through a mean absolute difference algorithm; the mean absolute difference algorithm calculates the absolute difference value of the gray value of the corresponding pixel position of the adjacent two frame images, and the arithmetic mean value of all pixel positions is taken as the inter-frame gray level fluctuation value;

[0107] Step 3, based on the inter-frame gray level fluctuation value and the preset gray level fluctuation threshold, a dynamic interference frame set in the image sequence to be recognized is obtained; wherein, if the inter-frame gray level fluctuation value is greater than the preset gray level fluctuation threshold, the next frame is determined as a dynamic interference frame;

[0108] Step 4, based on each dynamic interference frame in the dynamic interference frame set, the previous normal frame and the next normal frame of the interference frame are obtained as reference frames through a time sequence index algorithm; the normal frame is a frame with an inter-frame gray level fluctuation value less than or equal to the preset gray level fluctuation threshold;

[0109] Step 5, based on the interference frame and the front and rear reference normal frames, the repair pixel value data of the interference frame blur area is obtained through a bidirectional linear interpolation algorithm; the bidirectional linear interpolation algorithm calculates the weighted average of the gray values of the corresponding pixels of the front and rear reference normal frames, and the weight is dynamically adjusted according to the time interval;

[0110] Step 6, based on the repair pixel value data, the preliminary repaired dynamic interference frame is obtained through a pixel covering algorithm; the pixel covering algorithm replaces the original pixel value in the corresponding position of the interference frame with the pixel value obtained by interpolation;

[0111] Step 7, based on the preliminary repaired dynamic interference frame and the front and rear reference normal frames, the reference color mean value data is obtained through a mean value calculation algorithm; the mean value calculation algorithm calculates the arithmetic mean value of the R, G and B three channel pixel values of the front and rear reference normal frames respectively;

[0112] Step 8, based on the reference color mean value data and the preliminary repaired dynamic interference frame, the repaired interference frame after color feature calibration is obtained through a histogram matching algorithm; the histogram matching algorithm adjusts the color distribution of the interference frame to make it consistent with the color distribution feature of the reference frame;

[0113] Step 9, based on all dynamic interference frames in the local feature repaired to-be-recognized image sequence, the operation of steps 4 to 8 is repeatedly performed on each interference frame through a batch processing algorithm, and a dynamic interference suppressed to-be-recognized image sequence is obtained.

[0114] S2024, based on the dynamic interference suppressed image sequence, adjacent two frames of images are extracted in time stamp order, and an ORB feature matching algorithm is used to detect the overlapping area between adjacent frames; based on the overlapping area detection result, the blob feature and the corresponding three-dimensional space information mark result of the overlapping area in the previous frame image are retained, only the newly added area in the next frame image which is not overlapped is marked, and an inter-frame incremental area image set is obtained; based on the inter-frame incremental area image set, a complete inter-frame incremental area image sequence of any inspection robot is obtained through time sequence integration processing;

[0115] It needs to be further explained that the acquisition process of the inter-frame incremental area image set of any inspection robot in the embodiment includes:

[0116] Step 1, based on the time stamp order of the dynamic interference suppressed image sequence, adjacent two frames of images are obtained in time stamp increasing order through a time sequence traversal algorithm;

[0117] Step 2, based on the adjacent two frames of images, the ORB feature set of the front frame image and the ORB feature set of the rear frame image are obtained through an ORB feature extraction algorithm; the ORB feature extraction algorithm adopts oFAST key point detection and rBRIEF descriptor generation;

[0118] Step 3, based on the ORB feature set of the previous frame image and the ORB feature set of the next frame image, the overlapping area range data of the adjacent two frame images and the matching result data of the mushroom features in the overlapping area are obtained by the brute force matching algorithm; the brute force matching algorithm calculates the feature similarity by Hamming distance, and removes the false matching by distance ratio test;

[0119] Step 4, based on the overlapping area range data and the mushroom feature matching result data, the inter-frame incremental region image is obtained by the feature point screening and region division algorithm; the algorithm retains the matched feature points in the previous frame image and their corresponding world coordinate system three-dimensional space information, and only calculates and marks the world coordinate system three-dimensional space information of the new feature points in the next frame image which are not matched;

[0120] Step 5, based on the time stamps of the image sequence after dynamic interference suppression, the operation of steps 1 to 4 is repeatedly performed on all adjacent two frame image pairs in the sequence by the iterative processing algorithm, and the image set after frame redundancy removal is obtained.

[0121] S203, based on the complete inter-frame incremental region image sequence of any inspection robot, the standardized mushroom region image is obtained by image size standardization processing; the image size standardization processing process in this embodiment first obtains the width and height parameters of the original image, then calculates the scaling ratio according to the preset target resolution, and the ratio takes the smaller value of the original width-height ratio to ensure that the image is fully adapted; then the image is scaled using the bilinear interpolation algorithm, and the weighted average value of the four adjacent pixels around the interpolation point is calculated to maintain the image quality; finally, the edge zero value of the scaled image is filled, and the image size is unified to the target resolution, so as to ensure the consistency of the input size of the subsequent image processing algorithm.

[0122] S204, based on the standardized mushroom region image, the improved YOLOv8 algorithm is used for mushroom target detection to obtain the bounding box coordinates of each edible fungus; the bounding box coordinates include the pixel coordinates of the upper left corner and the lower right corner of the bounding box.

[0123] It needs to be further explained that the specific process of the improved YOLOv8 algorithm for mushroom target detection in this embodiment includes:

[0124] Step 1, based on the standardized mushroom region image, the normalized image is obtained by pixel value normalization processing; the pixel value normalization processing is to divide the RGB three-channel pixel values of the image by 255 respectively, so that the pixel value range is mapped to the interval of 0-1.

[0125] Step 2, based on the normalized image, a multi-scale basic feature map is obtained through a CSPDarknet backbone network; the CSPDarknet backbone network comprises 5 convolution modules and 4 CSP modules, and the multi-scale basic feature map comprises an 8x8 scale feature map, a 16x16 scale feature map and a 32x32 scale feature map, which correspond to different sizes of mushroom target features in the image respectively.

[0126] Step 3, based on the multi-scale basic feature map, an enhanced multi-scale feature map is obtained through a mushroom feature enhancement module, specifically:

[0127] Based on the multi-scale feature map extracted by the backbone network, the cross-scale feature fusion unit of the path aggregation network structure is used for processing, the cross-scale feature fusion unit adopts a bidirectional feature propagation path from top to bottom and from bottom to top, and through feature map splicing and 1x1 convolution operation, the 8x8 scale feature map and the 16x16 scale feature map, and the 16x16 scale feature map and the 32x32 scale feature map are fused pixel by pixel to obtain a fused multi-scale feature map; the feature propagation path from top to bottom realizes the fusion of shallow detail features to deep semantic features through up-sampling and feature splicing, and the feature propagation path from bottom to top realizes the fusion of deep semantic features to shallow detail features through down-sampling and feature splicing.

[0128] Based on the fused multi-scale feature map, a multi-scale receptive field feature is extracted through a dilated spatial pyramid pooling module, a plurality of dilated convolutions with different sampling rates are used for parallel processing to obtain an enhanced feature map containing multi-scale context information; in this embodiment, the process of extracting a multi-scale receptive field feature through the dilated spatial pyramid pooling module includes: based on the input feature map, four parallel branches are used for processing, 3x3 dilated convolution kernels with dilated rates of 1, 2, 4 and 8 are used for feature extraction respectively, and a global average pooling branch is reserved at the same time; the feature maps output by the five branches are spliced in the channel dimension, and then 1x1 convolution is used for feature fusion and dimension reduction, and finally an enhanced feature map with a multi-scale receptive field is output. The module obtains feature information with different receptive fields through convolution kernels with different dilated rates, wherein the dilated rate 1 corresponds to a standard convolution receptive field, the dilated rates 2, 4 and 8 obtain an equivalent receptive field of about 7x7, 15x15 and 31x31 respectively, the global average pooling branch provides image-level global context information, and finally the different scale features of the mushroom are completely captured through feature fusion.

[0129] Based on the enhanced feature map, an edge texture enhancement unit is used for processing, the edge texture enhancement unit adopts a multi-operator edge detection layer with a parallel convolution structure, and uses a Sobel operator to extract a first-order gradient feature, a Laplacian operator to extract a second-order differential feature, and a local binary pattern operator to extract a texture structure feature; the Sobel operator uses a 3*3 convolution kernel to calculate the gradient features in the horizontal and vertical directions, the Laplacian operator uses a second-order differential template to extract the curvature features of the mushroom surface, and the local binary pattern operator extracts the local texture pattern by comparing the gray value relationship between the center pixel and the neighborhood pixels.

[0130] Based on the extracted edge texture features, an edge-aware filter is used for edge-aware smoothing processing, which can suppress background noise while retaining the edge details of the mushroom, and obtain an enhanced edge texture feature map; the edge-aware filter uses a guide image to perform edge-preserving smoothing processing on the input feature map, wherein the guide image is the result of Gaussian filtering on the original input image.

[0131] The enhanced edge texture feature map and the fused multi-scale feature map are spliced in the channel dimension, and feature dimension reduction is performed through 1*1 convolution, and finally an enhanced multi-scale feature map is output. The feature dimension reduction operation reduces the number of channels of the spliced feature map to a preset dimension through 1*1 convolution, while keeping the spatial size of the feature map unchanged.

[0132] Step 4, based on the enhanced multi-scale feature map, a PAN-FPN neck network is used to obtain a fused detection feature map; the PAN-FPN neck network includes an up-sampling module, a down-sampling module and a convolution fusion module, which fuses 8*8 scale feature maps and 16*16 scale feature maps through up-sampling, fuses the fusion results and 32*32 scale feature maps through down-sampling, and finally outputs 8*8, 16*16 and 32*32 scale detection feature maps.

[0133] Step 5, based on the fused detection feature map, a YOLOv8 detection head is used to obtain the bounding box coordinates of each edible mushroom; the YOLOv8 detection head includes 3 prediction branches corresponding to 8*8, 16*16 and 32*32 scale detection feature maps respectively, each prediction branch outputs target prediction information through a 1*1 convolution layer, and the target prediction information includes the x-coordinate of the upper left corner of the bounding box, the y-coordinate of the upper left corner, the x-coordinate of the lower right corner, the y-coordinate of the lower right corner, the mushroom class confidence and the target existence confidence; the prediction information is filtered through a non-maximum suppression algorithm, and the prediction results with a confidence higher than a preset threshold are retained, and finally the bounding box coordinates of each edible mushroom are obtained.

[0134] S205, based on the bounding box coordinates of each edible mushroom, a single mushroom region image is obtained through a region segmentation algorithm, and based on the single mushroom region image, an edge contour data of the single mushroom is obtained through an edge detection algorithm; it needs to be further explained that the process of obtaining the edge contour data of the single mushroom in the embodiment includes:

[0135] Step 1, based on the bounding box coordinates of each edible mushroom and the standardized mushroom region image, an effective bounding box coordinate is obtained through coordinate boundary verification processing; the bounding box coordinates include the upper left corner x coordinate, the upper left corner y coordinate, the right lower corner x coordinate and the right lower corner y coordinate, the standardized mushroom region image has an image width and an image height, and the coordinate boundary verification processing is: if the upper left corner x coordinate is less than zero, it is adjusted to zero; if the upper left corner y coordinate is less than zero, it is adjusted to zero; if the right lower corner x coordinate is greater than the image width, it is adjusted to the image width; if the right lower corner y coordinate is greater than the image height, it is adjusted to the image height, so as to ensure that the bounding box coordinates are within the image size range.

[0136] Step 2, based on the effective bounding box coordinates, a single mushroom region image is obtained through pixel region extraction processing; the pixel region extraction processing is: from the standardized mushroom region image, all pixel points with the horizontal coordinate range of the upper left corner x coordinate to the right lower corner x coordinate and the vertical coordinate range of the upper left corner y coordinate to the right lower corner y coordinate are extracted, and these pixel points are arranged according to the original position to form a new image, which is the single mushroom region image, and the width of the new image is the difference between the right lower corner x coordinate and the upper left corner x coordinate, and the height of the new image is the difference between the right lower corner y coordinate and the upper left corner y coordinate.

[0137] Step 3, based on the single mushroom region image, a denoised single mushroom region image is obtained through Gaussian filtering processing; the Gaussian filtering processing adopts Gaussian kernel for convolution operation, the Gaussian kernel is a three-by-three size matrix, and the matrix element value is calculated by Gaussian function, which is used to smooth the image and reduce the interference of noise on edge detection.

[0138] Step 4, based on the denoised single mushroom region image, a gradient amplitude image and a gradient direction image are obtained through Sobel operator calculation processing; the Sobel operator includes x direction operator and y direction operator, both of which are three-by-three size matrices, the x direction operator is used to calculate the gradient value of the pixel point in the x direction, and the y direction operator is used to calculate the gradient value of the pixel point in the y direction; the gradient amplitude is obtained by squaring and square root operation of the sum of x direction gradient value and y direction gradient value, and the gradient direction is obtained by arctangent operation of x direction gradient value and y direction gradient value, the gradient amplitude image is composed of gradient amplitudes of all pixel points, and the gradient direction image is composed of gradient directions of all pixel points.

[0139] Step 5, based on the gradient magnitude image and the gradient direction image, a refined gradient magnitude image is obtained through non-maximum suppression processing; the non-maximum suppression processing is: for each pixel point in the gradient magnitude image, two adjacent pixel points (pixels before and after along the gradient direction) are determined according to the gradient direction of the pixel point, if the gradient magnitude of the pixel point is greater than the gradient magnitudes of the two adjacent pixel points, the gradient magnitude of the pixel point is retained, otherwise the gradient magnitude of the pixel point is set to zero, which is used to eliminate non-edge pixels and refine the edge contour.

[0140] Step 6, based on the refined gradient magnitude image, a strong edge pixel set and a weak edge pixel set are obtained through double-threshold screening processing; the double-threshold screening processing sets a high threshold and a low threshold, if the gradient magnitude of a pixel point is greater than the high threshold, the pixel point is included in the strong edge pixel set, if the gradient magnitude of a pixel point is between the low threshold and the high threshold, the pixel point is included in the weak edge pixel set, and if the gradient magnitude of a pixel point is less than the low threshold, the pixel point is excluded, the high threshold and the low threshold are set according to the gradient magnitude distribution of the single mushroom region image.

[0141] Step 7, based on the strong edge pixel set and the weak edge pixel set, edge contour data of a single mushroom is obtained through edge connection processing; the edge connection processing is: each pixel point in the weak edge pixel set is traversed, if the pixel point is adjacent (up, down, left, right or diagonal direction) to any pixel point in the strong edge pixel set, the pixel point is included in the strong edge pixel set, finally all pixel points in the strong edge pixel set form the edge contour data, the edge contour data is a set of pixel coordinates of the single mushroom edge.

[0142] S206, based on the single mushroom edge contour data, cap contour data and stem contour data are obtained through a contour segmentation algorithm; the contour segmentation algorithm is an algorithm for segmenting the edge contour according to the morphological difference between the cap and the stem, the cap is an arc-shaped contour with a wide top, and the stem is a columnar contour with a narrow bottom.

[0143] S207, based on the cap profile data, obtain the cap diameter parameter, the cap edge curvature parameter, the cap color RGB mean parameter and the gill clarity parameter; in this embodiment, the calculation of the cap diameter parameter first obtains the coordinate set of all edge pixel points based on the cap profile data, then calculates the distance between any two pixel points in the set through the Euclidean distance formula, and finally determines the maximum value in all distances as the cap diameter parameter through traversal comparison. The calculation process of the cap edge curvature parameter in this embodiment includes: extracting the continuous pixel point coordinate subset of the arc-shaped edge from the cap profile data; obtaining the profile function through quadratic curve fitting by least squares method; calculating the curvature value of each pixel point position based on the curvature formula; finally taking the arithmetic mean of all curvatures as the cap edge curvature parameter. The calculation process of the cap color RGB mean parameter in this embodiment includes: first, based on the cap profile data, determine all internal pixel points of the cap region through the boundary filling algorithm to form a cap region pixel set; then traverse each pixel point in the set, extract the values of R, G and B three color channels respectively to form the corresponding channel value set; then perform accumulation summation operation on each channel set respectively to obtain the value sum of each channel; finally, divide the sum of each channel by the total number of pixel points in the cap region to obtain the mean value of R, G and B three channels respectively through arithmetic mean calculation, and the combination of the mean values of the three channels constitutes the RGB mean parameter representing the overall color feature of the cap. The calculation process of the gill clarity parameter in this embodiment includes: first, based on the cap profile data and the single mushroom region image, determine the annular gill texture region in the range of 1 / 3 to 2 / 3 height above the bottom edge of the cap, obtain all pixel points in the region to form a texture region pixel set; then use the gradient calculation method to calculate the horizontal direction brightness difference absolute value of each pixel point in the set with the right adjacent pixel and the vertical direction brightness difference absolute value with the lower adjacent pixel, add the brightness difference values in the two directions to obtain the gradient value of the pixel point, which reflects the degree of local texture change; finally, divide the sum of the gradient values of all pixel points by the total number of pixel points to obtain the gradient mean value as the gill clarity parameter. The calculation process of the stem height parameter in this embodiment includes two key steps: first, based on the stem profile data, identify the bottom endpoint corresponding to the maximum vertical coordinate value and the top endpoint corresponding to the minimum vertical coordinate value among all pixel points on the stem edge through the extreme value search algorithm to obtain the complete coordinate information of the two key points; then calculate the vertical coordinate value of the top endpoint minus the vertical coordinate value of the bottom endpoint to obtain the absolute value as the stem height parameter; this parameter quantifies the extension degree of the stem in the vertical direction, accurately reflects the morphological characteristics of the edible mushroom, and provides an important geometric feature basis for maturity evaluation.

[0144] S208, based on the stem profile data, obtain the stem height parameter through the height calculation algorithm; wherein the height is the vertical distance between the upper and lower endpoints in the stem profile.

[0145] S209, based on the bounding box coordinates of all edible fungi, the total number of detected edible fungi in the preset initial path is obtained by quantity statistics, denoted as the total number of area mushroom bodies;

[0146] S210, based on the initial maturity discrimination score in the current preset initial path and the single mushroom feature parameter set extracted from the corresponding area, the maturity score and the corresponding maturity level of the edible fungi at each position in the preset initial path of any inspection robot are obtained through the maturity growth prediction model and the pre-trained expert evaluation system;

[0147] S211, based on the initial maturity discrimination score in the current preset initial path and the single mushroom feature parameter set extracted from the corresponding area, the maturity score and the corresponding maturity level of the edible fungi at each position in the preset initial path of any inspection robot are obtained through the maturity growth prediction model and the pre-trained expert evaluation system, including the following steps:

[0148] S2111, based on the initial maturity discrimination score in the current preset initial path, the fused maturity reference value is obtained by a weighted fusion algorithm; the weighted fusion algorithm is to weight and calculate the initial maturity discrimination score and the environmental parameters of the corresponding area; the environmental parameters include environmental temperature value, environmental humidity value and carbon dioxide concentration value; the weight of each environmental parameter is preset according to its influence degree on maturity;

[0149] S2112, based on the single mushroom feature parameter set extracted from the corresponding area, the standardized single mushroom feature parameter is obtained by feature standardization processing; the single mushroom feature parameter set includes cap diameter parameter, cap edge curvature parameter, cap color RGB mean value parameter, gill clarity parameter and stem height parameter; the feature standardization processing adopts the min-max normalization algorithm, and each feature parameter value is mapped to the interval of zero to one;

[0150] S2113, based on the standardized single mushroom feature parameter, the principal component feature vector is obtained by principal component analysis algorithm for feature dimension reduction; the principal component analysis algorithm calculates the covariance matrix of the feature parameter, performs eigenvalue decomposition on the covariance matrix, selects the feature vectors corresponding to the first k largest eigenvalues to form a projection matrix, and multiplies the standardized single mushroom feature parameter by the projection matrix to obtain the principal component feature vector;

[0151] S2114, based on the fused maturity reference value and the principal component feature vector, the maturity preliminary prediction value is obtained through the maturity growth prediction model; the maturity growth prediction model is a long short-term memory network model, which takes the principal component feature vector as the input feature and takes the fused maturity reference value as the initial state, and calculates the maturity preliminary prediction value through the gating mechanism of the long short-term memory network;

[0152] S2115, based on the maturity preliminary prediction value, obtaining the maturity score correction amount through the pre-trained expert evaluation system; the expert evaluation system is a fuzzy logic-based inference system, including a knowledge base and a reasoning machine, the knowledge base stores maturity evaluation rules, and the reasoning machine matches the corresponding rules according to the maturity preliminary prediction value to output the maturity score correction amount;

[0153] S2116, based on the maturity preliminary prediction value and the maturity score correction amount, obtaining the final maturity score of each position edible fungus through a linear superposition algorithm; the linear superposition algorithm adds the maturity preliminary prediction value and the maturity score correction amount to obtain the final maturity score in the range of zero to one hundred;

[0154] S2117, based on the final maturity score, obtaining the maturity grade corresponding to each position edible fungus through a threshold comparison algorithm; the threshold comparison algorithm sets a first threshold and a second threshold, if the final maturity score is greater than or equal to the first threshold, it is determined as a picking grade, if the final maturity score is greater than or equal to the second threshold and less than the first threshold, it is determined as a maturity grade, and if the final maturity score is less than the second threshold, it is determined as an immature grade.

[0155] S212, based on the maturity grade of each position edible fungus in the initial path preset by any inspection robot, combining the corresponding single mushroom characteristic parameters, taking the ratio of single mushroom characteristic parameter similarity and maturity score deviation as the clustering density parameter, and performing regional clustering division on the initial path preset by the adaptive clustering algorithm, to obtain the edible fungus evaluation clustering set on the current initial path; the single mushroom characteristic parameters in this embodiment include cap diameter parameter, cap edge curvature parameter, cap color RGB mean value parameter, gill clarity parameter, and stem height parameter;

[0156] It needs to be further explained that the process of performing regional clustering division on the initial path preset by the adaptive clustering algorithm in this embodiment includes:

[0157] S2121, based on the single mushroom characteristic parameters of each position edible fungus in the initial path preset, constructing a feature data matrix; the rows of the feature data matrix correspond to each edible fungus sample, and the columns correspond to five characteristic parameter dimensions;

[0158] S2122, based on the feature data matrix, obtaining the feature similarity value between any two edible fungus samples through the Euclidean distance algorithm; the Euclidean distance algorithm is used to calculate the straight line distance of two samples in a five-dimensional feature space, and the smaller the distance value, the higher the feature similarity;

[0159] S2123, based on the maturity score of each position mushroom, the maturity score deviation value between any two mushroom samples is obtained by an absolute difference calculation algorithm; the absolute difference calculation algorithm calculates the absolute difference of the maturity score of two samples;

[0160] S2124, based on the feature similarity matrix and the maturity score deviation matrix, a clustering density parameter matrix is obtained by a ratio calculation algorithm; the ratio calculation algorithm divides each feature similarity value by its corresponding maturity score deviation value, and when the maturity score deviation value is zero, the deviation value is replaced by a preset non-zero minimum value;

[0161] S2125, based on the clustering density parameter matrix, an initial clustering center set is obtained by a density peak search algorithm; the density peak search algorithm calculates the local density value and the relative distance value of each sample, and selects the first k samples with the maximum product of local density value and relative distance value as the initial clustering center;

[0162] S2126, based on the initial clustering center set and the clustering density parameter matrix, a mushroom evaluation clustering cluster set is obtained by an improved k-means clustering algorithm; the improved k-means clustering algorithm takes the clustering density parameter as the distance weight between samples, and recalculates the clustering center in each iteration until the clustering center variation is less than a preset threshold;

[0163] S2127, based on the obtained mushroom evaluation clustering cluster set, the validity of the clustering result is verified by a silhouette coefficient calculation algorithm; the silhouette coefficient calculation algorithm calculates the silhouette coefficient value of each sample, and takes the average value of the silhouette coefficients of all samples as the clustering validity index; when the average value is greater than a preset validity threshold, the final mushroom evaluation clustering cluster set is output.

[0164] S213, based on the mushroom evaluation clustering cluster set on the current preset initial path and the preset picking judgment threshold, the to-be-picked mushroom evaluation clustering cluster set and the to-be-mature mushroom evaluation clustering cluster set are obtained; it needs to be further explained that the to-be-picked judgment threshold in the embodiment is set by a person skilled in the art;

[0165] S214, based on the ratio of the corresponding maturity score in each to-be-picked mushroom evaluation clustering cluster, the number of to-be-picked mushroom evaluation clustering clusters, and the total number of area mushroom bodies, and the preset mushroom growth curve combined with an expert evaluation algorithm, the to-be-picked priority of each to-be-picked mushroom evaluation clustering cluster is obtained.

[0166] It needs to be further explained that the process of obtaining the to-be-picked priority of each to-be-picked mushroom evaluation clustering cluster in the embodiment includes:

[0167] S2141, based on the to-be-picked edible mushroom evaluation cluster All edible mushroom samples in the maturity score, the average maturity score of the cluster is calculated by weighted average algorithm; based on the average maturity score, query the preset edible mushroom growth curve, get the corresponding remaining picking time parameter; based on the remaining picking time parameter, calculate the maturity urgency coefficient through the inverse operation;

[0168] S2142, based on the to-be-picked edible mushroom evaluation cluster Edible mushroom quantity and regional mushroom total count, get yield proportion parameter through ratio calculation algorithm; logarithmic normalization is performed on the yield proportion parameter to obtain yield weight coefficient;

[0169] S2143, the maturity urgency coefficient and yield weight coefficient are combined into a two-dimensional evaluation feature vector, which is used as the input of the expert evaluation system;

[0170] S2144, based on the evaluation feature vector, calculate the initial priority score through the fuzzy logic reasoning system; the fuzzy logic reasoning system includes the membership function of the maturity urgency coefficient and the membership function of the yield weight coefficient, and the initial priority score is obtained by reasoning through the fuzzy rule base and using the barycentric method to defuzzify;

[0171] S2145, based on the remaining picking time parameter, the initial priority score is corrected through the exponential decay model; the exponential decay model takes the remaining picking time parameter as the independent variable, calculates the time decay factor, and multiplies the initial priority score with the time decay factor to obtain the corrected priority score;

[0172] S2146, the corrected priority score of all to-be-picked edible mushroom evaluation clusters is processed by the min-max normalization algorithm, and the priority score is mapped to the range of zero to one hundred to obtain the final to-be-picked priority of each to-be-picked edible mushroom evaluation cluster.

[0173] S215, according to the number of edible mushrooms in each to-be-picked edible mushroom evaluation cluster, calculate the regional to-be-picked density of each to-be-picked edible mushroom evaluation cluster;

[0174] S216, using the to-be-picked edible mushroom evaluation cluster in each preset initial path The to-be-picked priority and the corresponding regional to-be-picked density of the to-be-picked edible mushroom evaluation cluster and the position coordinate information of the to-be-picked edible mushroom evaluation cluster are used to construct the to-be-picked task list in each preset initial path.

[0175] The process realizes high-precision quantitative evaluation of the maturity of edible fungi through a systematic image processing and feature extraction process, and builds a complete data processing chain from raw perception data to picking decision. First, through multi-level image preprocessing techniques such as coordinate mapping correction, super-resolution reconstruction, dynamic interference suppression, and inter-frame redundancy removal, the spatial accuracy and feature quality of image data are significantly improved, providing a reliable data foundation for subsequent analysis. Then, the improved YOLOv8-based target detection algorithm combined with multi-scale feature enhancement realizes accurate positioning and segmentation of the mushroom body, and through edge detection and contour analysis, key morphological parameters such as cap diameter, edge curvature, color features, and gill clarity are further extracted. On this basis, environmental parameters and morphological features are systematically integrated, and the LSTM prediction model and fuzzy logic expert system are used for maturity scoring, and an adaptive clustering algorithm based on density peak search is used to divide edible fungi into evaluation clusters with similar characteristics. Finally, combining the growth curve and yield weight, the fuzzy reasoning and exponential decay model are used to calculate the picking priority considering the time urgency and yield value, forming a complete task list including spatial coordinates, maturity level, picking density, and priority. The systematic integration of this series of technical means makes the maturity evaluation of edible fungi change from traditional subjective judgment to objective quantitative analysis based on multi-source data fusion, significantly improving the scientificity and accuracy of picking decisions.

[0176] It needs to be further explained that the generation of the coordinated picking task control instruction in the embodiment includes:

[0177] S301, the number, position coordinates of the picking robots in the idle state and in the working state at the current time, and the remaining working time of the picking robots in the working state in the corresponding evaluation cluster of the edible fungi to be picked are obtained, and the number of instant scheduling picking robots and the corresponding coordinate positions and the number of waiting picking robots and the corresponding waiting time length and the corresponding coordinate positions are obtained through a unified algorithm; the remaining working time is predicted according to the historical average picking speed of the picking robots, the number of remaining edible fungi in the corresponding evaluation cluster of the edible fungi to be picked, and the number of picking robots currently working in the corresponding evaluation cluster of the edible fungi to be picked combined with the LSTM algorithm;

[0178] It needs to be further explained that the process of obtaining the number of instant scheduling picking robots and the corresponding coordinate positions and the number of waiting picking robots and the corresponding waiting time length and the corresponding coordinate positions in the embodiment includes:

[0179] S3011, based on the working state data of all picking robots at the current time, the robots are divided into idle state robots and working state robots through a state recognition algorithm; the state recognition algorithm is realized by reading the state flag bit returned by the robot controller, and the state flag bit is zero indicating an idle state and the state flag bit is one indicating a working state;

[0180] S3012, the total number of idle state robots is counted as the number of instant scheduling picking robots, and the three-dimensional coordinate positions of these robots in the world coordinate system are obtained as the instant scheduling coordinate positions;

[0181] S3013, for each working state robot, based on the historical picking robot average picking rate parameter, the corresponding to-be-picked edible mushroom evaluation clustering cluster remaining edible mushroom quantity parameter, and the corresponding to-be-picked edible mushroom evaluation clustering cluster current working picking robot quantity parameter, a remaining working time parameter is predicted through a long short-term memory network algorithm; the input feature vector of the long short-term memory network algorithm includes a current working duration parameter, a completed picking quantity parameter, a current picking rate parameter, and an intra-cluster remaining edible mushroom density parameter;

[0182] S3014, based on the world coordinates of the current position of the working state robot and the center position of the target to-be-picked edible mushroom evaluation clustering cluster, a moving distance parameter is obtained through a Euclidean distance calculation algorithm; based on the moving distance parameter and the robot average moving speed parameter, a robot transfer time parameter is calculated through division operation;

[0183] S3015, based on the predicted remaining working time parameter and the robot transfer time parameter, a scheduling waiting time length parameter is obtained through addition operation;

[0184] S3016, the total number of working state robots is counted as the number of scheduling waiting picking robots, and the three-dimensional coordinate positions of these robots in the world coordinate system are obtained as the scheduling waiting coordinate positions, and the calculated scheduling waiting time length parameter and the corresponding coordinate position are associated and stored;

[0185] S3017, the instant scheduling picking robot number and the corresponding coordinate position, the scheduling waiting picking robot number and the corresponding waiting time length and the corresponding coordinate position are integrated into a structured scheduling resource list.

[0186] S302, based on the number of instant scheduling picking robots and the corresponding coordinate positions, and the number of scheduled waiting picking robots and the corresponding waiting time length and the corresponding coordinate positions, and the to-be-picked task list in each preset initial path, taking the shortest travel time and picking task completion time as the goal, through the optimal task allocation algorithm and the load balancing algorithm, the picking robot distribution label list corresponding to each to-be-picked mushroom evaluation clustering cluster and the corresponding travel path parameters are obtained.

[0187] S3021, based on the to-be-picked priority parameters, the remaining pickable time parameters, and the number of edible fungi in the clustering cluster of the to-be-picked edible fungi evaluation clustering cluster, combined with the instant scheduling picking robot coordinate position, the scheduled waiting picking robot coordinate position and the corresponding waiting time length parameter, a multi-dimensional decision feature matrix is constructed; the row dimension of the multi-dimensional decision feature matrix corresponds to the to-be-picked edible fungi evaluation clustering cluster identifier, the column dimension corresponds to the available picking robot identifier, and the matrix element contains the travel time estimation value, the waiting time compensation value and the urgency adjustment factor.

[0188] S3022, based on the current position coordinates of the picking robot and the center point coordinates of the to-be-picked edible fungi evaluation clustering cluster, the straight line distance parameter is obtained through the Euclidean distance calculation algorithm; based on the straight line distance parameter and the robot average moving speed parameter, the theoretical travel time is calculated by division operation; based on the environmental complexity coefficient, the theoretical travel time is modified to obtain the travel time estimation value, and the specific implementation process is: through the laser radar or visual sensor carried by the picking robot, the environmental data around the path from its current position to the center point of the to-be-picked edible fungi evaluation clustering cluster is collected, and a local environment map of the path is constructed combined with the SLAM technology; the obstacles in the local environment map are identified by using the DBSCAN clustering algorithm, and the obstacle density per unit area, the path width fluctuation value and the ground flatness deviation are extracted as environmental feature parameters; the weights of each environmental feature parameter are determined by using the analytic hierarchy process or the entropy weight method, and after the Min-Max standardization processing of each environmental feature parameter, the weighted sum is performed according to the weight, to obtain the environmental complexity coefficient with a value range of 0-1, and the larger the coefficient value, the more complex the path environment; based on the historical travel data of the robot, a travel time correction model is obtained by using the least square method, which takes the theoretical travel time and the environmental complexity coefficient as inputs, introduces a correction coefficient k calibrated by historical data, and calculates the travel time estimation value according to the calculation logic of travel time estimation value = theoretical travel time × (1+k×environmental complexity coefficient), and the travel time estimation value is obtained by substituting the theoretical travel time and the environmental complexity coefficient.

[0189] S3023, for the scheduled waiting picking robot, based on the waiting time length parameter, the waiting time compensation value is calculated by using the exponential decay function; the input of the exponential decay function is the waiting time length parameter, and the output is a compensation coefficient that decreases with time.

[0190] S3024, based on the remaining harvestable time parameter of the edible mushroom to be picked, the urgency adjustment factor is calculated by an inverse proportional function; the inverse proportional function outputs the maximum value when the remaining harvestable time tends to zero, and outputs smaller values when the remaining harvestable time is larger;

[0191] S3025, based on the travel time estimate value, the waiting time compensation value and the urgency adjustment factor, the comprehensive cost matrix is constructed by a linear weighted combination algorithm; the linear weighted combination algorithm is to weight and sum the three parameters according to the preset weight coefficient, and the weight coefficient is dynamically adjusted according to the system optimization target;

[0192] S3026, based on the comprehensive cost matrix, the Hungarian algorithm is used to solve the optimal allocation scheme; the Hungarian algorithm finds the task-robot matching relationship that minimizes the total comprehensive cost through three sub-steps of matrix reduction, independent zero element marking and augmented path search;

[0193] S3027, based on the optimal allocation scheme, the number of tasks allocated to each picking robot is calculated, and the load distribution balance degree is evaluated by a standard deviation algorithm; the standard deviation algorithm calculates the deviation degree of the task number of each robot from the average task number;

[0194] S3028, when the load balance degree exceeds the preset threshold, the task reassignment algorithm is used for adjustment; the task reassignment algorithm is based on the hill climbing search strategy, and under the premise of keeping the total comprehensive cost increasing to the minimum, part of the tasks of the high-load robot is redistributed to the low-load robot;

[0195] S3029, based on the final allocation scheme, a corresponding picking robot allocation tag list is generated for each edible mushroom cluster to be picked; the allocation tag list includes the unique identifier of the allocated picking robot, the task priority serial number and the predicted start working timestamp;

[0196] S30210, based on the current position coordinates of the picking robot and the target cluster center coordinates, the path planning algorithm calculates the optimal travel path; the path planning algorithm manages the search nodes through open list and closed list, evaluates the path cost based on heuristic function, and outputs the travel path parameters including path point coordinate sequence, total path length and predicted travel time.

[0197] S303, based on each to be picked edible mushroom evaluation clustering cluster corresponding picking robot distribution label list and corresponding travel path parameters combined fuzzy control algorithm, generate each to be picked edible mushroom evaluation clustering cluster corresponding picking robot optimal path instruction; the picking robot optimal path instruction corresponding to each to be picked edible mushroom evaluation clustering cluster includes the number of distributed picking robots, the travel starting point and the travel endpoint of each picking robot, the travel starting time point and the travel time length;

[0198] S304, based on the maturity score and the corresponding maturity level of each position edible mushroom in the to-be-picked edible mushroom evaluation clustering cluster, and combining the maturity-gripping posture parameter table, the gripping posture parameter of each picking robot in picking a single edible mushroom in the current to-be-picked edible mushroom evaluation clustering cluster is obtained.

[0199] It needs to be further explained that the construction process of the maturity-gripping posture parameter table in the embodiment includes:

[0200] S3041, based on the edible mushroom maturity score parameter, the mapping relationship between the maturity level and the score interval is established by threshold division algorithm; the maturity level includes immature level, mature level and to-be-picked level, and the corresponding score interval is zero to fifty-nine minutes, sixty to seventy-nine minutes and eighty to one hundred minutes;

[0201] S3042, based on historical picking experiment data, the basic gripping posture parameter corresponding to each maturity level is determined by statistical analysis algorithm; the basic gripping posture parameter includes the gripping angle reference value, the gripping depth reference value and the approach angle reference value;

[0202] S3043, based on the edible mushroom cap diameter parameter and the stem height parameter, the basic posture parameter is corrected by multivariate linear regression algorithm; the multivariate linear regression model takes the cap diameter parameter and the stem height parameter as the independent variable, and takes the posture parameter adjustment amount as the dependent variable;

[0203] S3044, based on the maturity score parameter of the current edible mushroom, the corresponding gripping posture parameter is obtained by table lookup interpolation algorithm; the table lookup interpolation algorithm first determines the maturity level, and then performs linear interpolation calculation in the parameter range of the corresponding level;

[0204] S3045, based on the physical limitation of the mechanical system, the feasibility of the gripping posture parameter is verified by boundary constraint algorithm; the boundary constraint algorithm checks whether the gripping angle value is within the range of fifteen degrees to ninety degrees, and whether the gripping depth value is within the range of five millimeters to thirty millimeters;

[0205] S3046, encapsulate the clamping posture parameters that pass the verification into a control instruction set, including the specific value of the gripper opening angle, the specific value of the clamping depth, and the specific value of the approach angle, and issue them to the end effector of the corresponding picking robot.

[0206] S305, based on the maturity score and the corresponding maturity level of each position of the edible fungi in the current edible fungi evaluation clustering cluster to be picked, and combining the maturity-gripping force curve, obtain the gripping force parameter of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering cluster to be picked;

[0207] It should be further explained that the construction process of the maturity-gripping force curve in the embodiment includes:

[0208] S3051, based on historical picking experiment data, a function relationship between the maturity score and the reference gripping force is established by a polynomial fitting algorithm; the polynomial fitting algorithm uses a quadratic polynomial model, with the maturity score parameter as the independent variable and the reference gripping force parameter as the dependent variable;

[0209] S3052, based on the influence of the cap diameter parameter on the gripping force, a diameter compensation coefficient is determined by a linear interpolation algorithm; the linear interpolation algorithm calculates the corresponding compensation coefficient value according to the position of the cap diameter parameter in the preset diameter range;

[0210] S3053, based on the reference gripping force parameter and the diameter compensation coefficient, a comprehensive gripping force parameter is calculated by a product algorithm; the product algorithm multiplies the reference gripping force parameter, the diameter compensation coefficient, and the environmental humidity correction factor to obtain the final gripping force value;

[0211] S3054, based on the safety requirements of the mechanical system, the reasonableness of the gripping force parameter is verified by a threshold comparison algorithm; the threshold comparison algorithm checks whether the comprehensive gripping force parameter is within the allowed range of 0.5N to 5N;

[0212] S3055, the verified gripping force parameter is converted into a motor torque control signal, including the force setting value and the torque conversion coefficient, and is issued to the torque controller of the corresponding picking robot.

[0213] S306, based on the gripping force parameter of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering cluster to be picked, and combining the force-lifting speed curve, obtain the picking lifting speed control parameter of each picking robot in picking a single edible fungus in the current edible fungi evaluation clustering cluster to be picked.

[0214] It should be further explained that the construction process of the force-lifting speed curve in the embodiment includes:

[0215] S3061, based on historical picking data, a mathematical relationship between the clamping force parameter and the mushroom damage probability is established by a logistic regression algorithm; the logistic regression algorithm takes the clamping force parameter as the independent variable and the mushroom damage probability as the dependent variable;

[0216] S3062, based on the edible mushroom damage probability threshold, the interpolation algorithm is used to determine the pulling speed adjustment coefficient; the interpolation algorithm calculates the corresponding speed adjustment coefficient value according to the position of the damage probability in the preset probability range;

[0217] S3063, based on the cap diameter parameter and the cap edge curvature parameter, the cap fragility index is calculated by principal component analysis; the principal component analysis reduces the cap diameter parameter and the cap edge curvature parameter to a single fragility index;

[0218] S3064, based on the clamping force parameter and the target damage probability, the reference pulling speed control parameter is calculated by the inverse function solving algorithm; the inverse function solving algorithm solves the pulling speed value that meets the target damage probability by Newton iteration method;

[0219] S3065, based on the reference pulling speed control parameter, the speed adjustment coefficient and the cap fragility correction factor, the optimal pulling speed control parameter is calculated by the weighted fusion algorithm;

[0220] S3066, based on the mechanical system performance parameter, the feasibility of the pulling speed control parameter is verified by the constraint satisfaction algorithm; the constraint satisfaction algorithm checks whether the pulling speed control parameter is within the minimum speed and maximum speed range allowed by the device;

[0221] S3067, the verified pulling speed control parameter is converted into a motor control signal including a target speed value and an acceleration limit value, and is sent to the motion control system of the corresponding picking robot.

[0222] S307, based on the clamping posture parameter, the force parameter and the pulling speed control parameter of each picking robot in the evaluation cluster of the current edible mushroom to be picked, the PID control algorithm is combined to generate the picking task control instruction of each picking robot;

[0223] S308, based on the picking task control instruction of all picking robots combined with the simulation algorithm, the minimum picking damage ratio is taken as the target for optimization training to generate the coordinated picking task control instruction in each edible mushroom evaluation cluster to be picked, and real-time picking control is performed.

[0224] The process realizes the global optimization of picking robot resource scheduling and operation parameters by building a multi-level collaborative control system. First, based on the LSTM algorithm, the remaining operation time of the working state robot is accurately predicted, and combined with the real-time position information, a complete scheduling resource list containing instant available resources and expected available resources is constructed, providing accurate spatiotemporal resource data support for task allocation; secondly, the travel time, waiting time and picking urgency are innovatively integrated into the comprehensive cost matrix, and the initial optimal allocation is realized through the Hungarian algorithm, and then the load balancing optimization is carried out combined with the hill climbing search strategy, which not only ensures the time efficiency of task allocation, but also avoids the overload of a single robot; more importantly, a complete mapping chain from maturity to specific operation parameters is established: the maturity level is converted into specific mechanical operation parameters through the maturity-posture parameter table; the intensity curve based on polynomial fitting combined with environmental factors realizes accurate control of intensity; the intensity-speed relationship constructed by the logistic regression model effectively controls the damage risk while ensuring the picking efficiency. Finally, through PID control and simulation optimization, the coordination of each parameter is ensured. This whole-link optimization from macro resource scheduling to micro operation parameters enables the system to significantly improve the operation efficiency while ensuring the picking quality, and realizes the efficient collaborative operation of multiple robots.

[0225] It should be further pointed out that the process of triggering the secondary imaging confirmation of the picking area in the embodiment includes:

[0226] S401, based on the state signal of the end effector of the picking robot completing the picking action, a picking completion signal is sent to the nearest inspection robot through a wireless communication protocol; the picking completion signal contains the edible mushroom evaluation cluster identifier to be picked, the picking completion timestamp and the robot identifier performing the picking;

[0227] S402, based on the received picking completion signal, the inspection robot adjusts the current inspection path through a path re-planning algorithm; the path re-planning algorithm uses the Dijkstra shortest path algorithm to calculate the optimal path of the robot to the center coordinates of the edible mushroom evaluation cluster to be picked;

[0228] S403, based on the multi-spectral imaging system carried by the inspection robot, the region scanning algorithm is used to conduct secondary imaging on the edible mushroom evaluation cluster region to be picked; the region scanning algorithm uses a raster scanning mode to collect image data containing visible light and near-infrared bands;

[0229] S404, based on the secondary imaging data, the improved YOLOv8 algorithm is used to identify the residual edible mushrooms; after the identification is completed, the density calculation algorithm is used to count the number of residual edible mushrooms in the edible mushroom evaluation cluster to be picked, and the residual area picking density is calculated combined with the actual effective planting area of the region;

[0230] S405, based on the real-time picking quantity parameter, the secondary confirmation missing quantity parameter and the picking task completion time length parameter, obtaining the current picking efficiency value through the efficiency calculation algorithm; the efficiency calculation algorithm is the real-time picking quantity minus the secondary confirmation missing quantity and divided by the picking task completion time length;

[0231] S406, based on the picking efficiency value and the residual area to be picked density parameter, updating the preset initial path planning strategy through the Q-learning algorithm; the Q-learning algorithm takes the maximization of picking efficiency as the reward function and the minimization of residual area to be picked density as the constraint condition;

[0232] S407, based on the optimized path planning parameter, solving the optimal inspection path sequence through the genetic algorithm; the genetic algorithm takes the maximization of picking efficiency as the fitness function, and outputs the updated preset initial path;

[0233] S408, the updated preset initial path is issued to all inspection robots, and the real-time optimization adjustment of the inspection path is completed.

[0234] The embodiment constructs a full-link intelligent operation system from environmental perception to picking optimization, realizes the precision and efficient management of edible mushroom planting through the collaborative innovation of multiple core technologies; among them, in the inspection and perception link, the optimal inspection path is generated based on grid division and efficiency target, and the path inside and outside double mode data acquisition mechanism is innovatively adopted, combined with the spatial matching algorithm to realize the cross-robot data sharing, and through the high-precision world coordinate system construction and multi-source fusion positioning technology, the spatial consistency of all robots and mushroom coordinates is ensured, which fundamentally solves the perception blind area problem of traditional single path inspection, and significantly improves the coverage range and data accuracy of global monitoring of mushroom house; in the data processing link, the system adopts multi-level optimization strategy, improves the spatial positioning accuracy through coordinate mapping correction algorithm, enhances the image detail features based on super-resolution reconstruction technology, uses dynamic interference suppression method to ensure the stability of image sequence, and uses frame redundancy removal mechanism to improve the data processing efficiency, which provides a high-quality data basis for subsequent analysis. In the maturity evaluation link, the system accurately identifies the mushroom through the improved multi-scale target detection network, quantifies the key parameters combined with the morphological feature extraction algorithm, realizes the accurate maturity determination by using the evaluation method of time series prediction model and expert system fusion, and realizes the scientific grouping of mushrooms based on the density adaptive clustering algorithm, establishing a complete evaluation system from image recognition to maturity grading; in the task scheduling and picking control link, the system innovatively constructs a multi-dimensional resource scheduling model, comprehensively considers the robot state, environmental complexity and time urgency and other factors, and realizes the efficient allocation of tasks through optimization algorithm; at the same time, a complete mapping relationship from maturity to specific operation parameters is established to ensure the accuracy and adaptability of picking action; in the secondary optimization link, the system realizes real-time evaluation of picking effect through multispectral imaging and intelligent algorithm, and continuously optimizes the operation strategy based on reinforcement learning and evolutionary algorithm, forming a complete closed-loop optimization system.

[0235] Embodiment 2

[0236] Please refer to Figure 2 The application provides another embodiment of a robot collaborative operation control system in an edible mushroom planting process, which comprises a data acquisition module, an identification module, an instruction module, a response module and a feedback confirmation module.

[0237] The data acquisition module is used for scanning the mushroom bed by the inspection robot along the preset initial path, and acquiring the environmental perception data of the edible mushroom.

[0238] The identification module is used for determining the maturity of at least one edible mushroom and calculating the area picking density based on the environmental perception data of the edible mushroom combined with an image recognition algorithm, and generating a picking task list containing the coordinates of the area to be picked, the maturity grade and the area picking density; the environmental perception data of the edible mushroom at least contains the image and three-dimensional spatial information of the edible mushroom.

[0239] an instruction module, based on the to-be-picked task list, in combination with a load balancing strategy and a preset maturity-gripping posture parameter table, a maturity-gripping force curve and a force-pulling speed curve, generating a coordinated picking task control instruction; the coordinated picking task control instruction at least includes target picking coordinates, a number of picking robots and corresponding gripping posture, gripping force and pulling speed control parameters of each picking robot;

[0240] a response module, in response to the coordinated picking task control instruction, executing picking operation through an end gripper of the picking robot in the gripping posture, gripping force and pulling speed, and feeding back a picking completion signal to a nearest inspection robot from the picking area after picking is completed;

[0241] a feedback confirmation module, configured to trigger secondary imaging confirmation of the picking area, and if there is picking residue, generate a supplementary picking task list according to the to-be-picked density of the residue area and repeat response to the corresponding coordinated picking task control instruction.

[0242] The embodiments of the present application are described above in combination with the drawings, but the present application is not limited to the above specific embodiments, and the above specific embodiments are only illustrative and not restrictive, and those skilled in the art can make changes, modifications, replacements and variations to the above embodiments without departing from the purpose of the present application and the scope of protection, which are all within the protection of the present application.

[0243] If the technical solution of the present disclosure involves personal information, the product applying the technical solution of the present disclosure has clearly informed the personal information processing rules before processing the personal information and obtained the personal independent consent. If the technical solution of the present disclosure involves sensitive personal information, the product applying the technical solution of the present disclosure has obtained the personal independent consent before processing the sensitive personal information, and at the same time meets the requirement of "explicit consent". For example, at the personal information collection device such as camera, a clear and prominent sign is set to inform that the personal information collection range has been entered and the personal information will be collected, and if the individual voluntarily enters the collection range, it is considered to agree to collect the personal information; or on the device for processing personal information, through the pop-up information or by uploading the personal information by the individual, the personal authorization is obtained under the condition that the obvious sign / information informs the personal information processing rules; wherein, the personal information processing rules can include personal information processor, personal information processing purpose, processing method and personal information type, etc.

Claims

1. A method for controlling the cooperative work of robots in the cultivation process of edible fungi, characterized in that, The method comprises the following steps: The inspection robot scans the mushroom bed along the preset initial path, and collects environmental perception data of the edible fungi; Based on the environmental perception data of the edible fungi and an image recognition algorithm, the maturity of at least one edible fungus is determined and the area density of the to-be-picked area is calculated, and a to-be-picked task list containing the coordinates of the to-be-picked area, the maturity level and the area density of the to-be-picked area is generated; the environmental perception data of the edible fungi at least contains the image and three-dimensional space information of the edible fungi; Based on the to-be-picked task list, a load balancing strategy, and a preset maturity-gripping posture parameter table, a maturity-gripping force curve and a force-pulling speed curve, a coordinated picking task control instruction is generated; the coordinated picking task control instruction at least contains target picking coordinates, the number of picking robots and the corresponding gripping posture, gripping force and pulling speed control parameters of each picking robot; In response to the coordinated picking task control instruction, the end gripper of the picking robot performs picking operation with the gripping posture, gripping force and pulling speed, and after picking is completed, a picking completion signal is fed back to the nearest inspection robot of the picking area, and secondary imaging confirmation of the picking area is triggered; if there is picking residue, a supplementary picking task list is generated according to the residue area density, and the corresponding coordinated picking task control instruction is repeatedly responded.

2. The method of claim 1, wherein the robot cooperative work control method is used in a mushroom cultivation process. The collection of the image and three-dimensional space information of the edible fungi comprises: S101. The first inspection robot continuously collects first environmental perception data within the preset initial path coverage range and second environmental perception data within the preset initial path coverage range and the collection angle range during the travel of the first inspection robot in the preset world coordinate system along the preset initial path; S102. Based on the second environmental perception data, the position of the second to-be-evaluated mushroom area in the second environmental perception data is identified in real time through a lightweight convolutional neural network, and based on the position of the second to-be-evaluated mushroom area, a second inspection robot corresponding to the preset initial path to which the position of the second to-be-evaluated mushroom area belongs is queried through an information sharing algorithm.

3. The method of claim 2, wherein the robot cooperative work control method is used in a mushroom cultivation process. The collection of the image and three-dimensional space information of the edible fungi further comprises: S103. Based on the second environmental perception data, the initial maturity discrimination score of the second to-be-evaluated mushroom area is calculated; the initial maturity discrimination score is obtained based on the estimated density of the mushroom in the second to-be-evaluated mushroom area, the Euclidean distance between the second to-be-evaluated mushroom area and the current position of the inspection robot, and the growth state of the edible fungi predicted based on a preset maturity growth prediction model in combination with an expert evaluation system; S104. When the initial maturity discrimination score of at least one second to-be-evaluated mushroom area exceeds a set threshold, the first inspection robot shares the second environmental perception data and the initial maturity discrimination score to the second inspection robot through an information sharing algorithm according to the second inspection robot obtained by querying, and constructs the third environmental perception data of the second inspection robot in combination with the first environmental perception data collected by the second inspection robot in real time.

4. The method of claim 3, wherein the robot cooperative work control method is used in a mushroom cultivation process. The acquisition process of the to-be-picked task list comprises: Based on the third environment perception data of any inspection robot, arrange in chronological order to obtain a sequence of images to be identified, wherein each image is associated with three-dimensional spatial information markers corresponding to the time stamp; Coordinate mapping, interference suppression, feature restoration, and inter-frame redundancy reduction are performed on each image in the sequence of images to be identified to obtain a complete inter-frame incremental area image sequence of any inspection robot; the length of the complete inter-frame incremental area image sequence is the same as the preset initial path length of the corresponding inspection robot; Based on the complete inter-frame incremental area image sequence of any inspection robot, image size normalization processing is performed to obtain a normalized mushroom body area image; Based on the normalized mushroom body area image, a modified YOLOv8 algorithm is used for mushroom target detection to obtain the bounding box coordinates of each edible mushroom; Based on the bounding box coordinates of each edible mushroom, a region segmentation algorithm is used to obtain a single mushroom area image, and an edge detection algorithm is used to obtain single mushroom edge profile data based on the single mushroom area image; Based on the single mushroom edge profile data, a contour segmentation algorithm is used to obtain cap contour data and stem contour data.

5. The method of claim 4, wherein the robot cooperative work control method is characterized by, The process of obtaining the to-be-picked task list further includes: Based on the cap contour data, the cap diameter parameter, the cap edge curvature parameter, the cap color RGB mean parameter, and the gill clarity parameter are obtained; Based on the stem contour data, the stem height parameter is obtained through a height calculation algorithm; Based on the bounding box coordinates of all edible mushrooms, the total number of detected edible mushrooms in the preset initial path is obtained through quantity statistics, denoted as the area mushroom total count.

6. The method of claim 5, wherein the robot cooperative work control method is used in a mushroom cultivation process. The process of obtaining the to-be-picked task list further includes: Based on the initial maturity discrimination score in the current preset initial path and the corresponding single mushroom feature parameter set extracted, the maturity score and corresponding maturity level of edible mushrooms at each position in the preset initial path of any inspection robot are obtained through the maturity growth prediction model and the pre-trained expert evaluation system; the single mushroom feature parameters include cap diameter parameter, cap edge curvature parameter, cap color RGB mean parameter, gill clarity parameter, and stem height parameter; Based on the maturity level of edible mushrooms at each position in the preset initial path of any inspection robot and the corresponding single mushroom feature parameters, the ratio of single mushroom feature parameter similarity to maturity score deviation is used as the clustering density parameter, and the adaptive clustering algorithm is used to divide the preset initial path into regions to obtain the edible mushroom evaluation cluster set on the current preset initial path; Based on the edible mushroom evaluation cluster set on the current preset initial path and the preset to-be-picked evaluation threshold, the to-be-picked edible mushroom evaluation cluster set and the to-be-mature edible mushroom evaluation cluster set are obtained.

7. The method of claim 6, wherein the robot cooperative work control method is used in a mushroom cultivation process. The process of obtaining the to-be-picked task list further includes: Based on the ratio of the corresponding maturity score, the number of to-be-picked edible mushroom evaluation clusters, and the area mushroom total count in each to-be-picked edible mushroom evaluation cluster, the preset edible mushroom growth curve, and the expert evaluation algorithm, the to-be-picked priority of each to-be-picked edible mushroom evaluation cluster is obtained. Meanwhile, according to the number of edible fungi in each to-be-picked edible fungi evaluation cluster, the area to-be-picked density of each to-be-picked edible fungi evaluation cluster is calculated; The to-be-picked priority of each to-be-picked edible fungi evaluation cluster in each preset initial path and the corresponding area to-be-picked density and the position coordinate information of the to-be-picked edible fungi evaluation cluster are used to construct a to-be-picked task list in each preset initial path.

8. The method of claim 7, wherein the robot cooperative work control method is used in a mushroom cultivation process. The generated coordinated picking task control instruction comprises: The number, position coordinates of idle and working picking robots at the current moment and the remaining working time of the working picking robots in the corresponding to-be-picked edible fungi evaluation cluster are obtained, and the number of instantaneously scheduled picking robots and the corresponding coordinate positions and the number of scheduled waiting picking robots and the corresponding waiting time length and the corresponding coordinate positions are obtained by using a statistical algorithm; the remaining working time is predicted by using an LSTM algorithm according to the historical average picking rate of the picking robots, the remaining number of edible fungi in the corresponding to-be-picked edible fungi evaluation cluster and the number of currently working picking robots in the corresponding to-be-picked edible fungi evaluation cluster; Based on the number of instantaneously scheduled picking robots and the corresponding coordinate positions and the number of scheduled waiting picking robots and the corresponding waiting time length and the corresponding coordinate positions and the to-be-picked task list in each preset initial path, the shortest travel time and picking task completion time are taken as the target, and the optimal task allocation algorithm and the load balancing algorithm are used to obtain the picking robot allocation tag list corresponding to each to-be-picked edible fungi evaluation cluster and the corresponding travel path parameters; Based on the picking robot allocation tag list corresponding to each to-be-picked edible fungi evaluation cluster and the corresponding travel path parameters, the fuzzy control algorithm is used to generate the optimal path instruction of the picking robot corresponding to each to-be-picked edible fungi evaluation cluster; the optimal path instruction of the picking robot corresponding to each to-be-picked edible fungi evaluation cluster comprises the number of allocated picking robots, the travel starting point and ending point of each picking robot, the travel starting time point and the travel time length.

9. The method of claim 8, wherein the robot cooperative work control method is used in a mushroom cultivation process. The generated coordinated picking task control instruction further comprises: Based on the maturity score and the corresponding maturity level of each position edible fungus in the to-be-picked edible fungi evaluation cluster, the maturity-gripping posture parameter table is used to obtain the gripping posture parameter of each picking robot in picking a single edible fungus in the current to-be-picked edible fungi evaluation cluster; Meanwhile, based on the maturity score and the corresponding maturity level of each position edible fungus in the to-be-picked edible fungi evaluation cluster, the maturity-gripping force curve is used to obtain the gripping force parameter of each picking robot in picking a single edible fungus in the current to-be-picked edible fungi evaluation cluster; Based on the gripping force parameter of each picking robot in picking a single edible fungus in the current to-be-picked edible fungi evaluation cluster, the force-lifting speed curve is used to obtain the picking lifting speed control parameter of each picking robot in picking a single edible fungus in the current to-be-picked edible fungi evaluation cluster; The picking posture parameter, the force parameter and the pulling speed control parameter of each picking robot in the cluster of the edible mushroom to be picked are evaluated based on the current picking edible mushroom, and a PID control algorithm is combined to generate the picking task control instruction of each picking robot; Based on the picking task control instruction of all picking robots, a simulation algorithm is combined to optimize training with the minimum picking damage ratio as the target to generate the coordinated picking task control instruction in each cluster of the edible mushroom to be picked, and real-time picking control is performed.

10. A robot cooperative work control system for use in a process for cultivating edible mushrooms, for implementing the robot cooperative work control method for use in a process for cultivating edible mushrooms according to any one of claims 1 to 9, characterized by It includes: A data acquisition module, an identification module, an instruction module, a response module and a feedback confirmation module; The data acquisition module is used for the scanning of the mushroom bed by the inspection robot along the preset initial path to collect the environmental perception data of the edible mushroom; The identification module determines the maturity of at least one edible mushroom and calculates the regional picking density based on the environmental perception data of the edible mushroom combined with an image recognition algorithm to generate a picking task list containing the coordinates of the picking area, the maturity level and the regional picking density; the environmental perception data of the edible mushroom at least contains the image and three-dimensional space information of one edible mushroom; The instruction module generates the coordinated picking task control instruction based on the picking task list combined with the load balancing strategy and the preset maturity-gripping posture parameter table, maturity-gripping force curve and force-pulling speed curve; the coordinated picking task control instruction at least contains the target picking coordinates, the number of picking robots and the corresponding gripping posture, gripping force and pulling speed control parameters of each picking robot; The response module responds to the coordinated picking task control instruction, and the end gripper of the picking robot performs the picking operation with the gripping posture, gripping force and pulling speed, and feeds back the picking completion signal to the inspection robot closest to the picking area after picking; The feedback confirmation module is used to trigger the secondary imaging confirmation of the picking area, and if there is picking residue, a supplementary picking task list is generated according to the residual area picking density and the corresponding coordinated picking task control instruction is repeatedly responded.

Images