A machine vision-based inverted spoon type normal bud device and method

Abstract

The application discloses a machine vision-based turnover spoon type normal bud device and method, which comprises a controller and a seed taking device, a visual detection system and a normal bud device electrically connected with the controller, the seed taking device is used for picking up and conveying garlic seeds, the visual detection system is used for identifying the orientation of garlic scale buds, and the controller controls the posture adjustment of the normal bud device on the garlic seeds according to the garlic scale bud orientation information provided by the visual detection system, so that normal bud sowing is realized. Through the cooperative control of the rotatable profiled spoon body curved surface sliding structure and the visual detection system, the accurate identification of the orientation of the garlic scale buds is realized, meanwhile, the resistance of the silica gel soft head in the push-pull electromagnet is utilized to adjust the posture of the garlic scale buds, so that the garlic seeds fall into the duckbill inserter in the posture of the scale buds upward, and normal bud sowing is realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of agricultural machinery, in particular to a turning spoon type bud alignment device and method based on machine vision. BACKGROUND

[0002] Garlic is an important economic crop in the world, and its planting mode is experiencing a critical stage from traditional manual to mechanization. Although traditional manual seeding can accurately control the orientation of the bud, it has high labor intensity and low efficiency, which is difficult to match the demand of large-scale planting. The existing mechanical bud alignment equipment mostly relies on physical feature recognition, and the bud alignment rate of regular varieties is relatively high, but it is not suitable for hybrid varieties with short bud tips and various shapes, which leads to a decrease in the uniformity of seedling emergence after sowing, directly affecting the commodity rate of garlic buds. Moreover, the existing mechanical seeding devices generally face the dual challenges of structural complexity and functional limitation. For example, the vibration type seed metering mechanism is prone to pipe cavity blockage due to garlic seed friction, and the duckbill type directional device has high dependence on the shape of the bud tip, which is difficult to adapt to the differences in physical properties of different varieties.

[0003] Therefore, how to accurately identify and correct the orientation of the garlic bud to improve the bud alignment rate of the garlic seed is a technical problem that needs to be solved in the field. SUMMARY

[0004] To solve the above technical problems, the present application provides the following technical solutions:

[0005] In a first aspect, the present application provides a turning spoon type bud alignment device based on machine vision, comprising a controller and a seed taking device, a vision detection system and a bud alignment device electrically connected to the controller, the seed taking device is used to pick up and convey garlic seeds, the vision detection system is used to identify the orientation of the garlic bud, and the controller controls the posture adjustment of the garlic seed by the bud alignment device according to the garlic bud orientation information provided by the vision detection system, so as to realize the bud alignment seeding.

[0006] In a possible implementation, the seed taking device includes a seed box, the seed box is fixedly arranged on a seed box support, a plurality of seed plates are uniformly arranged in the seed box, and a circulating conveying mechanism is arranged around the seed plates.

[0007] In a possible implementation, the circulating conveying mechanism includes a transmission shaft arranged at the bottom of the seed box support, a chain wheel arranged on the transmission shaft, a transmission wheel arranged at the corner and the middle position of the seed plate, the chain wheel and the transmission wheel are connected by a chain transmission, a groove is arranged at the position where the seed plate contacts the chain, and the transmission wheel is arranged in the groove; the chain is provided with a turning seed spoon at equal intervals.

[0008] In a possible implementation, the seed box comprises a seed box bottom plate, a plurality of seed impacting plates are arranged between the seed box bottom plate and a seed box front plate, the seed impacting plates divide the seed box into a plurality of seed taking openings corresponding to the number of the circulating conveying mechanisms, a plurality of spring beams are arranged on the seed box support, spring seats are fixedly arranged on the spring beams, and an inclined guide device is fixedly arranged on the spring seats. The inclined guide device is a transparent groove structure, a drop prevention box is arranged at the top end of the inclined guide device, the width and height of the inclined guide device are matched with the flip seed taking spoon, and the flip seed taking spoon passes through the inclined guide device when the flip seed taking spoon is driven by the chain.

[0009] In a possible implementation, the flip seed taking spoon comprises a rotatable profiled spoon body, one end of the rotatable profiled spoon body is fixedly connected with one end of a torsion spring, the second end of the torsion spring is fixedly connected with the first end of a bottom support, the second end of the bottom support is fixedly connected with an accessory outer chain piece, and the accessory outer chain piece is arranged on the chain.

[0010] In a possible implementation, the normal bud device comprises a duckbill inserter arranged at the bottom of the inclined guide device, push-pull electromagnets are arranged at both ends of the entrance of the duckbill inserter, a retractable silica gel soft head is arranged at one end of each of the push-pull electromagnets, the silica gel soft heads are directed towards the flip seed taking spoon, and the control ends of the push-pull electromagnets are electrically connected with a controller.

[0011] In a possible implementation, the visual detection system comprises an industrial camera and a visual processing system, the industrial camera is arranged on the seed box support, the lens of the industrial camera is aligned with a preset flip seed taking spoon position, the industrial camera is electrically connected with the visual processing system, and the visual processing system is electrically connected with the controller.

[0012] In a second aspect, the embodiments of the present application provide a flip spoon type normal bud method based on machine vision, comprising:

[0013] The chain wheel drives the flip seed taking spoon to reciprocate after the flip seed taking spoon completes seed taking from the seed box, when the flip seed taking spoon is flipped to a preset inclination angle, the garlic seeds slide along the curved surface of the rotatable profiled spoon body, are limited by the inclined guide device, and slide to the back of the previous flip seed taking spoon to complete relay transmission and prepositioning of the garlic seeds.

[0014] An industrial camera is used to collect image information of the garlic seeds at the preset flip seed taking spoon position, and a YOLO11 target detection model is used to identify the orientation of the garlic seed scale buds in a plurality of continuous frames of images.

[0015] A BOT-SORT tracking algorithm is used to track the spatial trajectory of the garlic seeds in the relay transmission process of the flip seed taking spoon in real time.

[0016] When the garlic seed moves to the last mechanical positioning point before falling into the duckbill planter, the push-pull electromagnet on the same side of the scale bud is driven to extend a certain distance according to the identification result, the flip seed scoop is guided to flip to the corresponding side, and the garlic seed is slid into the duckbill planter in a scale bud upward posture to realize normal bud planting.

[0017] In a possible implementation, the identification of the garlic seed scale bud orientation in the continuous multiple frames of images by the YOLO11 target detection model comprises:

[0018] A frame buffer list and a garlic seed orientation determination list are constructed, and the garlic seed orientation determination list adopts a circular storage structure with end writing and head reading;

[0019] When a single garlic seed enters the visual recognition area, images are collected at a preset time as a period, and the confidence of the scale bud and the stem disc of the garlic seed is calculated by the YOLO11 target detection model;

[0020] When the scale bud recognition confidence in a single frame of image is greater than or equal to a preset value, element 1 is added in the frame buffer list, and when the stem disc recognition confidence is greater than or equal to a preset value, element 0 is added in the frame buffer list;

[0021] The number of 1 and 0 in the frame buffer list is counted;

[0022] If the number of 1 is greater than the number of 0, 1 is written at the end of the garlic seed orientation determination list and the frame buffer list is emptied;

[0023] If the number of 0 is greater than the number of 1, 0 is written at the end of the garlic seed orientation determination list and the frame buffer list is emptied.

[0024] In a possible implementation, when the garlic seed moves to the last mechanical positioning point before falling into the duckbill planter, the push-pull electromagnet on the same side of the scale bud is driven to extend a certain distance according to the identification result, the flip seed scoop is guided to flip to the corresponding side, and the garlic seed is slid into the duckbill planter in a scale bud upward posture to realize normal bud planting, which comprises:

[0025] When the garlic seed moves to the last mechanical positioning point before falling into the duckbill planter, the head element of the garlic seed orientation determination list is extracted, if it is 1, a left push-pull electromagnet trigger signal is sent to the controller to drive the left silica gel soft head to extend;

[0026] If it is 0, a right push-pull electromagnet trigger signal is sent to the controller to drive the right silica gel soft head to extend;

[0027] The silica gel soft head guides the flip seed scoop to flip to one side, so that the garlic seed slides to one side and ensures that the garlic seed scale bud falls into the duckbill planter upward.

[0028] Compared with the prior art, the application has the beneficial effects that:

[0029] The application realizes accurate identification of scale bud orientation through the cooperation of the profiled spoon body curved surface sliding structure and the visual detection system, combines the push-pull electromagnet to drive the silicone soft head on the same side of the scale bud to extend a certain distance, so that it can contact one side of the turnover seed taking spoon to trigger the turnover function of the turnover seed taking spoon. When the turnover seed taking spoon moves to the height of the push-pull electromagnet, the rotatable profiled spoon body part will contact the silicone soft head, and the turnover seed taking spoon will be deflected at a certain angle under the resistance of the silicone soft head to the garlic clove stem disc direction. The garlic clove changes from the original posture of lying flat on the back of the turnover seed taking spoon to the posture of the scale bud being inclined upward. With the continuous downward displacement of the turnover seed taking spoon, the deflection angle of the turnover seed taking spoon becomes larger and larger, and finally the garlic clove falls into the duck-bill planter in the posture of the scale bud being upward under the influence of gravity to realize the normal bud sowing. After the garlic clove slides off the back of the turnover seed taking spoon, the push-pull electromagnet pulls back the silicone soft head, and the turnover seed taking spoon is reset under the action of the internal torsional spring, so as to continue the cycle of taking seeds-transporting-turning-over-resetting.

[0030] Compared with the traditional gravity guide groove scheme, the application effectively solves the problem of random sliding caused by the weight difference between the scale bud and the stem disc. Through the cooperative working mechanism of the frame buffer list and the orientation determination list, continuous multi-frame checking is carried out during the period when the garlic clove enters the visual recognition area to the period when it moves out, which can effectively overcome the problem of instantaneous misidentification caused by mechanical vibration and sudden change of light in a single frame image. Through the visual detection of 4 seed taking spoon points in advance, the time and space matching of the algorithm processing period and the stable transmission period of the machine is formed, the problems of motion blur and posture distortion existing in the traditional scheme during image acquisition in the stage of rapid acceleration before falling are avoided, the correction accuracy of the scale bud orientation is further improved, and the trigger of the push-pull electromagnet is synchronized with the mechanical phase error controlled within 0.3 ms. BRIEF DESCRIPTION OF DRAWINGS

[0031] Figure 1 A whole structure schematic diagram of a turnover spoon type normal bud device based on machine vision provided for an embodiment of the application;

[0032] Figure 2 A structure schematic diagram of a circulating conveying mechanism provided for an embodiment of the application;

[0033] Figure 3 A taking seed process schematic diagram provided for an embodiment of the application;

[0034] Figure 4 A structure schematic diagram of a normal bud device provided for an embodiment of the application;

[0035] Figure 5 An exploded view of a turnover seed taking spoon provided for an embodiment of the application;

[0036] Figure 6A flowchart of a machine vision-based flip spoon type normal bud method provided in the embodiment of the present application is shown in the figure.

[0037] wherein, Figures 1-5 The symbols in the figure represent: 1-chain wheel, 2-industrial camera, 3-spring beam, 4-spring seat, 5-anti-falling box, 6-flip seed picking spoon, 7-seed box side plate, 8-seed box front plate, 9-seed pressing plate, 10-seed box, 11-transmission shaft, 12-seed plate, 13-pull-type electromagnet, 14-transmission wheel, 15-chain, 16-attachment outer chain piece, 17-rotatable profiled spoon body, 18-inclined guide device, 19-bottom support, 20-duckbill planter, 21-silicone soft head, 22-torsional spring. DETAILED DESCRIPTION

[0038] The present scheme will be described below in conjunction with the accompanying drawings and specific embodiments.

[0039] The machine vision-based flip spoon type normal bud device in the embodiment comprises a controller and a seed picking device, a vision detection system and a normal bud device electrically connected to the controller, wherein the seed picking device is used to pick up and convey garlic seeds, the vision detection system is used to identify the orientation of garlic scale buds, and the controller controls the posture adjustment of the normal bud device according to the garlic scale bud orientation information provided by the vision detection system to achieve normal bud seeding.

[0040] Figure 1 The overall structure of the machine vision-based flip spoon type normal bud device provided in the embodiment of the present application is shown in the figure. Figure 1 The seed picking device in the embodiment comprises a seed box 10, which is fixedly arranged on a seed box support. A plurality of seed plates 12 are uniformly arranged in the seed box 10, and a circulating conveying mechanism is arranged around the seed plates 12. Figure 2 In the embodiment, the circulating conveying mechanism comprises a transmission shaft 11 arranged at the bottom of the seed box support. A chain wheel 1 is arranged on the transmission shaft 11. A transmission wheel 14 is arranged at the corner and the middle position of each seed plate 12. The chain wheel 1 is in driving connection with the transmission wheel 14 through a chain 15. A groove is arranged at the position where the seed plate 12 contacts the chain 15. The transmission wheel 14 is arranged in the groove. Attachment outer chain pieces 16 are equally arranged on the chain 15. A flip seed picking spoon 6 is fixedly arranged on each attachment outer chain piece 16.

[0041] In the embodiment, the seed box 10 comprises a seed box bottom plate. A plurality of seed pressing plates 9 are arranged between the seed box bottom plate and the seed box front plate 8. The seed pressing plates 9 divide the seed box 10 into a plurality of seed picking ports corresponding to the number of the circulating conveying mechanism. A plurality of spring beams 3 are arranged on the seed box support. A spring seat 4 is fixedly arranged on the spring beam 3. An inclined guide device 18 is fixedly arranged on the spring seat 4. Figure 3In the embodiment, in order to accurately obtain the orientation of the scale bud of the garlic seed, the inclined guide device 18 is provided as a transparent groove structure, the top end is provided with the anti-falling box 5, the width and height of the inclined guide device 18 are matched with the turnover seed spoon 6, and the turnover seed spoon 6 is driven by the chain 15 to pass through the inclined guide device 18.

[0042] Referring to Figure 4 The turnover seed spoon 6 in the embodiment includes a rotatable profiled spoon body 17, one end of the rotatable profiled spoon body 17 is fixedly connected with one end of a torsional spring 22, the second end of the torsional spring 22 is fixedly connected with the first end of a bottom support 19, the second end of the bottom support 19 is fixedly connected with an accessory outer chain piece 16, the accessory outer chain piece 16 is arranged on the chain 15, and the restoring torque of the torsional spring 22 is 0.5-1.2 N·m, so that the spoon body is reset within 300 ms after the external force disappears.

[0043] Referring to Figure 5 In the embodiment, the positive bud device includes the duckbill planter 20 arranged at the bottom of the inclined guide device 18, the two ends of the duckbill planter 20 at the entrance are both provided with a push-pull electromagnet 13, one end of the push-pull electromagnet 13 is provided with a retractable silica gel soft head 21, the silica gel soft head 21 faces the turnover seed spoon 6, and the control end of the push-pull electromagnet 3 is electrically connected with the controller. In the embodiment, the visual detection system includes an industrial camera 2 and a visual processing system, the industrial camera frame 2 is arranged on the seed box support, the lens of the industrial camera 2 is aligned with the preset turnover seed spoon position, the industrial camera 2 is electrically connected with the visual processing system, and the visual processing system is electrically connected with the controller.

[0044] Corresponding to the turnover spoon type positive bud device based on machine vision provided in the above embodiment, an embodiment of a turnover spoon type positive bud method based on machine vision is also provided.

[0045] Figure 6 A flowchart of the turnover spoon type positive bud method based on machine vision provided in the embodiment is shown in Figure 6 The turnover spoon type positive bud method based on machine vision in the embodiment includes the following steps.

[0046] S101, the chain wheel drives the turnover seed spoon to reciprocate after completing seed picking from the seed box, when the turnover seed spoon is turned to a preset inclination angle, the garlic seed slides along the curved surface of the rotatable profiled spoon body, is limited by the inclined guide device, and slides to the back of the previous turnover seed spoon to complete relay transmission and prepositioning of the garlic seed.

[0047] In this embodiment, the turnover seed spoon rotates with the sprocket. When the turnover seed spoon passes the center axis of the sprocket by an angle of 15°-25°, the garlic seeds in the turnover seed spoon slide along the curved surface of the rotatable profiling spoon body under the action of gravity, and slide along the groove in the center of the inclined guide device to the back of the previous turnover seed spoon, realizing relay transfer and physical pre-positioning of the garlic seeds.

[0048] In S102, an industrial camera is used to collect image information of the garlic seeds at the preset turnover seed spoon positions, and a YOLO11 target detection model is used to identify the orientation of the garlic seed scale bud in the continuous multiple frames of images.

[0049] In this embodiment, when the garlic seeds are visually detected, only the two obvious characteristics of the scale bud and the stem disc of the garlic seeds need to be determined. An industrial camera is used to collect images of the garlic seeds at the fourth turnover seed spoon position in front of the duckbill planter inlet. When the garlic seeds are in the stable sliding stage in the middle section of the inclined guide device, the effective images are continuously identified at a period of 20 ms. At this time, the garlic seeds are in a state of dynamic equilibrium under the limiting action of the groove in the inclined guide device, and the image with the clearest scale bud feature can be obtained. Compared with the forward vertical projection detection, the algorithm complexity can be reduced, the computing power resources can be saved, and the detection accuracy can be improved.

[0050] After the corresponding images are obtained, the YOLO11 target detection model is used to identify the orientation of the garlic seed scale bud in the continuous multiple frames of images. In this embodiment, the YOLO11 target detection model uses a depth separable convolution to replace a standard convolution layer, and introduces an attention mechanism in a feature fusion module. The model inference speed is ≤50 ms / frame, and the identification accuracy is ≥98%. Through garlic seed image collection calibration and data enhancement, a garlic seed data set is made. The YOLO11 model trained through the data set can quickly identify the garlic seeds in the image and accurately identify the orientation of the garlic seeds. When the scale bud recognition confidence in a single frame of image is greater than or equal to 85%, element 1 is added to the pre-constructed frame buffer list. When the stem disc recognition confidence is greater than or equal to 85%, element 0 is added to the pre-constructed frame buffer list. When the garlic seed moves out of the identification area, the number of 1 and 0 in the frame buffer list is counted. If the number of 1 is greater than the number of 0, 1 is written to the end of the garlic seed orientation determination list and the frame buffer list is emptied. If the number of 0 is greater than the number of 1, 0 is written to the end of the garlic seed orientation determination list and the frame buffer list is emptied. The garlic seed orientation determination list is initialized to 4 elements and all are 0. A circular storage structure of end writing and head reading is adopted. When a new element is written each time, all existing elements are shifted one bit forward, and the original head element is removed.

[0051] In S103, the spatial trajectory of the garlic seeds in the relay transfer process in the turnover seed spoon is tracked in real time through the BOT-SORT tracking algorithm.

[0052] In this embodiment, after identifying the orientation of the garlic seed using the YOLO11 target detection model, the BoT-SORT tracking algorithm is used to track the spatial trajectory of the garlic seed in real time during the relay transmission process of the flipping seed-taking spoon. The BoT-SORT algorithm predicts the movement state of the garlic seed through Kalman filtering and uses a dual-weight matching mechanism of appearance similarity and motion consistency to align the garlic seed detection box in the continuous frame images of the industrial camera with the sprocket speed parameter in time and space, establishing an individual tracking sequence for the garlic seed. The BoT-SORT tracking algorithm records the garlic seeds that enter the detection area and assigns them a unique ID. It works in conjunction with the region detection algorithm to identify the state of the garlic seeds within the identification area, preventing extra garlic seeds outside the field of view of the industrial camera from affecting the identification results.

[0053] S104, when the garlic seed moves to the last mechanical positioning point before the duckbill planter falls, the push-pull electromagnet is controlled according to the recognition result to drive the silicone soft head on the same side of the scale to extend a certain distance, guiding the flipping seed-taking spoon to flip to the corresponding side, so that the garlic seed slides into the duckbill planter with the scale facing upward to achieve positive bud planting.

[0054] In this embodiment, when the garlic seed moves to the last mechanical positioning point before the duckbill planter falls, the first element of the garlic seed orientation determination list is extracted. If it is 1, a left push-pull electromagnet trigger signal is sent to the controller to drive the left silicone soft head to extend a certain distance.

[0055] If the value is 0, a right push-pull electromagnet trigger signal is sent to the controller to drive the right silicone soft head to extend a certain distance.

[0056] The silicone soft tip guides the flipping seed-taking spoon to flip to one side, causing the garlic seed to slide to one side, ensuring that the garlic seed scales fall upwards into the duckbill planter, achieving precise planting operation with a synchronization error of ≤0.3ms.

[0057] Furthermore, a communication middleware was constructed using Python scripts to establish a long-term connection with the push-pull electromagnet controller via the MC protocol. The middleware receives YOLO11 detection results and BoT-SORT tracking data in real time, converting the garlic seed orientation correction command into hexadecimal control words recognizable by the controller. In this embodiment, the Python script has a built-in exception handling mechanism. When a response delay of >50ms is detected in the push-pull electromagnet, a heartbeat detection is automatically triggered and the communication connection is reset, ensuring that the transmission reliability of the push-pull electromagnet control signal is ≥99.9%.

[0058] In the embodiments of the present application, "multiple" refers to two or more than two. The "and / or" describes the association between the associated objects, which means that there can be three kinds of relationships, for example, A and / or B, which can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Wherein A, B can be singular or plural. The character " / " generally represents that the associated objects before and after are a kind of "or" relationship.

[0059] It should be noted that in this paper, the term "includes", "contains" or any other variant thereof is intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements not only includes those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. Without more limitations, the element defined by the sentence "includes a…" does not exclude the existence of other identical elements in the process, method, article or device including the element.

[0060] The above is only a specific embodiment of the present application, and any person skilled in the art can easily think of changes or replacements within the technical scope disclosed by the present application, which should be covered by the protection scope of the present application. The protection scope of the present application should be subject to the protection scope of the claims.

Claims

1. A machine vision-based flipping spoon-type bud-correcting device, characterized in that, include: The system includes a controller and a seed-collecting device, a visual detection system, and a bud-aligning device electrically connected to the controller. The seed-collecting device is used to pick up and transport garlic seeds. The visual detection system is used to identify the orientation of the garlic seed scales. The controller controls the bud-aligning device to adjust the posture of the garlic seeds based on the garlic seed scale orientation information provided by the visual detection system, thereby achieving proper bud planting. The seed-collecting device includes a seed box, which is fixedly mounted on a seed box support. Multiple seed-distributing plates are evenly arranged inside the seed box, and a circulating conveyor is arranged around the seed-distributing plates. The mechanism includes a drive shaft located at the bottom of the seed box support, a sprocket mounted on the drive shaft, and drive wheels located at the corners and center of the seed metering plate. The sprocket and the drive wheels are connected by a chain drive. A groove is provided at the contact point between the seed metering plate and the chain, and the drive wheels are disposed within the groove. Accessory outer chain pieces are equidistantly arranged on the chain, and a flipping seed-taking spoon is fixedly mounted on each accessory outer chain piece. The seed box includes a seed box bottom plate, and multiple punches are provided between the seed box bottom plate and the seed box front plate. The seed box is divided into multiple seed-taking ports corresponding to the number of the circulating conveyor mechanism by a seed plate and a punch plate. Multiple spring beams are mounted on the seed box support, and spring seats are fixedly mounted on the spring beams. An inclined guide device is fixedly mounted on each spring seat. The inclined guide device has a transparent groove structure and an anti-drop box at its top. The width and height of the inclined guide device are adapted to the tilting seed-taking spoon. The tilting seed-taking spoon passes through the inclined guide device during chain transmission. The tilting seed-taking spoon includes a rotatable, contoured spoon body. One end of the shaped spoon body is fixedly connected to one end of the torsion spring, the second end of the torsion spring is fixedly connected to the first end of the bottom support, the second end of the bottom support is fixedly connected to the accessory outer chain piece, and the accessory outer chain piece is set on the chain; the budding device includes a duckbill inserter set at the bottom of the inclined guide device, and push-pull electromagnets are set at both ends of the inlet of the duckbill inserter. One end of the push-pull electromagnet is set with a retractable silicone soft head, and the silicone soft head faces the flipping seed-taking spoon. The control end of the push-pull electromagnet is electrically connected to the controller.

2. The machine vision-based flipping spoon-type bud-correcting device according to claim 1, characterized in that, The vision inspection system includes an industrial camera and a vision processing system. The industrial camera is mounted on a seed box support, and the lens of the industrial camera is aligned with a preset flip seed-picking spoon position. The industrial camera is electrically connected to the vision processing system, and the vision processing system is electrically connected to a controller.

3. A machine vision-based flipping spoon-type bud-correcting method, characterized in that, The method of using the machine vision-based flipping spoon bud-correcting device according to any one of claims 1-2 includes: The sprocket-driven flipping seed-taking spoon reciprocates after taking seeds from the seed box. When it flips to a preset tilt angle, the garlic seed slides down the curved surface of the rotatable contoured spoon body and is limited by the tilting guide device to slide to the back of the previous flipping seed-taking spoon, completing the relay transfer and pre-positioning of the garlic seed. An industrial camera was used to collect images of garlic cloves at a preset location for flipping the seed-taking spoon, and the orientation of the garlic clove scales in multiple consecutive frames of images was identified using the YOLO11 target detection model. The spatial trajectory of garlic seeds during the relay transfer process of the flipping seed-taking spoon is tracked in real time using the BOT-SORT tracking algorithm. When the garlic seed moves to the last mechanical positioning point before the duckbill planter falls, the push-pull electromagnet is controlled to drive the silicone soft head on the same side of the scale and bud to extend a certain distance according to the recognition result, guiding the flipping seed-taking spoon to flip to the corresponding side, so that the garlic seed slides into the duckbill planter with the scale and bud facing upwards to achieve positive bud planting.

4. The machine vision-based flipping spoon-type bud-correcting method according to claim 3, characterized in that, The step of identifying the orientation of garlic seed scales in multiple consecutive frames of images using the YOLO11 target detection model includes: Construct a frame buffer list and a garlic seed orientation determination list, wherein the garlic seed orientation determination list adopts a circular storage structure of end-to-end writing and beginning-to-end reading; When a single garlic clove enters the visual recognition area, images are acquired at preset time intervals, and the confidence scores of the garlic clove scales and stem discs are calculated using the YOLO11 target detection model. When the confidence level of bud identification in a single frame image is greater than or equal to the preset value, add element 1 to the frame buffer list; when the confidence level of stem plate identification is greater than or equal to the preset value, add element 0 to the frame buffer list. Count the number of 1s and 0s in the frame buffer list; If the number of 1s is greater than the number of 0s, write 1s to the end of the garlic seed orientation judgment list and clear the frame buffer list; If the number of 0s is greater than the number of 1s, write 0s to the end of the garlic seed orientation judgment list and clear the frame buffer list.

5. The machine vision-based flipping spoon-type bud-correcting method according to claim 3, characterized in that, When the garlic seed moves to the last mechanical positioning point before the duckbill planter falls, the push-pull electromagnet is controlled according to the recognition result to drive the silicone soft head on the same side of the scale to extend a certain distance, guiding the flipping seed-taking spoon to flip to the corresponding side, so that the garlic seed slides into the duckbill planter with the scale facing upwards to achieve positive bud planting, including: When the garlic seed moves to the last mechanical positioning point before the duckbill planter falls, the first element of the garlic seed orientation determination list is extracted. If it is 1, a left push-pull electromagnet trigger signal is sent to the controller to drive the left silicone soft head to extend. If the value is 0, a right push-pull electromagnet trigger signal is sent to the controller to drive the right silicone soft head to extend. The silicone soft tip guides the flipping seed-taking spoon to flip to one side, causing the garlic seed to slide to one side and ensuring that the garlic seed scales fall into the duckbill planter with the scales facing upwards.

Images