Device and method for detecting imperfect wheat grains

Abstract

The invention belongs to the technical field of wheat grain imperfection detection. The imperfect wheat grain detection device comprises a material conveying pipe, a material bin, a linear feeder, an upper disc, a lower disc, a transparent rotary supporting table, a first image collecting mechanism, a second image collecting mechanism, a servo motor and a collecting mechanism. A detection area is formed between an inner circular ring and an outer circular ring of the upper disc and connected with an involute structure through a supporting structure, the upper disc and the lower disc are provided with light grooves and triangular check blocks, the upper disc is further provided with a cleaning cotton cloth groove, the servo motor drives the transparent rotary supporting table to rotate, the rotating speed of the servo motor is in the optimal interval, and it is guaranteed that wheat grains are evenly distributed and not stacked; the double-image acquisition mechanism acquires upper and lower images of wheat grains, precise detection is realized in combination with a pre-training deep learning model, and the wheat grains fall into the collection mechanism through an involute structure. According to the invention, the problems of low efficiency, poor precision, large equipment size and the like of traditional detection are solved, efficient and accurate detection is realized, and small scenes are adapted.

Description

Technical Field

[0001] This invention relates to the field of wheat grain imperfection detection technology, and in particular to a device and method for detecting imperfect wheat grains. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Wheat is one of the major food crops, and its grain quality directly affects the product quality of grain processing enterprises and the planting income of farmers. The proportion of imperfect wheat grains (including broken grains, black-germinated grains, and insect-damaged grains) is a core indicator for measuring wheat quality. Therefore, accurate and efficient detection of imperfect wheat grains is crucial. Currently, traditional detection of imperfect wheat grains mainly relies on manual sorting, where inspectors judge the presence of defects by visually observing the appearance of the grains. This method has many drawbacks: on the one hand, manual inspection is extremely inefficient and cannot meet the inspection needs of large-scale grain purchasing and processing; on the other hand, the test results are greatly affected by subjective factors such as personnel experience and fatigue, and the accuracy and consistency of the inspection cannot be guaranteed, easily leading to missed or false detections.

[0004] With the development of machine vision and deep learning technologies, some automated inspection equipment has begun to be applied in the field of grain inspection. Existing automated inspection devices mostly use linear conveyor belts as the conveying mechanism, spreading the grains evenly on the conveyor belt for image acquisition and recognition. However, such equipment has the following technical drawbacks: Firstly, wheat grains tend to pile up. During the linear conveyor belt transport process, wheat grains are prone to stacking due to uneven conveying speed and fluctuations in feeding volume, making it difficult for the camera to clearly capture images of the lower layers of wheat grains, thus affecting detection accuracy. Secondly, the equipment is large in size. The linear conveyor belt requires a long conveying path to complete detection and sorting, resulting in a large overall footprint, making it unsuitable for small grain collection points or other spaces with limited space. Thirdly, detection efficiency is limited. To avoid wheat grain stacking, the linear conveyor equipment often needs to reduce the feeding speed, and some wheat grains may be repeatedly detected due to conveyor belt jamming, further reducing overall detection efficiency. Fourthly, the imaging quality is poor. The supplementary lighting structure design of traditional detection devices is unreasonable, easily producing reflections and light strip shadows on the detection platform surface, interfering with the image recognition algorithm's judgment and leading to a decrease in the accuracy of identifying imperfect wheat grains. Furthermore, the core recognition algorithms of existing detection equipment mostly adopt traditional machine vision feature extraction methods, which have weak ability to distinguish between different types of imperfect wheat grains, especially when the wheat grain morphology differences are small, making accurate classification difficult and failing to meet the market demand for high-precision detection. Summary of the Invention

[0005] To address the shortcomings of existing technologies, this invention provides an imperfect wheat grain detection device and method, which achieves uniform delivery and non-stacking imaging of wheat grains. Combined with a pre-trained deep learning model, it improves the detection efficiency and accuracy of imperfect wheat grains. The device has a small overall size and strong adaptability.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: In a first aspect, the present invention provides an imperfect wheat grain detection device.

[0007] An imperfect wheat grain detection device includes: a conveying pipe, a hopper, a linear feeder, an upper disc, a transparent rotating support platform, a lower disc, a first image acquisition mechanism, a servo motor, a second image acquisition mechanism, and a collection mechanism. The upper part of the conveying pipe is used to receive the wheat grains to be tested. The lower outlet of the conveying pipe is at a set distance from the inner bottom of the hopper. The upper disc includes an inner ring and an outer ring. The area between the inner ring and the outer ring is the detection area. The inner ring and the outer ring are connected by a support structure and an involute structure. The outlet of the hopper is located above the detection area. A linear feeder is arranged at the bottom of the hopper to convey wheat grains by vibration. The involute structure is an outwardly convex involute shape, and the collection mechanism is located outside the involute structure in the detection area. A transparent rotating support platform is located between the upper and lower discs. The transparent rotating support platform is used to rotate horizontally under the drive of a servo motor. The first image acquisition mechanism is arranged directly above the detection area, and the second image acquisition mechanism is arranged directly below the detection area. Based on the images acquired by the first and second image acquisition mechanisms, a pre-trained deep learning model is used to detect imperfect wheat grains.

[0008] In one implementation of the first aspect of the present invention, the upper disc is provided with a groove for installing a cleaning cloth, and the groove is formed on the connector between the inner ring and the outer ring.

[0009] In one implementation of the first aspect of the present invention, the upper edge of the outer ring is an inwardly protruding triangular stop to confine the wheat grains within the detection area.

[0010] As a further limitation of the first aspect of the present invention, a light groove is provided on the inner sidewall of the outer ring, the lower disk is circular, a light groove is provided on the inner sidewall of the lower disk, and an inwardly protruding triangular stop is provided on the outer edge of the lower disk.

[0011] In one implementation of the first aspect of the present invention, the collecting mechanism includes a feeding trough and a feeding pipe. The feeding trough is located outside the involute structure of the detection area, and the upper end of the feeding pipe is located at the outlet of the feeding trough.

[0012] In one implementation of the first aspect of the present invention, a servo motor is connected to a gearbox, and an optimal speed range for the servo motor is determined so that the wheat grains can be evenly distributed in the detection area and avoid grain accumulation. Under the action of centrifugal force, friction, and extrusion force, the wheat grains can all fall along an involute curve into the feeding mechanism to avoid repeated detection and grain accumulation. The lower limit of the optimal speed range is the maximum value of the centrifugal force dispersion constraint, the filling rate constraint, and the linear velocity matching constraint, and the upper limit of the optimal speed range is... ,in, This represents the coefficient of static friction between the transparent rotating support platform and the grain of wheat. Represents gravitational acceleration. The radius of the entrance where wheat grains in the detection area enter the involute structure represents the radius of the entrance. This represents the proportion by which the involute structure counteracts centrifugal force.

[0013] As a further limitation of the first aspect of the invention, the centrifugal force dispersion constraint is as follows: ,in, Represents the static friction coefficient between wheat grains. Represents the degree of grain stacking. Represents the compressive strength of wheat grains. Represents the volume of wheat grains. Represents grain density. This represents the initial radius from which the wheat grains begin to disperse. This represents the particle stacking coefficient.

[0014] As a further limitation of the first aspect of the invention, the fill rate constraint is as follows: ,in, This represents the feeding speed of the linear feeder. Represents the volume of wheat grains. The radius of the entrance where wheat grains in the detection area enter the involute structure represents the radius of the entrance. This represents the initial radius from which the wheat grains begin to disperse. This represents the angle at which the wheat grains are covered on the transparent rotating support platform. Represents the area redundancy coefficient.

[0015] As a further limitation of the first aspect of the invention, the linear velocity matching constraint is: ,in, Represents the volume of wheat grains. Represents the safe spacing between wheat grains. Represents the initial velocity compensation coefficient. This represents the feeding speed of the linear feeder. The radius of the involute structure represents the entry point of wheat grains into the detection area.

[0016] Secondly, the present invention provides an imperfect method for detecting wheat grains.

[0017] A method for detecting imperfect wheat grains, utilizing the imperfect wheat grain detection device of the first aspect of the present invention, includes the following steps: The wheat grains to be tested are conveyed to the upper part of the conveying pipe. The wheat grains fall into the hopper through the discharge port at the lower part of the conveying pipe. The linear feeder is started and conveys the wheat grains in the hopper to the discharge port of the hopper through vibration. The wheat grains fall into the detection area between the inner and outer rings through the discharge port of the hopper. Start the servo motor and adjust its speed to the optimal speed range. The lower limit of the optimal speed range is the maximum value of the centrifugal force dispersion constraint, the filling rate constraint, and the linear velocity matching constraint. The upper limit of the optimal speed range is the value calculated by the corresponding formula. The servo motor drives the transparent rotating support platform to rotate horizontally. Under the action of centrifugal force, friction force, and extrusion force, the wheat grains are evenly distributed in the detection area without stacking. Simultaneously, the first image acquisition mechanism and the second image acquisition mechanism are activated. The first image acquisition mechanism acquires the upper image of the wheat grain from directly above the detection area, and the second image acquisition mechanism acquires the lower image of the wheat grain from directly below the detection area through the transparent rotating support platform. Based on the acquired upper and lower images of the wheat grain, the characteristics of the wheat grain are analyzed to determine whether the wheat grain is an imperfect wheat grain. After the test is completed, the wheat grains move outward along the involute structure under the action of centrifugal force and finally fall into the collection mechanism located outside the involute structure.

[0018] Compared with the prior art, the beneficial effects of the present invention are: This invention designs a transparent rotating support platform as the detection carrier and proposes a multi-parameter coupled servo motor speed determination method (setting an optimal speed range) to achieve uniform distribution of wheat grains and zero residual falling. Under the set optimal speed range, wheat grains can be evenly distributed in the specified area, avoiding wheat grain accumulation and ensuring good imaging quality. At the same time, under the action of centrifugal force, friction force, and extrusion force, the wheat grains can fall into the collection device along the involute curve, avoiding repeated detection and wheat grain accumulation.

[0019] This invention achieves an integrated structure for anti-reflective supplementary lighting and area limitation. The invention sets inverted triangular blocks on both the upper and lower discs. This structure has the dual functions of limiting the grain area and preventing reflection: on the one hand, the blocks can limit the grain to the detection area of ​​the inner and outer rings, preventing the grain from deviating from the field of view; on the other hand, the physical blocking eliminates the reflection on the surface of the glass disc and the reflection of the light strip, solving the technical problem of imaging interference in traditional supplementary lighting structures, and providing high-quality image input for deep learning models.

[0020] This invention achieves a collaborative design of involute feeding and rotary conveying. The design of the involute structure, together with the rotary support platform, forms a collaborative conveying mechanism: after the wheat grains move to the involute area with the rotary platform, they fall along the involute trajectory under the action of centrifugal force and extrusion force. This not only avoids the jamming and repeated detection problems of traditional linear conveying, but also greatly reduces the overall size of the equipment, adapting to the space requirements of small detection scenarios.

[0021] Advantages of additional aspects of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0022] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0023] Figure 1 A schematic diagram of an imperfect grain detection device provided as an exemplary embodiment of the present invention; Figure 2 A schematic diagram of the structure of the lower disk provided as an exemplary embodiment of the present invention; Figure 3 A schematic diagram of the upper disk structure is provided for an exemplary embodiment of the present invention; Figure 4 A flowchart illustrating an imperfect wheat grain detection method provided as an exemplary embodiment of the present invention; Figure 5 A schematic diagram of the recognition result of an image acquired by a first image acquisition mechanism, provided as an exemplary embodiment of the present invention; Figure 6 A schematic diagram of the recognition result of an image acquired by a second image acquisition mechanism, provided as an exemplary embodiment of the present invention; The components include: 1. Feed pipe; 2. First image acquisition mechanism; 3. Upper disc; 4. Transparent rotating support platform; 5. Lower disc; 6. Servo motor; 7. Feed pipe; 8. Support base; 9. Linear feeder; 10. Hopper; 11. Feed pipe position adjustment plate; 12. Discharge pipe; 13. First outer LED strip mounting slot; 14. Connecting hole; 15. First outer triangular stop block; 16. First inner LED strip mounting slot; 17. First inner triangular stop block; 18. Wire groove; 19. Involute structure; 20. Second inner triangular stop block; 21. Second outer LED strip mounting slot; 22. Through hole; 23. Second outer triangular stop block; 24. Cleaning strip mounting slot; 25. Second inner LED strip mounting slot. Detailed Implementation

[0024] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0025] It should be noted that the following detailed descriptions are exemplary and intended to provide further illustration of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0026] like Figure 1 As shown, this implementation proposes an imperfect grain detection device, including: a conveying pipe 1, a hopper 10, a linear feeder 9, an upper disc 3, a transparent rotating support platform 4, a lower disc 5, a first image acquisition mechanism 2, a servo motor 6, a second image acquisition mechanism, and a collection mechanism. The linear feeder 9 is fixed on a support base 8, and a conveying pipe position adjustment plate 11 is detachably connected to the support base 8. The conveying pipe position adjustment plate 11 is fixedly connected to the conveying pipe 1, and the vertical position of the conveying pipe 1 is adjusted by adjusting the position of the conveying pipe position adjustment plate 11 in the vertical direction of the support base 8.

[0027] The conveying pipe 1 of this implementation is designed with a discharge pipe 12 to control the weight of the wheat grain sample for a single test. At the same time, the distance between the conveying pipe 1 and the bottom of the hopper 10 can be adjusted (the conveying pipe 1 is slidably connected to the base, and the base is provided with a corresponding positioning hole. After sliding to the set position, it can be fixed by bolts or pins). This controls the stacking height and conveying speed of the wheat grains in the subsequent testing mechanism, and avoids wheat grain accumulation due to excessive feeding.

[0028] like Figure 2 As shown, the lower disc 5 includes an inner ring and an outer ring. The lower disc 5 has a connecting hole 14 for connecting with the upper disc 3. The inner ring has a first inner triangular stop 17 and a first inner LED strip mounting groove 16. The outer ring has a first outer LED strip mounting groove 13 and a first outer triangular stop 15. LED strips are installed in the first inner LED strip mounting groove 16 and the first outer LED strip mounting groove 13. The inner ring has a wire groove 18. Figure 3 As shown, the upper disk 3 also includes an inner ring and an outer ring. The purpose of designing the inner and outer rings is to limit the detection area and ensure that the wheat grains are distributed in the area between the two rings. This ensures that the camera has a good field of view. The area between the inner ring and the outer ring is the detection area. The inner ring and the outer ring are connected by a support structure and an involute structure 19.

[0029] The discharge port of the hopper 10 is located above the detection area; the linear feeder 9 is arranged at the bottom of the hopper 10 to convey wheat grains by vibration; the involute structure 19 is an outwardly convex involute shape; the collection mechanism is located outside the involute structure 19 in the detection area; the transparent rotating support platform 4 is located between the upper disk 3 and the lower disk 5; the transparent rotating support platform 4 is used to rotate horizontally under the drive of the servo motor 6; the first image acquisition mechanism 2 is arranged directly above the detection area; the second image acquisition mechanism is arranged directly below the detection area; based on the images acquired by the first image acquisition mechanism 2 and the second image acquisition mechanism, a pre-trained deep learning model is used to detect imperfect wheat grains (the first image acquisition mechanism 2 and the second image acquisition mechanism are symmetrically arranged above the transparent rotating support platform 4 to acquire wheat grain images from multiple angles to ensure the integrity of image information; the acquired images are transmitted to the pre-trained deep learning model for processing).

[0030] In this implementation, by setting an involute structure 19 and designing it using the involute principle, the wheat grains can fall along the involute line into the collection mechanism under the action of friction, centrifugal force, and extrusion force.

[0031] In this implementation, the upper disc 3 is provided with a cleaning strip mounting groove 24 for installing a cleaning cloth (the designed cleaning cloth can be installed to ensure the cleanliness of the glass turntable). The cleaning strip mounting groove 24 is opened on the connector between the inner and outer rings. The inner ring is provided with a second inner triangular stop 20 and a second inner light strip mounting groove 25, and the outer ring is provided with a second outer triangular stop 23 and a second outer light strip mounting groove 21. The second inner triangular stop 20 (inverted triangle) and the second outer triangular stop 23 (inverted triangle) restrict the wheat grains within the detection area. The second outer light strip mounting groove 21 and the second inner light strip mounting groove 25 contain light strips, avoiding the effects of reflection from the glass disc and the reflection of the light strip. The outer ring is provided with a through hole 22.

[0032] In this implementation, the transparent rotating support platform 4 is rotated by the servo motor 6. After the wheat grains fall onto the transparent rotating support platform 4, they pass through the detection area with the transparent rotating support platform 4. After the image is captured, they fall into the collection mechanism. The design of the transparent rotating support platform 4 can avoid repeated detection of wheat grains and greatly reduce the size of the device compared with the detection of linear conveyor belts.

[0033] In this implementation, the collection mechanism includes a feeding trough and a feeding pipe 7. The feeding trough is located outside the involute structure 19 of the detection area, and the upper end of the feeding pipe 7 is located at the outlet of the feeding trough. The detected wheat grains are transported to a designated external container through the feeding pipe 7 to achieve unified collection of the detected wheat grains.

[0034] In this implementation, a deep learning model is used. After the collected wheat grain images are input into the deep learning model, different types of targets such as broken grains, black germ grains, insect-eaten grains, and wheat stalks can be automatically identified. At the same time, the proportion of each type of imperfect wheat grain is calculated, and accurate detection results are output.

[0035] Based on the aforementioned imperfect grain detection device, and using a transparent rotating support platform 4 as the detection carrier, this invention proposes a method for determining the rotational speed range of a multi-parameter coupled servo motor 6 to achieve uniform distribution of grains and zero residual falling: within this rotational speed range, grains can be evenly distributed in a specified area, avoiding grain accumulation and ensuring good imaging quality; simultaneously, under the action of centrifugal force, friction, and extrusion force, all grains can fall along the involute curve into the collection device, avoiding repeated detection and grain accumulation.

[0036] The method for determining the speed range is as follows: To establish a servo motor speed constraint relationship that integrates variables such as wheat volume, weight, involute parameters, friction coefficient, and stacking degree, it is necessary to refine it from three dimensions: disk fabric mechanics, involute kinematics, and material property coupling, and finally obtain a complex function system containing multi-parameter coupling.

[0037] Servo motor 6 speed The constraint is essentially that the lower limit is determined by the transparent rotating support platform 4 fabric being "not stacked and uniform", and the upper limit is determined by the involute being "without residual falling". The two boundary conditions need to be derived separately to finally form the feasible region.

[0038] (1) Lower limit of speed Derivation (the transparent rotating support platform 4 is not stacked and is evenly distributed).

[0039] Lower speed limit It is the minimum rotation speed to ensure that the wheat does not stack and is evenly covered on the transparent rotating support platform 4. It needs to meet three independent constraints and take the maximum value of the three (failure to meet any constraint will result in stacking or uneven distribution of the material).

[0040] Constraint 1: The lower limit of centrifugal force overcoming interparticle friction (centrifugal force dispersion constraint).

[0041] The wheat grain moves in uniform circular motion on the transparent rotating support platform 4, and is subjected to three forces: centrifugal force : Drive wheat dispersion radially outward along the transparent rotating support platform 4: (1); in, The outer radius of the transparent rotating support platform 4 is taken as the minimum inner radius. At this point, the centrifugal force is minimal, making it easiest to stack, hence the use Calculate the lower limit; This represents the weight of a single grain of wheat (in kg). (wheat density). Represents the volume of a single grain of wheat (unit) ), disk angular velocity (rad / s), representing the speed of disk rotation. This represents a servo motor speed of 6.

[0042] Static friction between wheat grains (Radially inward, hindering dispersion): (2); in, This is the coefficient of static friction between wheat grains, used to hinder the radial dispersion of particles. Represents gravitational acceleration, unit Generally, constants are used. .

[0043] Total interparticle extrusion pressure (Radially outward, aiding dispersion): (3); in, Represents the particle stacking coefficient. Represents the degree of grain stacking. Represents the compressive strength of wheat grains. Represents the radius of the equivalent sphere of a grain of wheat.

[0044] The equilibrium condition is: (centrifugal force + total extrusion force) ≥ static friction between wheat grains, that is: (4); Formula simplification, derivation, and substitution Eliminate quality have to: (5); Divide both sides of the equation by Separation containing Item: (6); Solve The lower limit: (7); Constraint 2: The lower limit of the fill rate (fill rate constraint).

[0045] The total area occupied by wheat grains fed per unit time must be less than or equal to the "effective sweeping area rate × maximum filling rate × area redundancy coefficient" of the transparent rotating support table 4 to avoid local density exceeding the standard and causing stacking.

[0046] Transparent rotating support platform with 4 horizontal effective areas : (8); in, The outer radius of the transparent rotating support platform 4 To cover the central angle.

[0047] Horizontal projected area of ​​a single grain of wheat : (9); in, A grain of wheat is equivalent to a sphere.

[0048] The balance between the loading rate and the sweeping rate of the transparent rotating support table 4 is as follows: (10); in, Feeding speed, unit: particles / s Represents the area redundancy coefficient.

[0049] (11); Organized The lower limit: (12); Constraint 3: Lower limit of the linear velocity matching between the feeding and transparent rotating support platform 4 (linear velocity matching constraint).

[0050] The linear velocity of the outer edge of the transparent rotating support platform 4 must be greater than or equal to "feeding circumferential speed × initial velocity compensation coefficient" to offset the congestion effect of the initial velocity of the feeder outlet tangent.

[0051] Linear velocity of the outer edge of the transparent rotating support platform 4 (The maximum linear velocity of the wheat grain moving with the transparent rotating support platform 4) is: (13); The circumferential length that wheat fed per unit time needs to occupy on the transparent rotating support table 4 : (14); in, Total spacing between individual grains: (15); in, Represents the safe spacing between wheat grains. To avoid localized accumulation due to excessively fast feeding speed, the linear speed of the transparent rotating support platform must be greater than or equal to the feeding circumferential length rate. The equilibrium condition after compensation is: (16); in, This represents the initial velocity compensation coefficient.

[0052] Substitution Solve The lower limit: (17); All three constraints mentioned above are "minimum speed" requirements, which must be met simultaneously; therefore, the maximum value is taken. (18); (2) Maximum speed Derivation (to ensure that wheat grains "do not splatter and can slide down")

[0053] Constraint 1: Anti-splash constraint for involute guide compensation.

[0054] outer radius A grain of wheat is subjected to centrifugal force (outward), static friction between the grain and glass (inward), and involute normal support force (which partially counteracts the centrifugal force). The equilibrium condition is: (19) in, Represents the static friction coefficient between glass and wheat grains. The involute guiding compensation coefficient (taken as 0.25, representing that the involute counteracts 25% of the centrifugal force) is used to derive the upper limit of the effective centrifugal force: (20); , eliminate After sorting, we get: (twenty one); Constraint 2: Involute glide constraint (geometric requirement), only the involute development angle is required. This ensures that the wheat grains slide down the involute structure 19, regardless of the rotation speed, and does not affect the speed. Take values, where, This represents the involute curve result and the static friction coefficient of wheat grains. Therefore, the upper limit of rotational speed is integrated. Determined by the anti-splash constraint with guidance compensation: (twenty two); In summary, the range of rotational speeds is: (twenty three).

[0055] In this implementation, an imperfect method for detecting wheat grains is also proposed, such as... Figure 4As shown, the imperfect wheat grain detection device described above includes the following process: S401: The wheat grains to be tested are conveyed to the upper part of the conveying pipe 1. The wheat grains fall into the hopper 10 through the discharge port at the lower part of the conveying pipe 1. The linear feeder 9 is started and conveys the wheat grains in the hopper 10 to the discharge port of the hopper 10 through vibration. The wheat grains fall into the detection area between the inner ring and the outer ring through the discharge port of the hopper 10. S402: Start the servo motor 6 and adjust the speed of the servo motor 6 to the optimal speed range. The lower limit of the optimal speed range is the maximum value of the centrifugal force dispersion constraint, the filling rate constraint and the linear velocity matching constraint. The upper limit of the optimal speed range is the value calculated by the corresponding formula. The servo motor 6 drives the transparent rotating support platform 4 to rotate horizontally. Under the action of centrifugal force, friction force and extrusion force, the wheat grains are evenly distributed in the detection area and do not stack. S403: Simultaneously activate the first image acquisition mechanism 2 and the second image acquisition mechanism. The first image acquisition mechanism 2 acquires an upper image of the wheat grain from directly above the detection area, and the second image acquisition mechanism acquires a lower image of the wheat grain from directly below the detection area through the transparent rotating support platform 4. Based on the acquired upper and lower images of the wheat grain, the characteristics of the wheat grain are analyzed, such as... Figure 5 and Figure 6 As shown, determine whether a wheat grain is an imperfect grain; S404: After the detection is completed, the wheat grains move outward along the involute structure 19 under the action of centrifugal force and finally fall into the collection mechanism located outside the involute structure 19.

[0056] In this implementation, optionally, the neural network model for analyzing the characteristics of wheat grains can be implemented using existing deep learning network models, which will not be described in detail here.

[0057] The above description is merely a preferred embodiment of the present invention and is not intended to limit the invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. An imperfect wheat grain detection device, characterized in that, include: Material conveying pipe, hopper, linear feeder, upper disc, transparent rotating support platform, lower disc, first image acquisition mechanism, servo motor, second image acquisition mechanism and collection mechanism; The upper part of the conveying pipe is used to receive the wheat grains to be tested. The lower outlet of the conveying pipe is at a set distance from the inner bottom of the hopper. The upper disc includes an inner ring and an outer ring. The area between the inner ring and the outer ring is the detection area. The inner ring and the outer ring are connected by a support structure and an involute structure. The outlet of the hopper is located above the detection area. A linear feeder is arranged at the bottom of the hopper to convey wheat grains by vibration. The involute structure is an outwardly convex involute shape, and the collection mechanism is located outside the involute structure in the detection area. A transparent rotating support platform is located between the upper and lower discs. The transparent rotating support platform is used to rotate horizontally under the drive of a servo motor. The first image acquisition mechanism is arranged directly above the detection area, and the second image acquisition mechanism is arranged directly below the detection area. Based on the images acquired by the first and second image acquisition mechanisms, a pre-trained deep learning model is used to detect imperfect wheat grains.

2. The imperfect wheat grain detection device as described in claim 1, characterized in that, The upper disc has a groove for installing a cleaning cloth, and the groove is located on the connector between the inner and outer rings.

3. The imperfect wheat grain detection device as described in claim 1, characterized in that, The upper edge of the outer ring has an inwardly protruding triangular stop to confine the wheat grains within the detection area.

4. The imperfect wheat grain detection device as described in claim 3, characterized in that, The inner wall of the outer ring has a light groove, the lower disc is circular, the inner wall of the lower disc has a light groove, and the outer edge of the lower disc has an inwardly protruding triangular stop.

5. The imperfect wheat grain detection device as described in claim 1, characterized in that, The collection mechanism includes a feeding trough and a feeding pipe. The feeding trough is located outside the involute structure of the detection area, and the upper end of the feeding pipe is located at the outlet of the feeding trough.

6. The imperfect wheat grain detection device as described in claim 1, characterized in that, The servo motor is connected to the gearbox, and the optimal speed range of the servo motor is determined to ensure that the wheat grains are evenly distributed in the detection area and to avoid grain accumulation. Furthermore, under the influence of centrifugal force, friction, and extrusion, the wheat grains should fall entirely along an involute curve into the feeding mechanism to avoid repeated detection and grain accumulation. The lower limit of the optimal speed range is the maximum value of the centrifugal force dispersion constraint, the filling rate constraint, and the linear velocity matching constraint, while the upper limit of the optimal speed range is... ,in, This represents the coefficient of static friction between the transparent rotating support platform and the grain of wheat. Represents gravitational acceleration. The radius of the entrance where wheat grains in the detection area enter the involute structure represents the radius of the entrance. This represents the proportion by which the involute structure counteracts centrifugal force.

7. The imperfect wheat grain detection device as described in claim 6, characterized in that, Centrifugal force dispersion constraint is ,in, Represents the static friction coefficient between wheat grains. Represents the degree of grain stacking. Represents the compressive strength of wheat grains. Represents the volume of wheat grains. Represents grain density. This represents the initial radius from which the wheat grains begin to disperse. This represents the particle stacking coefficient.

8. The imperfect wheat grain detection device as described in claim 6, characterized in that, Fill rate constraint is ,in, This represents the feeding speed of the linear feeder. Represents the volume of wheat grains. The radius of the entrance where wheat grains in the detection area enter the involute structure represents the radius of the entrance. This represents the initial radius from which the wheat grains begin to disperse. This represents the angle at which the wheat grains are covered on the transparent rotating support platform. Represents the area redundancy coefficient.

9. The imperfect wheat grain detection device as described in claim 6, characterized in that, Linear velocity matching constraint is ,in, Represents the volume of wheat grains. Represents the safe spacing between wheat grains. Represents the initial velocity compensation coefficient. This represents the feeding speed of the linear feeder. The radius of the involute structure represents the entry point of wheat grains into the detection area.

10. An imperfect method for detecting wheat grains, characterized in that, The imperfect wheat grain detection device according to any one of claims 1-9 includes the following process: The wheat grains to be tested are conveyed to the upper part of the conveying pipe. The wheat grains fall into the hopper through the discharge port at the lower part of the conveying pipe. The linear feeder is started and conveys the wheat grains in the hopper to the discharge port of the hopper through vibration. The wheat grains fall into the detection area between the inner and outer rings through the discharge port of the hopper. Start the servo motor and adjust its speed to the optimal speed range. The lower limit of the optimal speed range is the maximum value of the centrifugal force dispersion constraint, the filling rate constraint, and the linear velocity matching constraint. The upper limit of the optimal speed range is the value calculated by the corresponding formula. The servo motor drives the transparent rotating support platform to rotate horizontally. Under the action of centrifugal force, friction force, and extrusion force, the wheat grains are evenly distributed in the detection area without stacking. Simultaneously, the first image acquisition mechanism and the second image acquisition mechanism are activated. The first image acquisition mechanism acquires the upper image of the wheat grain from directly above the detection area, and the second image acquisition mechanism acquires the lower image of the wheat grain from directly below the detection area through the transparent rotating support platform. Based on the acquired upper and lower images of the wheat grain, the characteristics of the wheat grain are analyzed to determine whether the wheat grain is an imperfect wheat grain. After the test is completed, the wheat grains move outward along the involute structure under the action of centrifugal force and finally fall into the collection mechanism located outside the involute structure.

Images