Automatic crab classification system based on image recognition

Abstract

The invention relates to an automatic crab classification system based on image recognition. The automatic crab classification system comprises a controller, and a sorting turntable driving device, a camera, a plurality of groups of photoelectric sensors and an air pump which are respectively connected with the controller, the controller is used for determining the type of the crab by adopting a pre-trained image recognition algorithm according to the image features of the abdomen image; wherein the variety comprises male crabs or female crabs; responding to the first identification result, controlling a sorting turntable driving device to act, and moving the crabs to a preset first area; and in response to the second identification result, controlling the sorting turntable driving device to act, and moving the crabs to a preset second area. According to the technical scheme provided by the invention, the type of the crab can be determined by adopting the pre-trained image recognition algorithm according to the image features of the abdomen image; and the sorting turntable is controlled to move different types of crabs to different areas, so that automatic sorting according to the types of the crabs is realized, the efficiency of crab classification work is improved, and manpower is saved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of image recognition, and in particular to an automatic crab classification system based on image recognition. Background Art

[0002] With the enrichment of marine fishery resources and the annual growth of crab fishing output, crab sorting has become an important and tedious link in fishery processing. Traditional crab sorting mainly relies on manual sorting, which is inefficient and has high labor costs. Summary of the invention

[0003] In order to overcome the problems existing in the related art, the present disclosure provides an automatic crab classification system based on image recognition to solve the problem of low efficiency of manual sorting in the related art.

[0004] According to a first aspect of an embodiment of the present disclosure, there is provided a crab automatic classification system based on image recognition, comprising: A controller, a sorting turntable drive device, a camera, and multiple groups of photoelectric sensors and an air pump respectively connected to the controller; The controller is used to control the air pump to start in response to the crab detection electrical signal sent by the photoelectric sensor, so as to adsorb the crab on the suction head; Controlling the sorting turntable driving device to drive the sorting turntable to rotate, and rotating the suction head and the crabs adsorbed on the suction head to a predetermined image detection position; At the image detection position, control the camera to turn on; take a video of the crab; and obtain an image of the abdomen of the crab; Determine the type of the crab using a pre-trained image recognition algorithm based on the image features of the abdomen image; wherein the type includes a male crab or a female crab; In response to the first recognition result, controlling the sorting turntable driving device to move the crabs to a predetermined first area; The first recognition result is that the crab is a male crab; In response to the second recognition result, controlling the sorting turntable driving device to move the crabs to a predetermined second area; The second identification result is that the crab is a female crab.

[0005] In some embodiments, multiple groups of photoelectric sensors and air pumps are located on the above-mentioned sorting turntable; The above-mentioned sorting turntable is also provided with multiple groups of rotating arms, suction heads, air pipes and photoelectric sensors; Among them, multiple rotating arms are evenly distributed; For any set of rotating arm, suction head, air pipe and air pump: The end of the rotating arm is provided with the air suction head; the air suction head is connected to the air pump through the air pipe.

[0006] In some embodiments, a hollow frame for fixing the air pipe is further provided in the middle of the rotating arm; the air pipe passes through the hollow frame to facilitate fixation.

[0007] In some embodiments, the sorting turntable drive device is located in a cabinet below the sorting turntable; The sorting turntable driving device at least includes: a hollow shaft, a chain, a reduction motor, a first bearing and a second bearing; The hollow shaft is vertically arranged in the cabinet below the sorting turntable; The first bearing is arranged in the top end cover of the lower cabinet; The first end of the hollow shaft is sleeved in the inner sleeve of the first bearing. The second end of the hollow shaft is fixed to the bearing seat of the second bearing; The output shaft of the reduction motor and the gear sleeved on the hollow shaft are located on the same horizontal plane; when the output shaft of the reduction motor rotates, the gear on the hollow shaft and the hollow shaft are driven to rotate through the chain; The hollow shaft drives the sorting turntable to rotate.

[0008] In some embodiments, the sorting turntable drive device further includes a speed regulator; the speed regulator is used to adjust the set rotation speed to the input rotation speed of the reduction motor.

[0009] In some embodiments, a movable reflector is provided at the image detection position; when in operation, the reflector is arranged upwards and at a predetermined inclination angle with the horizontal plane, so as to reflect the abdominal image of the crab; The camera of the video camera is aimed at the reflector to shoot, and the image of the crab abdomen is captured from the reflector.

[0010] In some embodiments, the controller is further used to control the sorting turntable drive device to rotate after the image detection is completed, so that the crab moves from the image detection position to the weighing area; The air pump is controlled to stop working so that the crab falls from the air suction head into the target sub-area in the weighing turntable.

[0011] In some embodiments, it also includes a weighing turntable driving device, multiple groups of baffle driving devices and pressure sensors respectively connected to the above controller; The weighing turntable includes a plurality of sub-areas; a group of baffles, hinges, the baffle driving device and the pressure sensor are arranged at the bottom of each sub-area; The controller is further used to obtain a pressure value detected by a pressure sensor disposed in the target sub-region having the crab; Determine the end of the interval to which the above pressure value belongs; Determine the target slideway according to the correspondence between the preset interval segments and the sequence numbers of the target slideways; Controlling the weighing turntable driving device to drive the weighing turntable to rotate, and rotating the sub-area with crabs to the position of the target slideway; The baffle driving device at the bottom of the target sub-area is controlled to operate, driving the baffle to open, so that the crabs fall from the target sub-area and slide along the target slideway into the target bucket, so as to achieve classification according to the weight of the crabs.

[0012] In some embodiments, the weighing turntable includes a plurality of sub-areas; a group of baffles, hinges, the baffle driving device and the pressure sensor are arranged at the bottom of each sub-area; For any set of baffles, hinges and baffle drive devices; The baffle is fixed on the first page of the hinge, and the second page of the hinge is fixed on the column; A hinge support is provided at the bottom of the baffle; The baffle driving device comprises: a baffle driving motor, a first driving hinge and a second driving hinge; The output shaft of the baffle driving motor is connected to the first end of the first driving hinge; The second end of the first driving hinge is movably connected to the first end of the second driving hinge; The second end of the second driving hinge is movably connected to the hinge support; When the baffle driving motor rotates in the first direction, the first driving hinge and the second driving hinge are driven to move toward the bottom direction, so that the baffle is opened; When the baffle driving motor rotates in the second direction, the first driving hinge and the second driving hinge are driven to move in the top direction, so that the baffle is closed.

[0013] In some embodiments, a display screen connected to the controller is also included. The controller is further used to determine the weight value of the crab according to the pressure value; The display screen is used to display the weight value of the crab.

[0014] The technical solution provided by the embodiments of the present disclosure may have the following beneficial effects: Compared with the prior art, the technical solution of the present application can determine the type of the crab according to the image features of the abdominal image by using a pre-trained image recognition algorithm; the sorting turntable is controlled to move different types of crabs to different areas, thereby realizing automatic sorting by type, improving the efficiency of crab classification work and saving manpower.

[0015] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present disclosure and, together with the description, serve to explain the principles of the present disclosure.

[0017] Figure 1 is a block diagram of a crab automatic classification system based on image recognition according to an exemplary embodiment; Figure 2 is a structural diagram of a sorting turntable according to an exemplary embodiment; Figure 3 is an image of a crab abdomen according to an exemplary embodiment; Figure 4 is another crab abdomen image shown according to an exemplary embodiment; Figure 5 is a diagram showing the internal structure of a sorting turntable according to an exemplary embodiment; Figure 6 is a structural diagram of a sorting turntable driving device according to an exemplary embodiment; Figure 7 is a structural diagram of image detection according to an exemplary embodiment; Figure 8 is a structural diagram of a weighing turntable according to an exemplary embodiment; Fig. 9 is a structural diagram of a baffle driving device according to an exemplary embodiment; Fig.10 is a structural diagram of another baffle driving device according to an exemplary embodiment; Fig.11 is a structural diagram of another baffle driving device according to an exemplary embodiment. DETAILED DESCRIPTION

[0018] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, the same numbers in different drawings represent the same or similar elements unless otherwise indicated. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure as detailed in the appended claims.

[0019] It should be noted that all actions of acquiring signals, information or data in this application are carried out in compliance with the relevant data protection laws and policies of the country where they are located, and with the authorization given by the owner of the corresponding device.

[0020] Based on this, this application proposes an automatic crab classification system based on image recognition, see the attached Figure 1 and attached Figure 2 ; A crab automatic classification system based on image recognition, comprising: A controller 01, a sorting turntable driving device 02, a camera 03, and multiple groups of photoelectric sensors 04 and an air pump 05 respectively connected to the controller 01.

[0021] The sorting turntable drive device 02 is located in the control cabinet 23 below the sorting turntable.

[0022] The controller 01 is used to control the air pump 05 to start in response to the crab detection electrical signal sent by the photoelectric sensor 04 so as to adsorb the crab on the suction head 21 .

[0023] The sorting turntable driving device 02 is controlled to operate, so as to drive the sorting turntable 26 to rotate, and the suction head 21 and the crabs adsorbed on the suction head 21 are rotated and driven to a predetermined image detection position.

[0024] At the image detection position 25, the camera 03 is controlled to be turned on; the crab is photographed; and the abdomen image of the crab is obtained.

[0025] According to the image features of the above-mentioned abdomen image, a pre-trained image recognition algorithm is used to determine the type of the above-mentioned crab; wherein the above-mentioned type includes male crabs or female crabs.

[0026] In this embodiment, the image recognition model can be implemented using a neural network. The image recognition model can use a compact model. The advantages of this model are memory saving and high operating efficiency. The principle of the model is a SqueezeNet network. The SqueezeNet network includes a squeeze layer and an expand layer. The squeeze layer is composed of 1×1 convolutions, and the expand layer uses 1×1 and 3×3 convolutions to form a multi-branch structure to extract input features, thereby improving the accuracy of the network.

[0027] First, collect a large number of sample images of river crabs, then label and classify the river crabs, and save them as .dltp files for pre-training. For example, you can collect 4,000 river crab images for image labeling. Male and female crabs have different abdominal patterns, so you can use the abdominal image features to identify male and female crabs.

[0028] See attached Figure 2 The abdomen of the female crab is shown with Figure 3 is the abdomen image of a male crab. From the image, we can see that the abdomen image features of male and female crabs are significantly different.

[0029] During model training, the image of the river crab to be identified is input into the model, and the identification result is compared with the standard result. If it is inaccurate, the parameters of the model are adjusted and training continues.

[0030] The image recognition model can be trained on a large scale, and after a lot of training, the accuracy of the image recognition model can be improved. In a specific recognition experiment, the correct prediction of female crabs accounted for 99.35%; the correct prediction of male crabs accounted for 98.39%, and the model prediction results were very consistent with the actual situation.

[0031] In response to the first recognition result, the sorting turntable driving device 02 is controlled to operate to move the crab to a predetermined first area.

[0032] The first area can be flexibly set, for example, it can be set at the left side of the sorting turntable 26.

[0033] The first recognition result is that the crab is a male crab.

[0034] In response to the second recognition result, the sorting turntable drive device 02 is controlled to move the crab to a predetermined second area.

[0035] The second area can be flexibly set, for example, it can be set at the right side of the sorting turntable 26.

[0036] The second identification result is that the crab is a female crab.

[0037] In this embodiment, the controller 01 includes a processor and a memory. The processor is provided with a crab classification, identification and sorting module 0111 and a weighing module 0112.

[0038] Among them, the crab classification, identification and sorting module 0111 is used to realize crab classification, identification and sorting.

[0039] The weighing module 0112 is used to weigh the crabs after they have been classified and sorted.

[0040] The technical solution of the present application can determine the type of the crab according to the image features of the abdominal image by using a pre-trained image recognition algorithm; control the sorting turntable to move different types of crabs to different areas, thereby realizing automatic sorting by type, improving the efficiency of crab classification work and saving manpower.

[0041] In some embodiments, see Appendix Figure 4 and attached Figure 5 , multiple groups of photoelectric sensors 04 and air pumps 05 are located on the above-mentioned sorting turntable 26.

[0042] The sorting turntable is also provided with multiple groups of rotating arms 24, suction heads 21, air pipes and photoelectric sensors 04. The multiple rotating arms 24 are evenly distributed. For any group of rotating arms 24, suction heads 21, air pipes and air pumps 05: the suction heads 21 are provided at the ends of the rotating arms 24; the suction heads 21 are connected to the air pumps 05 through the air pipes. The photoelectric sensor 04 is provided on the bottom surface of the rotating arms 24, facing the suction heads 21. When there are crabs under the suction heads 21, they can be detected by the photoelectric sensor 04.

[0043] When a crab enters the detection area of ​​the photoelectric sensor 04, the light will be blocked by the crab, causing the intensity of the light received by the receiver to change. The receiver converts the received light signal into an electrical signal, and the converted electrical signal is sent to the detection circuit of the photoelectric sensor 04 for processing. The detection circuit determines whether an object exists according to the strength of the signal, and outputs a switch signal to the controller 01. When the photoelectric sensor 04 detects an object, the controller 01 controls the vacuum pump to start or stop, thereby realizing the control of the vacuum pump. The suction head 21 is connected to the vacuum pump through a hose. When the photoelectric sensor 04 recognizes the presence of a crab, the vacuum pump starts to extract the air in the pipe, so that negative air pressure is generated inside the pipe, that is, a vacuum state. This process will reduce the air pressure in the suction head 21, forming a low-pressure area. When the air pressure inside the suction cup is lower than the atmospheric pressure of the external environment, due to the effect of the air pressure difference, the crab will be firmly adsorbed on the suction head to achieve the suction of the crab.

[0044] In this embodiment, the number of the plurality of rotating arms 24 can be flexibly set, and in one embodiment, can be set to 6. Each rotating arm 24 corresponds to an air suction head 21, an air pipe and an air pump 05. The air pump 05 can be arranged on the inner side of the sorting turntable. The rotating arm 24 is arranged on the outer side of the sorting turntable. The rotating arm 24 can be arranged as a hollow structure for passing the air pipe.

[0045] In some embodiments, a hollow frame for fixing the air pipe is further provided in the middle of the rotating arm 24. The air pipe passes through the hollow frame to facilitate fixing.

[0046] The hollow frame of the present application can be a rectangular parallelepiped structure, removing the materials in the four side elevations and the top surface, and only retaining the frame ribs. This design is conducive to weight reduction, cost saving, and avoiding the rotating arm 24 being too heavy to affect the control accuracy.

[0047] The technical solution of the present application can fix the air pipe by setting a hollow frame. Such a design is conducive to improving the simplicity of the entire product, improving the aesthetics, and avoiding the situation where the air pipe is too long and becomes messy. In addition, setting a hollow frame to fix the air pipe can also keep the air pipe as straight as possible. A straight air pipe is more conducive to the flow of gas. If the air pipe is in a bent state and the air pressure is too high, the fixed end of the air pipe may fall off from the suction head.

[0048] In some embodiments, see Appendix Figure 6 The sorting turntable driving device 02 is located in the cabinet below the sorting turntable.

[0049] The sorting turntable driving device 02 at least includes: a hollow shaft 52, a chain 55, a reduction motor 56, a first bearing and a second bearing.

[0050] The hollow shaft 52 is vertically arranged in the cabinet below the sorting turntable.

[0051] The first bearing is arranged in the top end cover of the lower cabinet.

[0052] The first end of the hollow shaft 52 is sleeved in the inner sleeve of the first bearing, and the second end of the hollow shaft 52 is fixed on the bearing seat 53 of the second bearing.

[0053] The output shaft of the reduction motor 56 and the gear sleeved on the hollow shaft 52 are located on the same horizontal plane; when the output shaft of the reduction motor 56 rotates, the gear 51 on the hollow shaft 52 and the hollow shaft 52 are driven to rotate through the chain 55. Specifically, the gear 51 is driven to rotate by the chain 55, and the gear 51 drives the hollow shaft 52 to rotate.

[0054] The hollow shaft 52 drives the sorting turntable to rotate.

[0055] In some embodiments, the sorting turntable drive device 02 further includes a speed regulator 54 ; the speed regulator 54 is used to adjust the set rotation speed to the input rotation speed of the reduction motor 56 .

[0056] The speed regulator 54 adjusts the input speed of the reduction motor 56 according to the set speed, and the reduction motor 56 then reduces the output speed according to the reduction ratio of the two sprockets. The speed regulator 54 is responsible for adjusting the speed of the motor, and the reduction motor 56 is responsible for reducing the speed and increasing the torque, and transmitting the power to the hollow shaft 52 through the chain 55. This cooperation can achieve precise control of the rotation speed. Since the hollow shaft 52 is integrally connected to the weighing turntable, precise control of the rotation speed of the turntable is achieved. The bearing seat 53 cooperates with the hollow shaft 52 to fix the hollow shaft 52 in an accurate position, ensure the correct axial distance between the inner and outer rings of the bearing and the end cover, and bear the axial and radial loads of the machine, as well as support the rotating shaft. The flange part and the groove part on the bearing seat 53 can reduce the transmission of vibration, reduce the noise of the machine, and improve the running stability of the equipment.

[0057] In some embodiments, see Appendix Figure 7 A movable reflector 41 is provided at the above-mentioned image detection position; when working, the above-mentioned reflector is arranged upward and has a predetermined inclination angle with the horizontal plane, so as to reflect the abdominal image of the above-mentioned crab; the camera of the above-mentioned camera 03 is aimed at the above-mentioned reflector 41 to shoot, and the abdominal image of the crab is captured from the above-mentioned reflector.

[0058] In some embodiments, the controller is also used to control the sorting turntable drive device to rotate after the image detection is completed, so that the crab moves from the image detection position to the weighing area.

[0059] The air pump 03 is controlled to stop working so that the crab falls from the air suction head 21 into the target sub-area in the weighing turntable.

[0060] In some embodiments, it also includes a weighing turntable driving device 06, multiple groups of baffle driving devices and pressure sensors 08 respectively connected to the above-mentioned controller.

[0061] The structure of the weighing turntable driving device 06 is the same as that of the sorting turntable driving device, and the motor is used to drive the central axis to rotate, thereby driving the weighing turntable to rotate.

[0062] The weighing carousel consists of several sub-areas.

[0063] For example, the turntable can be divided into six areas, each of which works independently, and multiple crabs can be weighed and sorted at the same time, which improves the speed and efficiency of sorting, reduces the mutual influence between crabs, and improves the accuracy of weight detection, thereby ensuring the accuracy of the sorting results.

[0064] A group of baffles, hinges, the baffle driving device and the pressure sensor 08 are arranged at the bottom of each sub-area.

[0065] The pressure sensor 08 is drilled and installed at the bottom of the turntable. After the sensor is installed in place, the positioning bolts are tightened to prevent looseness. The cable is led out from the pressure sensor 08 to complete the processing of the connecting line and the output line. This sensor can detect the pressure changes acting on the baffle.

[0066] The controller is further used to obtain a pressure value detected by a pressure sensor 08 disposed in the target sub-area having the crab.

[0067] Determine the interval segment to which the above pressure value belongs.

[0068] The target slideway is determined according to the correspondence between the preset interval segments and the serial numbers of the target slideways.

[0069] For example, some crabs weigh less than or equal to 0.5 kg, and this weight range corresponds to the first pressure interval segment, and the first pressure interval segment corresponds to the first slide.

[0070] Some crabs weigh more than 0.5 kg and less than or equal to 1 kg. This weight range corresponds to the second pressure interval, and the second pressure interval corresponds to the second slide.

[0071] Some crabs weigh more than 1 kg and less than or equal to 1.5 kg. This weight range corresponds to the third pressure interval, and the third pressure interval corresponds to the third slide.

[0072] Some crabs weigh more than 1.5 kg and less than or equal to 2 kg. This weight range corresponds to the fourth pressure interval, and the fourth pressure interval corresponds to the fourth slide.

[0073] The weighing turntable driving device 06 is controlled to operate, drive the weighing turntable to rotate, and rotate the sub-area with crabs to the position of the target slideway.

[0074] The baffle driving device at the bottom of the target sub-area is controlled to operate, driving the baffle to open, so that the crabs fall from the target sub-area and slide along the target slideway into the target bucket, so as to achieve classification according to the weight of the crabs.

[0075] In the technical solution of the present application, the interval to which the above pressure value belongs is determined. The target slide is determined according to the correspondence between the preset interval and the serial number of the target slide. The crab is moved to the target slide and then automatically falls into the target slide, thereby achieving the effect of classifying the crabs according to weight.

[0076] In some embodiments, when the pressure detected by the pressure sensor is greater than a preset pressure threshold, the damper is controlled to open.

[0077] In this embodiment, if the pressure detected by the pressure sensor is greater than the preset pressure threshold, the baffle is directly controlled to open and the crab falls into the slide. If the pressure detected by the pressure sensor is less than or equal to the preset pressure threshold, the baffle is not opened, and the weighing turntable can be controlled to rotate to move the crab to the corresponding slide area, and then the baffle is controlled to open.

[0078] In some embodiments, see Appendix Figure 8 and attached Fig. 9 The weighing turntable 61 includes a plurality of sub-areas; a group of baffles 09, hinges 011, the baffle driving device 07 and the pressure sensor 08 are arranged at the bottom of each sub-area.

[0079] For any set of baffle 09, hinge 011 and baffle 09 driving device 07. The baffle 09 is fixed on the first page of the hinge 011, and the second page of the hinge 011 is fixed on the column.

[0080] A hinge support 010 is provided at the bottom of the baffle 09. The baffle driving device 07 comprises: a baffle driving motor 071, a first driving hinge 072 and a second driving hinge 073. The output shaft of the baffle driving motor 071 is connected to the first end of the first driving hinge 072. The second end of the first driving hinge 072 is movably connected to the first end of the second driving hinge 073. The second end of the second driving hinge 073 is movably connected to the hinge support 010.

[0081] When the baffle driving motor 071 rotates in the first direction, the first driving hinge 072 and the second driving hinge 073 are driven to move toward the bottom direction, so that the baffle 09 is opened. When the baffle driving motor 071 rotates in the second direction, the first driving hinge 072 and the second driving hinge 073 are driven to move toward the top direction, so that the baffle 09 is closed.

[0082] In some embodiments, a display screen connected to the controller is further included, and the controller is further used to determine the weight value of the crab according to the pressure value. The display screen is used to display the weight value of the crab.

[0083] In this embodiment, the controller can convert the pressure value into the weight value of the crab, and send the weight value to the display screen for display. In the display screen, multiple display areas can be set, and each display area displays the weight of the crab in each area on the weighing turntable.

[0084] The following example illustrates the specific operation method: Place the crab's back on the suction head. When the crab enters the detection area of ​​the photoelectric sensor, the light will be blocked by the crab, causing the light intensity received by the receiver to change. At this time, the switch signal of the photoelectric sensor will trigger the vacuum pump to start. The suction head is connected to the vacuum pump through a hose. When the photoelectric sensor recognizes the presence of a crab, the vacuum pump starts to extract the air in the pipe, causing negative air pressure inside the pipe, that is, a vacuum state. This process will reduce the air pressure in the suction head and form a low-pressure area. When the air pressure inside the suction cup is lower than the atmospheric pressure of the external environment, due to the effect of the air pressure difference, the crab will be firmly attached to the suction head to achieve the suction of the crab. At this time, the sorting turntable always keeps a slow rotation state. When the crab is in the image detection position, the reflector is used to reflect the image of the crab's abdomen into the camera lens so that the camera can capture a clear image of the crab's abdomen. The design of the reflector at a certain angle to the camera enables the camera to capture the image of the crab's abdomen from below, thereby avoiding the problem of dust falling on the camera during long-term shooting and the situation where the crab's splashing liquid drips onto the camera. After the male and female crabs are sorted, the turntable will shut down the vacuum pump in the area corresponding to the male and female crabs and let the crabs fall. The crabs will fall into the corresponding weighing turntable according to their gender. A pressure sensor is installed at the bottom of the weighing turntable. The crabs are divided into several levels according to their weight, and the baffle can be controlled to open. By dividing the turntable into six areas, each area works independently, multiple crabs can be weighed and sorted at the same time, improving the speed and efficiency of sorting, reducing the mutual influence between crabs, and improving the accuracy of weight detection, thereby ensuring the accuracy of the sorting results.

[0085] The automatic crab sorting device of the present application can simulate the process of manual sorting, collect crab abdomen images and weight data, and quickly classify and grade crabs. It can identify the abdominal features of crabs through high-definition cameras, distinguish male and female characteristics, and use pressure sensors to detect weight and grade, greatly improving the efficiency and accuracy of sorting.

[0086] Those skilled in the art will readily appreciate other embodiments of the present disclosure after considering the specification and practicing the present disclosure. This application is intended to cover any variations, uses or adaptations of the present disclosure, which follow the general principles of the present disclosure and include common knowledge or customary technical means in the art that are not disclosed in the present disclosure. The specification and examples are to be considered as exemplary only, and the true scope and spirit of the present disclosure are indicated by the following claims.

[0087] It should be understood that the present disclosure is not limited to the exact structures that have been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof. The scope of the present disclosure is limited only by the appended claims.

Claims

1. A crab automatic classification system based on image recognition, characterized in that: include: A controller, a sorting turntable drive device, a camera, and multiple groups of photoelectric sensors and an air pump respectively connected to the controller; The controller is used to control the air pump to start in response to the crab detection electrical signal sent by the photoelectric sensor, so as to adsorb the crab on the suction head; Controlling the sorting turntable driving device to drive the sorting turntable to rotate, and rotating the suction head and the crabs adsorbed on the suction head to a predetermined image detection position; At the image detection position, controlling the camera to turn on; photographing the crab; obtaining an image of the crab's abdomen; Determine the type of the crab using a pre-trained image recognition algorithm according to the image features of the abdomen image; wherein the type includes a male crab or a female crab; In response to the first recognition result, controlling the sorting turntable driving device to move the crab to a predetermined first area; The first recognition result is that the crab is a male crab; In response to the second recognition result, controlling the sorting turntable driving device to move the crab to a predetermined second area; The second identification result is that the crab is a female crab.

2. The automatic crab classification system based on image recognition according to claim 1 is characterized in that: Multiple groups of photoelectric sensors and air pumps are located on the sorting turntable; The sorting turntable is also provided with multiple groups of rotating arms, suction heads, air pipes and photoelectric sensors; Among them, multiple rotating arms are evenly distributed; For any set of rotating arm, suction head, air pipe and air pump: The air suction head is arranged at the end of the rotating arm; the air suction head is connected to the air pump through the air pipe.

3. The automatic crab classification system based on image recognition according to claim 2 is characterized in that: A hollow frame for fixing the air pipe is also provided in the middle of the rotating arm; The trachea passes through the hollow frame to facilitate fixation.

4. The automatic crab classification system based on image recognition according to claim 1 is characterized in that: The sorting turntable driving device is located in a cabinet below the sorting turntable; The sorting turntable driving device at least includes: a hollow shaft, a chain, a reduction motor, a first bearing and a second bearing; The hollow shaft is vertically arranged in the cabinet below the sorting turntable; The first bearing is arranged in the top end cover of the lower cabinet; The first end of the hollow shaft is sleeved in the inner sleeve of the first bearing. The second end of the hollow shaft is fixed to the bearing seat of the second bearing; The output shaft of the reduction motor and the gear sleeved on the hollow shaft are located on the same horizontal plane; when the output shaft of the reduction motor rotates, the gear on the hollow shaft and the hollow shaft are driven to rotate through the chain; The hollow shaft drives the sorting turntable to rotate.

5. The automatic crab classification system based on image recognition according to claim 1 is characterized in that: The sorting turntable driving device also includes a speed regulator; the speed regulator is used to adjust the set rotation speed to the input rotation speed of the reduction motor.

6. The automatic crab classification system based on image recognition according to claim 1 is characterized in that: A movable reflector is provided at the image detection position; when in operation, the reflector is arranged upwards and has a predetermined inclination angle with the horizontal plane, so as to reflect the abdominal image of the crab; The camera of the video camera is aimed at the reflector to shoot, and the image of the crab abdomen is captured from the reflector.

7. The automatic crab classification system based on image recognition according to claim 1 is characterized in that: The controller is also used to control the sorting turntable drive device to operate after the image detection is completed, so as to drive the sorting turntable to rotate, so that the crab moves from the image detection position to the weighing area; The air pump is controlled to stop working so that the crab falls from the air suction head into a target sub-area in the weighing turntable.

8. The automatic crab classification system based on image recognition according to claim 7 is characterized in that: It also includes a weighing turntable driving device, a plurality of baffle driving devices and a pressure sensor respectively connected to the controller; The weighing turntable includes a plurality of sub-areas; a group of baffles, hinges, the baffle driving device and the pressure sensor are arranged at the bottom of each sub-area; The controller is further used to obtain a pressure value detected by a pressure sensor disposed in the target sub-area having the crab; determining the end of the interval to which the pressure value belongs; Determine the target slideway according to the correspondence between the preset interval segments and the sequence numbers of the target slideways; Controlling the weighing turntable driving device to drive the weighing turntable to rotate, and rotating the sub-area with crabs to the position of the target slideway; The baffle driving device at the bottom of the target sub-area is controlled to move, driving the baffle to open, so that the crabs fall from the target sub-area and slide along the target slideway into the target bucket, so as to achieve classification according to the weight of the crabs.

9. The automatic crab classification system based on image recognition according to claim 8 is characterized in that: The weighing turntable includes a plurality of sub-areas; a group of baffles, hinges, the baffle driving device and the pressure sensor are arranged at the bottom of each sub-area; For any set of baffles, hinges and baffle drive devices; The baffle is fixed on the first page of the hinge, and the second page of the hinge is fixed on the column; A hinge support is provided at the bottom of the baffle; The baffle driving device comprises: a baffle driving motor, a first driving hinge and a second driving hinge; The output shaft of the baffle driving motor is connected to the first end of the first driving hinge; The second end of the first driving hinge is movably connected to the first end of the second driving hinge; The second end of the second driving hinge is movably connected to the hinge support; When the baffle driving motor rotates in a first direction, the first driving hinge and the second driving hinge are driven to move toward the bottom direction, so that the baffle is opened; When the baffle driving motor rotates in the second direction, the first driving hinge and the second driving hinge are driven to move in the top direction, so that the baffle is closed.

10. The automatic crab classification system based on image recognition according to claim 8 is characterized in that: Also includes a display screen connected to the controller, The controller is also used to determine the weight value of the crab according to the pressure value; The display screen is used to display the weight value of the crab.

Images