A frozen dead cage silkworm identification method based on machine vision

Abstract

The present application relates to silkworm cocoon identification technical field, especially a kind of frozen dead cage silkworm identification method based on machine vision, comprising the following steps: A1, the first image of frozen silkworm is obtained using industrial camera, and the appearance characteristics of the frozen silkworm in the first image are identified by machine vision identification algorithm to determine whether the frozen silkworm is dead cage silkworm;A2, the dead cage silkworm of the frozen silkworm in step A1 is removed to obtain the primary selected silkworm.B1, the primary selected silkworm of step A1 is cooled by air draft using air draught fan, to obtain thawed silkworm;B2, the second image of the thawed silkworm in step B2 is obtained using industrial camera, and the thawed silkworm in the second image is identified by machine vision identification algorithm to determine whether the thawed silkworm is dead cage silkworm;B3, the dead cage silkworm in step B2 is removed to obtain intact silkworm.The present application can solve the problem that dead cage cocoon cannot be removed in time, and avoid the influence of dead cage cocoon on the water bath environment of silk reeling.

Description

Background Technology

[0001] This invention relates to the field of silkworm cocoon identification technology, and in particular to a machine vision-based method for identifying frozen dead silkworms. Technical Field

[0002] Silkworm cocoon reeling methods are divided into dry cocoon reeling and fresh cocoon reeling. Currently, silk reeling enterprises in Guangxi have begun to widely adopt fresh cocoon reeling, mainly because the process and technology of fresh cocoon reeling are relatively simpler than those of dry cocoon reeling. In particular, fresh cocoon pupae have high application value, which greatly increases the value of by-products.

[0003] The disadvantages of reeling silk from fresh cocoons compared to reeling silk from dried cocoons are also quite obvious. The main point is that the quality of raw silk from fresh cocoons is relatively low, fresh cocoons are not resistant to boiling, and dead cocoons release a large amount of biomass in the water bath environment of silk reeling, which affects the effect of drugs on the relaxation and cleaning of cocoon silk.

[0004] Dead cocoons refer to silkworm cocoons where the rotten fluids from the silkworm or pupa have seeped into the cocoon layer. Because fresh cocoons are frozen during reeling and refrigeration, most dead cocoons haven't yet exuded pus before it freezes, making them appear identical to good cocoons. Furthermore, the subsequent cocoon selection process, aimed at preserving the byproduct silkworm pupae, is conducted under frozen conditions, further complicating the identification of dead cocoons.

[0005] Because the preservation process of fresh cocoons often requires freezing, the freezing environment is generally between -20 and -5 degrees Celsius, depending on the production characteristics of different enterprises. In this low-temperature environment, the necrotic pupae of dead cocoons are frozen, so their appearance often remains the same as good cocoons. As they leave the frozen environment, the pupae begin to thaw, and the necrotic pus slowly appears in the cocoon layer. This results in dead cocoons not being completely removed in a timely manner, especially when the cocoon quality is poor and the proportion of dead cocoons is high. A large number of missed dead cocoons will seriously affect the water bath environment for silk reeling. Summary of the Invention

[0006] The present invention aims to solve at least one of the technical problems mentioned in the background art above, and provides a machine vision-based method for identifying frozen dead silkworm cocoons, which solves the problem that dead cocoons cannot be removed in time and avoids dead cocoons affecting the water bath environment of silk reeling.

[0007] To achieve the above objectives, the technical method employed in this invention is as follows:

[0008] A machine vision-based method for identifying frozen, dead silkworms includes primary and secondary identification.

[0009] The initial identification includes the following steps:

[0010] A1. An industrial camera is used to acquire a first image of the frozen silkworm. A machine vision recognition algorithm is used to identify the frozen silkworm based on its appearance features in the first image to determine whether the frozen silkworm is a dead silkworm.

[0011] A2. After removing the dead silkworms from the frozen silkworms described in step A1, the initial selected silkworms are obtained.

[0012] The secondary identification includes the following steps:

[0013] B1. Use an exhaust fan to ventilate the silkworms initially selected in step A1 with cold air to obtain thawed silkworms;

[0014] B2. Use an industrial camera to acquire a second image of the thawed silkworm described in step B2, and use a machine vision recognition algorithm to identify the thawed silkworm based on its external dimensions in the second image to determine whether the thawed silkworm is a dead silkworm.

[0015] B3. After removing the dead silkworms described in step B2, intact silkworms are obtained.

[0016] Furthermore, in step A1, the appearance features include dead cocoons, cotton cocoons, mouth cocoons, moldy cocoons, and cocoons with multiple defects.

[0017] Furthermore, the exhaust fan has a wind speed of 4-5 m / s and a wind temperature of 4-10℃.

[0018] Furthermore, the step of identifying the external dimensions of the thawed silkworm in the second image includes:

[0019] C1: Binarize the second image, extract the binarized second image, and obtain the binarized image;

[0020] C2: Calculate the aspect ratio of the binarized image and compare the aspect ratio of the binarized image with a preset threshold range. Determine the thawed silkworms corresponding to the binarized image that are outside the threshold range as dead silkworms.

[0021] Furthermore, it also includes an identification device used in the method for identifying frozen dead silkworms, the identification device comprising a conveyor belt, a first storage box, and a second storage box.

[0022] The conveyor belt is a U-shaped structure with openings at both ends, and the horizontal heights at both ends of the conveyor belt are different. The first storage box is located above one side of the conveyor belt. The first storage box has a recessed first cavity, and the bottom of the first storage box has a first discharge hole communicating with the first cavity. The first discharge hole has a discharge mechanism to allow a single frozen silkworm to be discharged into the conveyor belt through the discharge mechanism.

[0023] The second storage box is located at the opening of the conveyor belt, with the upper end of the conveyor belt above the second storage box and the lower end of the conveyor belt below the second storage box. The top of the second storage box has a recessed second cavity, and the bottom of the second storage box slopes from the upper end to the lower end of the conveyor belt. The side of the second storage box near the lower end of the conveyor belt has a first discharge hole communicating with the second cavity. The first discharge hole has a discharge mechanism to allow a single thawed silkworm to be discharged into the conveyor belt through the discharge mechanism.

[0024] The second storage box is equipped with an exhaust pipe, one end of which is connected to the second storage box through several pipes, and the exhaust fan is located at the end of the exhaust pipe away from the second storage box;

[0025] The industrial camera is positioned above the conveyor belt, and is located on one side of the first storage box along the direction of travel of the conveyor belt.

[0026] Furthermore, the discharge mechanism includes a housing, a discharge ball, and a discharge motor. The housing has a through discharge hole, which communicates with the corresponding first discharge hole. The discharge ball is located inside the discharge hole and is rotatably connected to the housing. The discharge ball has a recessed storage groove, allowing a single frozen silkworm or a single thawed silkworm to enter the storage groove through the discharge hole. The discharge motor drives the discharge ball so that the storage groove can face the corresponding first discharge hole or the first discharge hole, or face the conveyor belt.

[0027] Furthermore, the industrial camera is fixed above the conveyor belt by a support frame, and supplementary lights are provided on both sides of the support frame.

[0028] Furthermore, a screening mechanism is provided between the industrial camera and the second storage box. The screening mechanism includes a pneumatic push rod and a processing barrel. The pneumatic push rod and the processing barrel are arranged opposite each other on both sides of the conveyor belt. The pneumatic push rod is located on one side of the conveyor belt, and the drive rod of the pneumatic push rod faces the conveyor belt and is provided with a push plate, so that when the pneumatic push rod is extended, the push plate can push the dead silkworms into the processing barrel.

[0029] Furthermore, the conveyor belts on both sides of the industrial camera are respectively equipped with a first infrared sensor and a second infrared sensor. The first infrared sensor is located on the side of the industrial camera closer to the first storage box, so as to identify silkworms entering the industrial camera.

[0030] The second infrared sensor is located between the industrial camera and the screening mechanism, and is used to identify silkworms passing through the screening mechanism.

[0031] Furthermore, it also includes a third storage box, which is located on the side of the second storage box away from the conveyor belt. The third storage box is slidably provided with a guide groove, which is located between the third storage box and the second storage box. The guide groove is driven by a receiving motor to slide towards the third storage box. One end of the guide groove is located at the bottom of the conveyor belt, and the other end is located above the third storage box, or the guide groove is moved away from the second storage box.

[0032] The beneficial effects of this invention are:

[0033] 1. In the first identification stage, the appearance characteristics of frozen silkworms are identified to screen out obviously defective cocoons, thus obtaining the initial selected silkworms. In the second identification stage, the initial selected silkworms are thawed. The pus from dead cocoons seeps out, making the size of the dead cocoons different from that of intact silkworms. Thus, the machine vision recognition algorithm can determine whether the thawed silkworms are dead cocoons. At the same time, the invention uses cold air ventilation to thaw, which accelerates the appearance of dead cocoons, shortens the time of the entire process, and ensures the freshness of the silkworm pupae, which is conducive to the screening out of dead cocoons. Without affecting the silkworm pupae, the quality of raw silk reeling is improved, and the overall production efficiency and benefits of the enterprise are increased.

[0034] 2. Since dead silkworms typically ooze pus from their head, top, and sides after thawing, this invention binarizes the second image. In the second image, the black pixels represent the pus-oozing parts of the thawed silkworm, and the white pixels represent the intact parts. By extracting the pixels of the second image, the aspect ratio of the thawed silkworm can be obtained, thus accurately determining whether the thawed silkworm is a dead silkworm.

[0035] 3. By placing frozen silkworms in a first storage box and conveying them via a conveyor belt, the frozen silkworms pass through an industrial camera, which captures a first image to complete the first identification. The initially selected silkworms are then transported via conveyor belt to a second storage box. A fan in the second storage box evacuates the initially selected silkworms with cold air, and the thawed silkworms are then conveyed via conveyor belt to the industrial camera for a second image capture, thus completing the second identification. The identification device of this invention automatically identifies dead silkworms, effectively improving identification efficiency. Attached Figure Description

[0036] Figure 1 This is a flowchart of a preferred embodiment of the present invention, which describes a machine vision-based method for identifying frozen, dead silkworms.

[0037] Figure 2 This is a schematic diagram of the identification device structure of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0038] Figure 3 This is a schematic diagram of the first storage box structure of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0039] Figure 4 This is a schematic diagram of the second storage box structure of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0040] Figure 5 This is a schematic diagram of the third storage box structure of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0041] Figure 6 This is a schematic diagram of the discharge mechanism structure of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0042] Figure 7 This is a schematic diagram of the industrial phase structure of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0043] Figure 8 This is a binarized image of a machine vision-based method for identifying frozen dead silkworms according to a preferred embodiment of the present invention.

[0044] Figure 9 This is a shape and size recognition diagram of a preferred embodiment of the machine vision-based method for identifying frozen dead silkworms according to the present invention.

[0045] In the diagram, 1-conveyor belt, 101-industrial camera, 102-exhaust fan, 103-fill light, 11-support frame, 121-first transmission section, 122-second transmission section, 123-third transmission section, 13-pneumatic push rod, 131-processing tank, 132-push plate, 2-first storage box, 201-first cavity, 202-first discharge port, 3-second storage box, 301-second cavity, 302-first discharge port Hole, 303-vent hole, 31-exhaust pipe, 311-pipe, 4-outer shell, 401-discharge through hole, 41-discharge ball, 411-temporary storage tank, 42-discharge motor, 431-first gear, 432-second gear, 51-first infrared sensor, 52-second infrared sensor, 6-third storage box, 61-guide groove, 611-slide rail, 612-rack, 62-receiving motor, 621-drive gear. Detailed Implementation

[0046] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] It should be noted that when a component is described as "fixed to" another component, it can be directly on the other component or may have a component in between. When a component is considered "connected to" another component, it can be directly connected to the other component or may have a component in between. When a component is considered "set on" another component, it can be directly set on the other component or may have a component in between. The terms "vertical," "horizontal," "left," "right," and similar expressions used in this document are for illustrative purposes only.

[0048] Unless otherwise defined, 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. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0049] Please also see Figures 1 to 9 A preferred embodiment of the present invention provides a machine vision-based method for identifying frozen, dead silkworms, comprising primary identification and secondary identification.

[0050] A single identification process includes the following steps:

[0051] A1. An industrial camera 101 is used to acquire the first image of the frozen silkworm. A machine vision recognition algorithm is used to identify the frozen silkworm based on its appearance features in the first image to determine whether the frozen silkworm is a dead silkworm.

[0052] In step A1, the appearance characteristics include dead cocoons, cotton cocoons, mouth cocoons, moldy cocoons, and cocoons with multiple defects.

[0053] A2. After removing the dead silkworms from the frozen silkworms in step A1, the initial selected silkworms are obtained.

[0054] Secondary identification includes the following steps:

[0055] B1. Use a blower 102 to ventilate the silkworms initially selected in step A1 with cold air to obtain thawed silkworms. The wind speed of the blower 102 is 4-5 m / s and the wind temperature of the blower 102 is 4-10℃. In this embodiment, the blower 102 can ventilate for 5 minutes to accelerate the seepage of pus from dead cocoons, causing the cocoon layer to become sunken and discolored, so that they can be clearly screened out during secondary identification.

[0056] B2. Use an industrial camera 101 to acquire the second image of the thawed silkworm in step B2. Use a machine vision recognition algorithm and the size of the thawed silkworm in the second image to identify whether the thawed silkworm is a dead silkworm.

[0057] In this embodiment, the step of identifying the external dimensions of the thawed silkworm in the second image includes:

[0058] C1: Binarize the second image, extract the binarized second image, and obtain the binarized image, as shown below. Figure 8 As shown.

[0059] C2: Calculate the aspect ratio of the binarized image and compare it with a preset threshold range. The thawed silkworms corresponding to the binarized images located outside the threshold range are identified as dead silkworms.

[0060] Since dead silkworms typically ooze pus from their head, top, and sides after thawing, this invention binarizes the second image. In the second image, black pixels represent the pus-oozing portion of the thawed silkworm, and white pixels represent the intact portion. By extracting the pixels from the second image and obtaining the aspect ratio of the thawed silkworm, it is possible to accurately determine whether the thawed silkworm is a dead silkworm. In this embodiment, based on actual measurements, the normal aspect ratio of a cocoon is 1:1.7 to 1:2.1. One identification result of this embodiment is as follows: Figure 9 As shown, the thawed silkworm is of the head-end seepage type. At this time, its length-to-width ratio is calculated to be 1.6, which is significantly smaller than the set values ​​(1.7, 2.1).

[0061] B3. Remove the dead silkworms from step B2 to obtain intact silkworms.

[0062] In this embodiment, the machine vision recognition algorithm is implemented using a machine vision engineering machine.

[0063] In the first identification stage, the appearance characteristics of frozen silkworms are identified to screen out obviously defective cocoons, thus obtaining the initial selected silkworms. In the second identification stage, the initial selected silkworms are thawed. The pus from dead cocoons seeps out, making the size of the dead cocoons different from that of intact silkworms. Thus, the machine vision recognition algorithm can determine whether the thawed silkworms are dead cocoons. At the same time, the invention of using cold air ventilation for thawing accelerates the manifestation of the characteristics of dead cocoons, shortens the time of the entire process, and also ensures the freshness of the silkworm pupae, which is conducive to the screening out of dead cocoons. Without affecting the silkworm pupae, the quality of raw silk reeling is improved, and the overall production efficiency and benefits of the enterprise are increased.

[0064] This embodiment also includes an identification device used in the method for identifying frozen dead silkworms. The identification device includes a conveyor belt 1, a first storage box 2, a second storage box 3, and a third storage box 6.

[0065] The conveyor belt 1 has a U-shaped structure with openings at both ends, and the horizontal heights at the two ends of the conveyor belt 1 are different. In this embodiment, the conveyor belt 1 includes a semi-circular first conveyor section 121, a second conveyor section 122, and a third conveyor section 123. The first conveyor section 121 has a semi-circular structure, and its two ends are respectively connected to the second conveyor section 122 and the third conveyor section 123. The second conveyor section 122 is placed horizontally, and the third conveyor section 123 is placed at an angle so that the horizontal heights at the two ends of the conveyor belt 1 are different.

[0066] The first storage box 2 is located above the conveyor belt 1 on one side. The first storage box 2 is provided with a recessed first cavity 201, and the bottom of the first storage box 2 is provided with a first discharge hole 202 communicating with the first cavity 201. The first discharge hole 202 is provided with a discharge mechanism so that a single frozen silkworm can be discharged to the conveyor belt 1 through the discharge mechanism.

[0067] The second storage box 3 is located at the opening of the conveyor belt 1, with the upper end of the conveyor belt 1 above the second storage box 3 and the lower end of the conveyor belt 1 below the second storage box 3. The top of the second storage box 3 is recessed with a second cavity 301, and the bottom of the second storage box 3 is inclined from the upper end to the lower end of the conveyor belt 1. The side of the second storage box 3 near the lower end of the conveyor belt 1 is provided with a first discharge hole 302 communicating with the second cavity 301. The first discharge hole 302 is provided with a discharge mechanism so that a single thawed silkworm can be discharged to the conveyor belt 1 through the discharge mechanism.

[0068] The second storage box 3 is equipped with an exhaust pipe 31. One end of the exhaust pipe 31 is connected to the second storage box 3 through several pipes 311. The exhaust fan 102 is located at the end of the exhaust pipe 31 away from the second storage box 3. In this embodiment, the side wall of the second storage box 3 is provided with a through vent 303. Under the action of the vent 303, the air circulation speed inside the second storage box 3 can be increased, thereby improving the defrosting efficiency.

[0069] The third storage box 6 is located on the side of the second storage box 3 away from the conveyor belt 1. A guide groove 61 is slidably provided in the third storage box 6, located between the third storage box 6 and the second storage box 3. The guide groove 61 is driven by a take-up motor 62, causing it to slide towards the third storage box 6. One end of the guide groove 61 is located at the bottom of the conveyor belt 1, and the other end is located above the third storage box 6, or it can move away from the second storage box 3. In this embodiment, the guide groove 61 is slidably connected to the third storage box 6 on both sides by slide rails 611. A rack 612 is provided at the bottom of the guide groove 61. The take-up motor 62 is fixedly mounted on the side wall of the third box 6, and the drive shaft of the take-up motor 62 is provided with a drive gear 621 that meshes with the rack 612, so that the take-up motor 62 drives the guide groove 61 to move.

[0070] During the first identification, the receiving motor 62 is turned forward, and the guide trough 61 leaves the second storage box 3, so that the initially selected silkworms can fall into the third storage box 6. After the first identification is completed, the receiving motor 62 is reversed, and the guide trough 61 slides to the bottom of the conveyor belt 1 so that the intact silkworms can be collected in the third storage box 6 in the future.

[0071] In this embodiment, the first storage box 2 and the second storage box 3 can be mounted on the conveyor belt 1 using a fixed frame, and the second storage box 3 can be placed on the ground.

[0072] An industrial camera 101 is positioned above the conveyor belt 1, located on one side of the first storage box 2 along the forward direction of the conveyor belt 1. The industrial camera 101 is fixed above the conveyor belt 1 by a support frame 11, and supplementary lights 103 are provided on both sides of the support frame 11. The supplementary lights 103 are incandescent lamps.

[0073] By placing frozen silkworms in the first storage box 2 and conveying them via the conveyor belt 1, the frozen silkworms pass through the industrial camera 101, which acquires a first image to complete the first identification. The initially selected silkworms are then transported via the conveyor belt 1 into the second storage box 3. Under the action of the exhaust fan 102, the initially selected silkworms in the second storage box 3 are evacuated with cold air. The thawed silkworms are then conveyed via the conveyor belt 1 to the industrial camera 101 for a second image capture, thus completing the second identification. The identification device of this application automatically identifies dead silkworms, effectively improving identification efficiency.

[0074] like Figure 6As shown, the discharge mechanism includes a housing 4, a discharge ball 41, and a discharge motor 42. The housing 4 has a through discharge hole 401, which is connected to the corresponding first discharge hole 202 and first discharge hole 302. The discharge ball 41 is located inside the discharge hole 401 and is rotatably connected to the housing 4. The discharge ball 41 is recessed with a temporary storage groove 411 so that a single frozen silkworm or a single thawed silkworm can enter the temporary storage groove 411 through the discharge hole 401. The discharge motor 42 is used to drive the discharge ball 41 so that the temporary storage groove 411 can face the corresponding first discharge hole 202 and first discharge hole 302, or face the conveyor belt 1.

[0075] In this embodiment, the discharge ball 41 is fixedly provided with rotating shafts at both ends. The rotating shafts are rotatably connected to the outer shell 4. One of the rotating shafts passes through the outer shell 4 and is fixedly provided with a first gear 431. The discharge motor 42 is fixed outside the outer shell 4 and meshes with the first gear 431 through the second gear 432 to realize the driving of the discharge ball 41.

[0076] Several temporary storage slots 411 can be provided. In this embodiment, there is one temporary storage slot 411. After the discharge ball 41 rotates 90° for the first time, a single frozen silkworm or a single thawed silkworm in the temporary storage slot 411 can fall into the conveyor belt 1. After the discharge ball 41 rotates 90° again, the single frozen silkworm or a single thawed silkworm can enter the temporary storage slot 411 through the discharge through hole 401, completing one feeding cycle. The discharge motor 42 is a servo motor. It stops for 1 second each time the discharge ball 41 rotates 90°, so that the single frozen silkworm or the single thawed silkworm can fall into the temporary storage slot 411 or fall out of the temporary storage slot 411.

[0077] like Figure 2 As shown, a screening mechanism is provided between the industrial camera 101 and the second storage box 3. The screening mechanism includes a pneumatic push rod 13 and a processing tank 131. The pneumatic push rod 13 and the processing tank 131 are arranged opposite each other on both sides of the conveyor belt 1. The pneumatic push rod 13 is located on one side of the conveyor belt 1, and the drive rod of the pneumatic push rod 13 faces the conveyor belt 1 and is provided with a push plate 132, so that when the pneumatic push rod 13 is extended, the push plate 132 can push the dead silkworm into the processing tank 131.

[0078] The conveyor belts 1 on both sides of the industrial camera 101 are respectively equipped with a first infrared sensor 51 and a second infrared sensor 52. The first infrared sensor 51 is located on the side of the industrial camera 101 near the first storage box 2, so as to identify silkworms entering the industrial camera 101.

[0079] The second infrared sensor 52 is located between the industrial camera 101 and the screening mechanism. The second infrared sensor 52 is used to identify silkworms passing through the screening mechanism.

[0080] In this embodiment, the identification of frozen dead silkworms can be carried out in batches. Each batch identifies a number of frozen silkworms or a number of thawed silkworms. Within a certain batch, each batch is simultaneously photographed by the industrial camera 101. The first infrared sensor 51 is used to count the frozen or thawed silkworms. When the first infrared sensor 51 counts to a set value, the machine vision engineering machine controls the conveyor belt 101 and the discharge motor 42 to pause, and simultaneously controls the industrial camera 101 to take pictures. The machine vision engineering machine acquires the information from the industrial camera 101 for identification.

[0081] After a preset pause time, the machine vision engineering machine completes the identification and labels the dead silkworms in the image according to the sorting. Then, it controls the start of the conveyor belt 101 and the discharge motor 42 to make the frozen or thawed silkworms pass through the second infrared sensor 52. The second infrared sensor 52 counts, and when the second infrared sensor 52 counts to the label, it proves that the frozen or thawed silkworm is a dead silkworm. According to the speed of the conveyor belt 101, after the pushing delay, the machine vision engineering machine controls the pneumatic push rod 13 to extend and retract, thereby pushing the dead silkworms into the processing tank 131.

[0082] The operation process of the identification device using this embodiment is as follows:

[0083] First identification: The receiving motor 62 rotates to the forward position, causing the guide trough 61 to leave the second storage box 3.

[0084] Frozen silkworms are placed in the first storage box 2. The screening mechanism discharges the frozen silkworms in sequence. The frozen silkworms pass through the first infrared sensor 51. When the first infrared sensor 51 counts to the set value, the machine vision engineering machine controls the conveyor belt 101 and the discharge motor 42 of the first storage box 2 to stop. At the same time, it controls the industrial camera 101 to take pictures. The machine vision engineering machine acquires the information from the industrial camera 101 for identification.

[0085] After a preset pause time, the machine vision engineering machine completes the identification and labels the dead silkworms in the first image according to the sorting. Then, it controls the conveyor belt 101 and the discharge motor 42 of the first storage box 2 to start, so that the frozen silkworms pass through the second infrared sensor 52. The second infrared sensor 52 counts, and when the second infrared sensor 52 counts to the label, it proves that the frozen silkworm is a dead silkworm. According to the speed of the conveyor belt 101, after the pushing delay, the machine vision engineering machine controls the pneumatic push rod 13 to extend and retract, thereby pushing the dead silkworms into the processing bucket 131.

[0086] Secondary identification: The receiving motor 62 reverses, one end of the guide groove 61 slides to the bottom of the conveyor belt 1, and the other end is located above the third storage box 6.

[0087] After all the initially selected silkworms have entered the second storage box 3, the exhaust fan 102 ventilates the second storage box 3 with cold air. After thawing, the screening mechanism discharges the thawed silkworms one by one, and the conveyor belt 101 transports the thawed silkworms.

[0088] When the thawed silkworm passes through the first infrared sensor 51, and the first infrared sensor 51 counts to the set value, the machine vision engineering machine controls the conveyor belt 101 and the discharge motor 42 of the second storage box 2 to stop, and at the same time controls the industrial camera 101 to take pictures. The machine vision engineering machine acquires the industrial camera 101 for identification.

[0089] After a preset pause time, the machine vision engineering machine completes the identification and labels the dead silkworms in the second image according to the sorting. Then, it controls the start of the conveyor belt 101 and the discharge motor 42 of the second storage box 2, causing the frozen silkworms to pass through the second infrared sensor 52. The second infrared sensor 52 counts, and when the second infrared sensor 52 counts to the labeled number, it proves that the frozen silkworm is a dead silkworm. According to the speed of the conveyor belt 101, after a pushing delay, the machine vision engineering machine controls the pneumatic push rod 13 to extend and retract, thereby pushing the dead silkworms into the processing bucket 131. Intact silkworms enter the third storage box 6 through the guide groove 61.

Claims

1. A machine vision-based method for identifying frozen, dead silkworms, characterized in that, Includes primary recognition and secondary recognition. The initial identification includes the following steps: A1. An industrial camera (101) is used to acquire a first image of the frozen silkworm. A machine vision recognition algorithm is used to identify the frozen silkworm based on its appearance features in the first image, so as to determine whether the frozen silkworm is a dead cage silkworm. A2. After removing the dead silkworms from the frozen silkworms described in step A1, the initial selected silkworms are obtained. The secondary identification includes the following steps: B1. Use a fan (102) to ventilate the silkworms selected in step A2 with cold air to obtain thawed silkworms; B2. Use an industrial camera (101) to acquire a second image of the thawed silkworm described in step B1, and use a machine vision recognition algorithm to identify the thawed silkworm based on its external dimensions in the second image to determine whether the thawed silkworm is a dead silkworm. The steps for identifying the external dimensions of the thawed silkworm in the second image include: C1: Binarize the second image, extract the binarized second image, and obtain the binarized image; C2: Calculate the aspect ratio of the binarized image and compare the aspect ratio of the binarized image with a preset threshold range. Determine the thawed silkworms corresponding to the binarized images that are outside the threshold range as dead silkworms. B3. After removing the dead silkworms described in step B2, intact silkworms are obtained. It also includes an identification device used in the method for identifying frozen dead silkworms, the identification device comprising a conveyor belt (1), a first storage box (2) and a second storage box (3). The conveyor belt (1) is a U-shaped structure with openings at both ends, and the horizontal heights at both ends of the conveyor belt (1) are different. The first storage box (2) is located above the conveyor belt (1) on one side. The first storage box (2) has a recessed first cavity (201), and the bottom of the first storage box (2) is provided with a first discharge hole (202) communicating with the first cavity (201). The first discharge hole (202) is provided with a discharge mechanism so that a single frozen silkworm can be discharged to the conveyor belt (1) through the discharge mechanism. The second storage box (3) is located at the opening of the conveyor belt (1), and the high end of the conveyor belt (1) is located above the second storage box (3), and the low end of the conveyor belt (1) is located below the second storage box (3). The top of the second storage box (3) is recessed with a second cavity (301). The bottom of the second storage box (3) is inclined from the high end of the conveyor belt (1) to the low end of the conveyor belt (1). The side of the second storage box (3) near the low end of the conveyor belt (1) is provided with a second discharge hole (302) communicating with the second cavity (301). The second discharge hole (302) is provided with a discharge mechanism so that a single thawed silkworm is discharged to the conveyor belt (1) through the discharge mechanism. The second storage box (3) is provided with an exhaust pipe (31). One end of the exhaust pipe (31) is connected to the second storage box (3) through several pipes (311). The exhaust fan (102) is located at the end of the exhaust pipe (31) away from the second storage box (3). The industrial camera (101) is positioned above the conveyor belt (1), and the industrial camera (101) is located on one side of the first storage box (2) along the forward direction of the conveyor belt (1); A screening mechanism is provided between the industrial camera (101) and the second storage box (3). The screening mechanism includes a pneumatic push rod (13) and a processing bucket (131). The pneumatic push rod (13) and the processing bucket (131) are arranged opposite to each other on both sides of the conveyor belt (1). The pneumatic push rod (13) is located on one side of the conveyor belt (1), and the drive rod of the pneumatic push rod (13) faces the conveyor belt (1) and is provided with a push plate (132) so that when the pneumatic push rod (13) is extended, the push plate (132) can push the dead silkworm into the processing bucket (131). The conveyor belts (1) on both sides of the industrial camera (101) are respectively equipped with a first infrared sensor (51) and a second infrared sensor (52). The first infrared sensor (51) is located on the side of the industrial camera (101) close to the first storage box (2) to identify silkworms entering the industrial camera (101). The second infrared sensor (52) is located between the industrial camera (101) and the screening mechanism, and the second infrared sensor (52) is used to identify silkworms passing through the screening mechanism.

2. The method for identifying frozen dead silkworms based on machine vision according to claim 1, characterized in that: In step A1, the appearance features include dead cocoons, cotton cocoons, mouth cocoons, moldy cocoons, and cocoons with multiple defects.

3. The method for identifying frozen dead silkworms based on machine vision according to claim 1, characterized in that: The air velocity of the exhaust fan (102) is 4-5 m / s, and the air temperature of the exhaust fan (102) is 4-10℃.

4. The method for identifying frozen dead silkworms based on machine vision according to claim 1, characterized in that: The discharge mechanism includes a housing (4), a discharge ball (41), and a discharge motor (42). The housing (4) has a through discharge hole (401), which is connected to the corresponding first discharge hole (202) and second discharge hole (302). The discharge ball (41) is located in the discharge hole (401) and is rotatably connected to the housing (4). The discharge ball (41) has a recessed storage groove (411) so that a single frozen silkworm or a single thawed silkworm can enter the storage groove (411) through the discharge hole (401). The discharge motor (42) is used to drive the discharge ball (41) so that the storage groove (411) can face the corresponding first discharge hole (202) and second discharge hole (302), or face the conveyor belt (1).

5. The method for identifying frozen dead silkworms based on machine vision according to claim 1, characterized in that: The industrial camera (101) is fixed above the conveyor belt (1) by a support frame (11), and supplementary lights (103) are provided on both sides of the support frame (11).

6. The method for identifying frozen dead silkworms based on machine vision according to claim 1, characterized in that: It also includes a third storage box (6), which is located on the side of the second storage box (3) away from the conveyor belt (1). The third storage box (6) is slidably provided with a guide groove (61), which is located between the third storage box (6) and the second storage box (3). The guide groove (61) is driven by a receiving motor (62) so that the guide groove (61) slides toward the third storage box (6). One end of the guide groove (61) is located at the bottom of the conveyor belt (1), and the other end is located above the third storage box (6).

Images