A bird's nest automatic selection device and method based on visual recognition technology

Abstract

The present invention provides a bird's nest automatic fine picking device based on visual recognition technology, including a support frame, including an upper fixed plate, a middle fixed plate and a lower fixed plate, a light-through hole is provided on the upper fixed plate, and a port to be inspected is provided below the lower fixed plate; an illumination light source is installed on the light-through hole and is provided with a hollow area, an image collector is installed just above the hollow area; a picking and inspection mold is placed on the port to be inspected and just below the hollow area; an operating module is installed on the front end of the lower fixed plate; a mechanical arm is installed on the lower fixed plate; a picking cutter head is arranged at the upper end of the mechanical arm; a vacuum pump; a waste box, one end of which is connected to the vacuum pump and the other end is connected to the picking cutter head; and an electrical control box is connected to the image collector, the operating module, the mechanical arm and the vacuum pump respectively. The present invention also provides a bird's nest automatic fine picking method based on visual recognition technology, which replaces manual impurity picking of raw materials and solves the problems of low efficiency of manual picking.

Description

Technical Field

[0001] The present invention relates to the technical field of food processing, and in particular to an automatic bird's nest selection device and method based on visual recognition technology. Background Art

[0002] Bird's nest is a nest made of saliva or down feathers mixed with saliva of the swiftlet family and its relatives. Bird's nest can be divided into house bird's nest and cave bird's nest according to its source. House bird's nest comes from bird's nests built by humans, while cave bird's nest comes from caves on islands or seashores. The main production areas of bird's nest include the South China Sea Islands and Southeast Asian countries such as Malaysia, Indonesia, Thailand and Vietnam. As early as the Ming Dynasty, bird's nest had become a tribute to the imperial court by vassal states.

[0003] Bird's nest contains a rich variety of amino acids, high protein and sialic acid content, and has the effects of invigorating qi and resolving phlegm, nourishing the kidneys and lungs, etc. Modern medical research shows that bird's nest also has the effects of anti-virus, enhancing immunity, and promoting cell division, and is often used for pregnant women and middle-aged and elderly weak patients.

[0004] The products currently on the market include dry bird's nest and instant bird's nest. Instant bird's nest refers to bird's nest produced by canning technology, which is processed, blended, filled, sealed, steamed, sterilized, cooled, inspected, and packaged to meet commercial sterility requirements. It is a bird's nest food that can be stored for a long time at room temperature, easy to carry, and ready to eat after opening the bottle. As the nutritional value of bird's nest is gradually recognized, there are more and more bird's nest products on the market, from the traditional single pure tonic to the variety of packaging specifications and styles, and then to the upgrade of product forms. With the step-by-step evolution of new bird's nest products, the bird's nest industry has gradually grown and developed.

[0005] Picking impurities from bird's nests is an essential process in the production of instant bird's nests. The cleanliness of the bird's nest raw materials directly affects the efficiency of picking bird's nests during the production process and the quality of the finished product of instant bird's nests. The main sources of impurities in bird's nest raw materials are: first, brought in by the bird's nest raw materials themselves, such as bird hair, black spots, white blocks, egg membranes, egg shells, etc.; second, caused by external pollution, such as thin threads, plastic threads, wood chips, paper scraps, and sand.

[0006] At present, impurities in bird's nest are mainly picked out manually. The process of manual picking is as follows: Figure 1 As shown: white tray picking → tweezers cleaning → black tray picking → tweezers cleaning → quality inspection. This method has the following problems:

[0007] 1) Detecting impurities requires a lot of labor, resulting in high production costs; long-term visual picking sharply increases personnel fatigue, and there is a risk of missed defects, which affects the stability of product quality;

[0008] 2) Manual picking is affected by subjective factors and there are inconsistent judgment standards, resulting in uneven product cleanliness and quality; manual picking of impurities greatly limits the further increase in the production capacity of instant bird's nests. Therefore, how to improve the picking efficiency has become a problem that many production companies need to solve urgently. Summary of the invention

[0009] In order to solve the above problems, the present invention provides an automatic bird's nest picking device based on visual recognition technology, which replaces manual impurity picking of raw materials and solves the problem of low efficiency of manual picking.

[0010] The present invention is implemented as follows: an automatic bird's nest selection device based on visual recognition technology, comprising:

[0011] A support frame, the support frame comprises an upper fixing plate, a middle fixing plate and a lower fixing plate, the two ends of the middle fixing plate are respectively connected to the upper fixing plate and the lower fixing plate; a light-through hole is provided on the upper fixing plate, and an inspection port is provided below the lower fixing plate;

[0012] An illumination light source, the illumination light source is installed on the light-through hole, and a hollow area is provided in the middle of the illumination light source;

[0013] An image collector, which is installed just above the hollow area;

[0014] An inspection mold is placed on the inspection port and installed just below the hollow area;

[0015] An operating module, wherein the operating module is installed at the front end of the lower fixing plate;

[0016] A mechanical arm, the lower end of which is mounted on a lower fixing plate;

[0017] A picking cutter head, which is arranged at the upper end of the mechanical arm; when working, the picking cutter head is driven to the inspection mold by controlling the mechanical arm to inspect the impurities;

[0018] Vacuum pump;

[0019] A waste box, one end of which is connected to a vacuum pump through a first pipe, and the other end of which is connected to a picking cutter head through a second pipe; the impurities cut off by the picking cutter head are recovered from the second pipe to the waste box through the vacuum pump;

[0020] and an electrical control box, wherein the electrical control box is respectively connected to the image collector, the illumination light source, the operation module, the mechanical arm and the vacuum pump.

[0021] Furthermore, it also includes: a pressure sensor, which is arranged on the picking tool head and connected to the electrical control box.

[0022] Furthermore, it also includes: a water supply box, the lower fixed plate is provided with a water suction port, the water supply box is connected to the water suction port through a water inlet pipe; when cleaning the second pipe, the second pipe is moved to the water suction port, and water is introduced into the waste box from the second pipe through a vacuum pump;

[0023] The water supply tank includes a box body and a cover matching the box body, the cover is provided with a sealing hole, the inner wall of the box body is provided with a valve core, the valve core is connected to a float matching the sealing hole through a valve stem, and when the water level reaches a set position, the float drives the valve core on the valve stem to seal the sealing hole of the cover.

[0024] Furthermore, it also includes a cutter head cleaning needle and an air pressure sensor, wherein the cleaning needle is arranged on the lower fixing plate, and the air pressure sensor is installed in the first pipeline and connected to the electrical controller.

[0025] Furthermore, the electrical control box includes an input power supply, a PLC controller and an IO control module, an air switch box and a reset indicator light, and the PLC controller is respectively connected to the input power supply, the IO control module, the air switch box, the reset indicator light, the image collector, the lighting source, the operation module, the robotic arm, the vacuum pump, the pressure sensor and the air pressure sensor;

[0026] Both ends of the lower fixed plate are provided with enclosures, a plurality of drainage holes are provided around the inspection port, the image collector is an industrial camera, and the operating module is a key switch.

[0027] Furthermore, the inspection mold is a dry picking mold or a wet picking mold, the dry picking mold includes a placement plate, a connecting column and a support seat, the placement plate is connected to the support seat through the connecting column, the placement plate is a polypropylene plate, the support seat is a spherical solid rubber seat, and the support seat is used to place the dry bird's nest and bird's nest to be inspected; the wet picking mold includes a placement plate, the placement plate is a polypropylene plate, which is used to place the wet bird's nest material to be inspected.

[0028] To solve the above problems, the present invention also provides an automatic bird's nest selection method based on visual recognition technology, which replaces manual impurity selection of raw materials and solves the problem of low efficiency of manual selection.

[0029] The present invention is implemented as follows: a method for automatically selecting bird's nests based on visual recognition technology, the method for automatically selecting bird's nests needs to provide the above-mentioned automatic selection device for bird's nests based on visual recognition technology, the method comprising:

[0030] Step 1, select a picking and inspection mold according to the type of bird's nest material to be inspected, place the selected picking and inspection mold at the inspection port, and then place the bird's nest material on the picking and inspection mold; and install a picking cutter head matching the selected picking and inspection mold at the front end of the mechanical arm;

[0031] Step 2: Divide the bird's nest material to be inspected into a plurality of inspection areas of equal area according to the lens field of view of the image collector;

[0032] Step 3: Turn on the illumination light source for illumination, select a detection area, and the lens of the image collector collects a color image of the detection area through the hollow area, and transmits the color image to the electrical control box;

[0033] Step 4: After receiving the color image, the electrical control box converts the color image into a grayscale image;

[0034] Step 5: After filtering and enhancing the obtained grayscale image, the corresponding grayscale value is obtained and sent to the electrical control box;

[0035] Step 6, the electrical control box determines the impurity situation by comparing the gray value with the set value. If the gray value does not exceed the set value, it is determined that the pixel points in the detection area are normal, and there is no impurity in the detection area, and the process returns to step 3 to select the next detection area for processing; if the gray value exceeds the set value, it is determined that the pixel points in the detection area are impurities, and the process goes to step 7;

[0036] Step 7, the electrical control box obtains the coordinate position of the impurity, and controls the robot arm to move the picking cutter head to the coordinate position of the impurity through the operation module;

[0037] Step 8: After the picking cutter head descends, it cuts off the impurities and sucks the impurities out to the waste box through the negative pressure of the vacuum pump; select the next detection area and repeat steps 3-8 until the impurities in all detection areas are removed.

[0038] Furthermore, the method for automatically selecting bird's nests also needs to provide a pressure sensor, which is arranged on the picking tool head and connected to the electrical control box; the method comprises:

[0039] Step 9, the pressure sensor sends the pressure value to the electrical control box in real time, and the electrical control box determines whether the inspection mold is damaged based on the pressure value. If the pressure value does not exceed the set value, it means that no damage is caused and no processing is performed; if the pressure value exceeds the set value, it means that damage is caused, and the electrical control box controls the robotic arm through the operating module to move the picking head upward.

[0040] Furthermore, the automatic bird's nest selection method further needs to provide a water supply tank, the lower fixed plate is provided with a water suction port, and the water supply tank is connected to the water suction port through a water inlet pipe; the method comprises:

[0041] Step 10: When the second pipe is blocked, move the second pipe to the water suction port, and introduce water from the second pipe into the waste tank through a vacuum pump.

[0042] Furthermore, the automatic bird's nest selection method further needs to provide a cutter head cleaning needle and an air pressure sensor, wherein the cleaning needle is arranged on the lower fixing plate, and the air pressure sensor is installed in the first pipe and connected to the electrical controller; the method comprises:

[0043] Step 11, the air pressure sensor sends the air pressure value to the electrical control box in real time, and the electrical control box determines whether the picking cutter head is blocked according to the air pressure value. If the air pressure value does not exceed the set value, it means that the picking cutter head is not blocked and no processing is performed; if the air pressure value exceeds the set value, it means that the picking cutter head is blocked, and the electrical control box controls the robot arm through the operation module to move the picking cutter head upward to the cutter head cleaning needle for cleaning; there is no order among step 9, step 10 and step 11.

[0044] The beneficial effects of the present invention are:

[0045] 1. This device can improve the processing efficiency of bird's nest raw materials, partially or completely replace manual labor, and save labor costs;

[0046] 2. Through this device, you can choose dry picking mode or wet picking mode to meet the different picking requirements of dry bird's nest production line and instant bird's nest production line;

[0047] 3. Through the pressure sensor monitoring of the robotic arm, the movement trajectory of the robotic arm can be effectively controlled to protect the strips of bird's nest raw materials from damage to the greatest extent, which can effectively improve the quality of the finished product;

[0048] 4. Through the coordinated action of the visual system and the robot, impurities in the material can be accurately removed. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present invention and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0050] Figure 1 It is a schematic diagram of the manual picking process in the prior art.

[0051] Figure 2 A schematic structural diagram of an automatic bird's nest selection device based on visual recognition technology provided in an embodiment of the present invention.

[0052] Figure 3 Schematic diagram of the optical path of the image collector, the illumination light source and the inspection mold in the embodiment of the present invention.

[0053] Figure 4 The figure is a schematic diagram of the connection between the water inlet pipe, the first pipe and the second pipe according to an embodiment of the present invention.

[0054] Figure 5 Schematic diagram of the structure of a water supply tank according to an embodiment of the present invention.

[0055] Figure 6 A schematic structural diagram of a dry picking mold according to an embodiment of the present invention.

[0056] Figure 7 It is a schematic structural diagram of a wet picking mold according to an embodiment of the present invention.

[0057] Figure 8 It is a schematic structural diagram of a dry picking cutter head according to an embodiment of the present invention.

[0058] Fig. 9 Schematic diagram of the structure of a wet pick cutter head according to an embodiment of the present invention.

[0059] Fig.10 It is a schematic diagram of the circuit connection of the electrical control box of the present invention.

[0060] Fig.11 A flow chart of a method for automatically selecting bird's nests based on visual recognition technology provided in an embodiment of the present invention.

[0061] The accompanying drawings in the figure are marked as:

[0062] 1-support frame, 1a-upper fixed plate, 1b-middle fixed plate, 1c-lower fixed plate, 1d-light hole, 1e-inspection port, 2-image collector, 21-lens, 3-picking mold, 3a-dry picking mold, 3b-wet picking mold, 31-placement plate, 32-connecting column, 33-support seat, 34-placement plate, 4-operation module, 5-mechanical arm, 6-picking tool head, 6a-dry picking tool head, 6b-wet picking tool head, 61-groove, 62-spring, 63-ring, 7-vacuum pump, 8-waste box, 81- First pipeline, 82-second pipeline, 9-electrical control box, 91-input power supply, 92-PLC controller, 93-IO control module, 94-air switch box, 95-reset indicator light, 10-pressure sensor, 11-water supply tank, 111-water suction port, 112-water inlet pipe, 113-box body, 114-lid; 115-sealing hole, 116-valve core, 117-valve stem, 118-float, 12-knife head cleaning needle, 13-air pressure sensor, 14-lighting light source, 141-hollow area. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Therefore, the following detailed description of the embodiments of the present invention provided in the drawings is not intended to limit the scope of the invention claimed for protection, but merely represents selected embodiments of the present invention.

[0064] like Figure 2-10 As shown, in this embodiment, an automatic bird's nest selection device based on visual recognition technology includes:

[0065] The support frame 1 includes an upper fixing plate 1a, a middle fixing plate 1b and a lower fixing plate 1c. The two ends of the middle fixing plate 1b are respectively connected to the upper fixing plate 1a and the lower fixing plate 1c to form a U-shaped structure for easy installation; the upper fixing plate 1a is provided with a light hole 1d, and the lower fixing plate 1c is provided with an inspection port 1e; the support frame 1 is used to support and place other accessories; the inspection port 1e: a square area is marked out on the lower fixing plate 1c for placing the picking mold 3;

[0066] An illumination light source 14, which is installed on the light hole 1d. The illumination light source 14 is an LED ring light source, which can provide a bright lighting environment for the area to be inspected; and a hollow area 141 is provided in the middle of the illumination light source 14;

[0067] Image collector 2, the lens 21 of the image collector 2 is installed just above the hollow area 141, the image collector 2 is an industrial camera, used to photograph the bird's nest material on the picking mold 3 and collect information; Industrial camera: the image of the area to be inspected is photographed by the lens 21 of the industrial camera and transmitted to the visual recognition system;

[0068] The inspection mold 3 is placed on the inspection port 1e, and the inspection mold 3 is installed directly below the hollow area 141, that is, the lens 21 of the image collector 2, the hollow area 141 and the inspection mold 3 are located on a vertical line; the inspection mold 3 is used to place the bird's nest material; Figure 3 As shown, the illumination light source 14 emits light to the bird's nest material to be tested on the inspection mold 3, and the light is then reflected from the bird's nest material to be tested back to the lens 21, and is received and photographed by the lens 21 of the industrial camera;

[0069] In this embodiment, the inspection mold 3 is a dry inspection mold 3a or a wet inspection mold 3b. The dry inspection mold 3a includes a placement plate 31, a connecting column 32 and a support seat 33. The placement plate 31 is connected to the support seat 33 through the connecting column 32. The placement plate 31 is a polypropylene plate, and the support seat 33 is a spherical solid rubber seat. The support seat 33 is used to place the dry bird's nest to be inspected; the wet inspection mold 3b includes a placement plate 34, and the placement plate 34 is a polypropylene plate for placing the wet bird's nest material to be inspected;

[0070] The operation module 4 is installed at the front end of the lower fixing plate 1c; the operation module 4 is a key switch, and the key design is more convenient for operation; the operation buttons include physical buttons such as the opening and closing of the vacuum pump 7, the cleaning of the picking cutter head 6, the cleaning of the second pipeline 82, the reset, the start, the pause, and the emergency stop; all functions can be realized on the touch screen, and the physical key switch here can increase the convenience of operation;

[0071] A mechanical arm 5, the lower end of which is mounted on the lower fixing plate 1c; the mechanical arm 5 is an electrically driven mechanical arm, which is driven by an electrical control box 1 to imitate the movement functions of human hands and arms, and a corresponding picking cutter head 6 is mounted at the end thereof to match different picking modes;

[0072] A picking cutter head 6, which is arranged at the upper end of the mechanical arm 5; when working, the picking cutter head 6 is driven to the picking mold 3 by controlling the mechanical arm 5 to pick and inspect impurities on the bird's nest material; the picking cutter head 6 is a dry picking cutter head 6a or a wet picking cutter head 6b, and the dry picking cutter head 6a adopts a rotating cutter head with a diameter of 0.5-2.5mm and a spiral groove 61, which removes impurities from the dry bird's nest by high-speed rotation and vacuum; the wet picking cutter head 6b is a cylindrical cutter head with a diameter of 1-3mm, which is made of a concentric reducer made of metal, and the outer wall of the wet picking cutter head is covered with a spring 62, and the bottom of the wet picking cutter head 6b has a ring 63 for pressing the wet material; according to different working modes, two different cutter heads are used;

[0073] A vacuum pump 7 is used to establish a vacuum system to form a negative pressure to remove impurities; the impurities are sucked away by vacuum;

[0074] A waste box 8, one end of which is connected to the vacuum pump 7 through a first pipe 81, and the other end is connected to the picking cutter head 6 through a second pipe 82; the bird's nest material impurities cut off by the picking cutter head 6 are recovered from the second pipe 82 to the waste box 8 through the vacuum pump 7; the waste box 8 is used to collect the picked impurities;

[0075] And an electrical control box 9, the electrical control box 9 is respectively connected to the image collector 2, the lighting source 14, the operating module 4, the mechanical arm 5 and the vacuum pump 7, and is used to load the control circuit in the electrical control box 9 (including an all-in-one machine with a touch screen), which can realize real-time image display, as well as the setting of parameters such as industrial camera exposure parameters, motor motion parameters, and identification and rejection accuracy. The equipment operation is divided into two modes: dry material picking and wet material picking. The electrical control box 9 includes an input power supply 91, a PLC controller 92 and an IO control module 93, an air switch box 94 and a reset indicator light 95. The PLC controller 92 and the IO control module 93 are installed in the electrical control box 9, the air switch box 94 is installed outside the electrical control box 9, and the reset indicator light 95 is arranged on the side of the electrical control box 9; the PLC controller 92 is used to receive and process the image information uploaded by the image collector 2, and convert the color image into a black and white image through the conversion of the image algorithm, and mark the impurities according to the set black point value and gray value;

[0076] The PLC controller 92 is respectively connected to the input power supply 91 (220V), the IO control module 93, the air switch box 94, the reset indicator light 95, the image collector 2, the lighting source 14, the operation module 4, the robot arm 5 and the vacuum pump 7. When the equipment is reset, the reset indicator light 95 lights up.

[0077] In this solution, it also includes: a pressure sensor 10, which is set on the picking cutter head 6 and connected to the PLC controller 92 in the electrical control box 9; the pressure sensor 10 is used to sense the pressure between the picking cutter head 6 and the contact surface to avoid the picking die 3 being damaged by the picking cutter head 6, resulting in a decrease in the recognition accuracy of defects. When the pressure exceeds the set value, the PLC controller 92 sends a signal to the robot arm 5 to control the picking cutter head 6 to move upward. Through this setting, the original strip can be protected to the greatest extent during the picking process.

[0078] In the present scheme, it also includes: a water supply box 11, a water suction port 111 is provided on the lower fixed plate 1c, and the water supply box 11 is connected to the water suction port 1f through the water inlet pipe 112; when cleaning the second pipe, the second pipe 82 is moved to the water suction port 111, and the water is introduced into the waste bin 8 from the second pipe 82 through the vacuum pump 7; the water suction port 111: the water is introduced from the water supply box 11 to this place by using the water inlet pipe 112. After a certain number of pickings, the second pipe 82 is prone to blockage. The control robot 5 moves the port of the second pipe 82 fixed on the picking cutter head 6 to this place, and the water is sucked into the second pipe 82 by the vacuum of the vacuum pump 7, and the impurities in the second pipe 82 are brought to the waste bin together, which can better dredge the second pipe 82 and facilitate impurity removal.

[0079] Among them, the water supply tank 11 includes a box body 113 and a cover 113 matching the box body 113, the cover 113 is provided with a sealing hole 115, and the inner wall of the box body 113 is provided with a valve core 116, and the valve core 116 is connected to a float 118 matching the sealing hole 115 through a valve stem 117. When the water level reaches the set position, the float 118 drives the valve core 116 on the valve stem 117 to seal the sealing hole 115 of the cover 113, thereby achieving the purpose of automatically closing the water inlet of the water supply tank 11 and avoiding waste of water resources.

[0080] In this scheme, it also includes a cutter head cleaning needle 12 and an air pressure sensor 13. The cleaning needle 12 is arranged on the lower fixed plate 1c, and the air pressure sensor 13 is installed in the first pipeline 81 and is connected to the PLC controller 92 in the electrical controller 9; the cutter head cleaning needle 12: after the cutter head cleaning function is selected, the picking cutter head 6 will move here and move up and down to dredge the picking cutter head 6. When the negative pressure of the first pipeline 81 exceeds the set value, it is determined that the picking cutter head 6 is blocked. The PLC controller 92 orders the picking cutter head 6 to move to the cutter head cleaning needle 12, move up and down quickly, and dredge the picking cutter head 6 until the pipeline air pressure returns to normal, thereby completing the purpose of automatically cleaning the picking cutter head 6 and facilitating the cutting of impurities.

[0081] In this solution, baffles 13 are provided at both ends of the lower fixed plate 1c to prevent water from dripping during the picking process and keep the equipment clean.

[0082] In this solution, a plurality of drainage holes (not shown) are arranged around the inspection port 1e. The picking cutter head 6 may produce some water marks during the water absorption process and the process of picking wet materials, and the excess water can be discharged in time through the drainage holes.

[0083] like Fig.11 As shown, the present invention provides an automatic bird's nest selection method based on visual recognition technology, and the automatic bird's nest selection method needs to provide the above-mentioned automatic bird's nest selection device based on visual recognition technology, and the method includes:

[0084] Step 1, select a picking mold 3 (dry picking mold 3a or wet picking mold 3b) according to the variety of the bird's nest material to be inspected, place the selected picking mold 3 at the inspection port 1e, and then place the bird's nest material to be inspected on the picking mold 3; and install a picking knife head 6 (dry picking knife head 6a or wet picking knife head 6b) matching the selected picking mold 3 at the front end of the mechanical arm 5; select according to different varieties of bird's nests, use two kinds of picking molds 3 and picking knife heads 6, which can be applicable to all bird's nest materials on the market, and are convenient for detection and impurity removal;

[0085] Step 2: Divide the bird's nest material to be inspected into a plurality of inspection areas of equal area according to the lens field of view of the image collector 2;

[0086] In this embodiment, since the surface of the bird's nest material to be inspected is larger than the field of view of the lens, that is, the field of view of the lens does not cover the entire surface of the bird's nest material to be inspected, the lens cannot capture the entire picture of the bird's nest material to be inspected at one time, so it is necessary to divide the bird's nest material to be inspected into a plurality of detection areas of equal area within the field of view of the lens to ensure that each detection area can be captured by the lens; the area of ​​all detection areas superimposed is required to be larger than the area of ​​the bird's nest material to be inspected, that is, it is not required that the boundary of each detection area is precisely fitted, and there may be overlapping parts in each detection area, which can reduce the difficulty of shooting and ensure that every position of the bird's nest material to be inspected can be detected;

[0087] Step 3, turn on the illumination light source 14 for illumination, select a detection area, the lens 21 of the image collector 2 collects a color image of the detection area through the hollow area 141, and transmits the color image to the electrical control box 9;

[0088] In this embodiment, the illumination light source 14 is connected to the PLC controller 92 of the electrical control box 9. When the detection starts, the PLC controller 92 controls the illumination light source 14 to automatically turn on and keep it on, providing a continuous light source for the detection area.

[0089] Step 4: After receiving the color image, the electrical control box 9 converts the color image into a grayscale image;

[0090] In this embodiment, the color image is received by the PLC controller 92 of the electrical control box 9, and the color image is converted into a grayscale image using the rgb2gray() function. This has the advantage of reducing the amount of subsequent image calculations, thus preparing for subsequent operations such as image segmentation, image recognition, and image analysis.

[0091] Step 5: After filtering and enhancing the obtained grayscale image, the corresponding grayscale value is obtained and sent to the electrical control box 9 (in this embodiment, it is sent to the PLC controller 92 of the electrical control box 9);

[0092] In this embodiment, the gray image is filtered by distinguishing target areas and background areas of different gray levels by using appropriate thresholds, thereby generating a corresponding binary image and obtaining a corresponding gray value; that is, the entire image is presented with an obvious visual effect of only black and white;

[0093] The enhancement processing method of the gray image is: using contrast enhancement method, by modifying the grayscale of each pixel of the gray image, the dynamic range of the image grayscale is changed, and the imadjust function of MatLab is used to adjust the brightness of the gray image;

[0094] Step 6, the electrical control box 9 (in this embodiment, the PLC controller 92 sent to the electrical control box 9) determines the impurity situation by comparing the gray value with the set value. If the gray value does not exceed the set value, it is determined that the pixel point of the detection area is normal, and the detection area does not have impurities, and returns to step 3 to select the next detection area for processing; if the gray value exceeds the set value, it is determined that the pixel point of the detection area is impurity, and enters step 7;

[0095] Step 7, the electrical control box 9 (in this embodiment, the PLC controller 92 sent to the electrical control box 9) obtains the coordinate position of the impurity, and controls the robot arm 5 to move the picking cutter head 6 to the coordinate position of the impurity through the operating module 4;

[0096] In this embodiment, image calibration is used to match the target (impurity) with the image captured by the image collector 2, determine the proportional relationship between the two, and solve the corresponding relationship between the target (impurity) and the image by establishing a pinhole camera model and using a linear method. According to the proportional relationship and the corresponding relationship, the target (impurity) is converted from the image position to the coordinate position. The advantage of this is that after the impurity position is determined, the robot arm 5 can perform the impurity removal operation;

[0097] Step 8: After the picking cutter head 6 descends, it cuts off the impurities and sucks the impurities out to the waste box 8 through the negative pressure of the vacuum pump 7; select the next detection area and repeat steps 3-8 until the impurities in all detection areas are removed;

[0098] In this embodiment, the cutter head cutting is currently a more feasible automated impurity removal method. After the impurities are located, the impurities and the bird's nest are cut off and sucked out together; the impurities are removed by the cooperation of the picking cutter head 6 and the vacuum pump 7, which can achieve more accurate cutting;

[0099] Step 9, the pressure sensor 10 sends the pressure value to the electrical control box 9 (in this embodiment, it is sent to the PLC controller 92 of the electrical control box 9) in real time. The electrical control box 9 determines whether the picking mold 3 is damaged according to the pressure value. If the pressure value does not exceed the set value, it means that no damage is caused and no processing is performed; if the pressure value exceeds the set value, it means that damage is caused, and the electrical control box 9 controls the mechanical arm 5 through the operating module 4 to move the picking head 6 upward; the pressure sensor 10: is used to sense the pressure between the picking head 6 and the contact surface to avoid the picking mold 3 being damaged by the picking head 6, resulting in a decrease in the recognition accuracy of defects. When the pressure exceeds the set value, the PLC controller 92 sends a signal to the mechanical arm 5 to control the picking head 6 to move upward. Through this setting, the original strip can be protected to the greatest extent during the picking process.

[0100] Step 10, when the second pipe 82 is blocked, move the second pipe 82 to the water suction port 111, and introduce water from the second pipe 82 into the waste bin 8 through the vacuum pump 7; the water suction port 111: use the water inlet pipe 112 to introduce water from the water supply tank 11 here. After a certain number of pickings, the second pipe 82 is prone to blockage. Control the robot arm 5 to move the port of the second pipe 82 fixed on the picking head 6 to this place, and use the vacuum of the vacuum pump 7 to suck water into the second pipe 82, and bring the impurities in the second pipe 82 into the waste bin together, which can better dredge the second pipe 82 and facilitate impurity removal.

[0101] Step 11, the air pressure sensor 13 sends the air pressure value to the electrical control box 9 in real time (in this embodiment, it is sent to the PLC controller 92 of the electrical control box 9). The electrical control box 9 determines whether the picking cutter head 6 is blocked according to the air pressure value. If the air pressure value does not exceed the set value, it means that the picking cutter head 6 is not blocked and no processing is performed; if the air pressure value exceeds the set value, it means that the picking cutter head 6 is blocked, and the electrical control box 9 controls the mechanical arm 5 through the operating module 4 to move the picking cutter head 6 upward to the cutter head cleaning needle 12 for cleaning; the cutter head cleaning needle 12: after the cutter head cleaning function is selected, the picking cutter head 6 will move here and move up and down to dredge the picking cutter head 6. When the negative pressure of the first pipeline 81 exceeds the set value, it is determined that the picking cutter head 6 is blocked. The PLC controller 92 orders the picking cutter head 6 to move to the cutter head cleaning needle 12, move up and down quickly, and dredge the picking cutter head 6 until the pipeline air pressure returns to normal, completing the purpose of automatically cleaning the picking cutter head 6, and facilitating the cutting of impurities.

[0102] There is no order of precedence between step 9, step 10 and step 11.

[0103] The working principle of this device is: this device has two picking modes to choose from, namely dry picking mode and wet picking mode.

[0104] A. Dry material picking mode: Install the dry picking mold 3a on the inspection port 1e, replace the dry picking cutter head 6a dedicated to dry picking, and turn off the water absorption function at the same time. Place the dry bird's nest and bird's nest on the spherical support seat 33 of the dry picking mold 3a. After turning on the detection, the PLC controller 92 in the electrical controller 9 instructs the industrial camera lens to move to the specified position to take pictures of the area to be inspected. After the image is processed by the algorithm, the color image is converted into a black and white image, and the impurities are marked with a red circle according to the set black point value and gray value parameters. The picking cutter head 6 is moved to the location of the impurities by the mechanical arm 5, and the high-speed rotating dry picking cutter head 6a is lowered to take out and remove the impurities. During the descent of the dry picking cutter head 6a, the PLC controller 92 on the electrical control box 9 monitors the value of the pressure sensor 10 on the mechanical arm 5 in real time. When the pressure value exceeds the set value, the mechanical arm 5 automatically rises to complete a de-impurity. This process is repeated until all the impurities in the area to be inspected are removed.

[0105] B. Wet material picking mode: Spread the wet bird's nest after soaking and stripping on the wet picking mold 3b, and then place it in the area to be inspected. After the detection is turned on, the PLC controller 92 in the electrical controller 9 divides the area to be inspected into several blocks according to the field of view of the industrial camera, and takes pictures one by one from the origin. After the image is processed by the algorithm, the color image is converted into a black and white image, and the impurities are marked with red circles according to the set black point value and gray value parameters. The PLC controller 92 in the electrical controller 9 sends the coordinates of the location of the impurity to the robot arm, and the wet picking head 6b is moved to the location of the impurity by the robot arm. After the picking head 6b descends, the impurities are sucked out by negative pressure. During the descent of the picking head 6b, the PLC controller 92 monitors the value of the pressure sensor on the robot arm in real time. When the pressure value exceeds the set value, the wet picking head automatically rises to complete one impurity removal. This goes back and forth until all the impurities in the area to be inspected are removed. The impurities are uniformly collected in the waste barrel through the second pipeline, and these impurities can be sorted twice to reduce the overall picking loss.

[0106] Embodiment 1:

[0107] Take 100 grams of bird's nest raw materials, soak them in purified water for 4-6 hours, rinse them, drain them with a 60-mesh filter, and then divide them into two groups after drying. The experimental group was automatically selected, and the control group was manually selected.

[0108]

[0109] Embodiment 2:

[0110] Take 200 grams of bird's nest raw materials, soak them in purified water for 4-6 hours, rinse them, drain them with a 60-mesh filter, and then divide them into two groups after drying. The experimental group was automatically picked, and the control group was manually picked.

[0111]

[0112] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An automatic bird's nest selection device based on visual recognition technology, characterized in that: include: A support frame, the support frame comprises an upper fixing plate, a middle fixing plate and a lower fixing plate, the two ends of the middle fixing plate are respectively connected to the upper fixing plate and the lower fixing plate; a light-through hole is provided on the upper fixing plate, and an inspection port is provided below the lower fixing plate; An illumination light source, the illumination light source is installed on the light-through hole, and a hollow area is provided in the middle of the illumination light source; An image collector, which is installed just above the hollow area; An inspection mold is placed on the inspection port and installed just below the hollow area; An operating module, wherein the operating module is installed at the front end of the lower fixing plate; A mechanical arm, the lower end of which is mounted on a lower fixing plate; A picking cutter head, which is arranged at the upper end of the mechanical arm; when working, the picking cutter head is driven to the inspection mold by controlling the mechanical arm to inspect the impurities; Vacuum pump; A waste box, one end of which is connected to a vacuum pump through a first pipe, and the other end of which is connected to a picking cutter head through a second pipe; the impurities cut off by the picking cutter head are recovered from the second pipe to the waste box through the vacuum pump; and an electrical control box, wherein the electrical control box is respectively connected to the image collector, the illumination light source, the operation module, the mechanical arm and the vacuum pump; It also includes: a pressure sensor, which is arranged on the picking cutter head and connected to the electrical control box; It also includes a cutter head cleaning needle and an air pressure sensor, wherein the cleaning needle is arranged on the lower fixing plate, and the air pressure sensor is installed in the first pipeline and connected to the electrical controller; The electrical control box includes an input power supply, a PLC controller and an IO control module, an air switch box and a reset indicator light, and the PLC controller is respectively connected to the input power supply, the IO control module, the air switch box, the reset indicator light, the image collector, the lighting source, the operation module, the mechanical arm, the vacuum pump, the pressure sensor and the air pressure sensor; Both ends of the lower fixed plate are provided with enclosures; a plurality of drainage holes are provided around the inspection port; the image collector is an industrial camera, and the operating module is a key switch.

2. According to claim 1, the automatic bird's nest selection device based on visual recognition technology is characterized in that: Also includes: A water supply box, wherein the lower fixed plate is provided with a water suction port, and the water supply box is connected to the water suction port through a water inlet pipe; when cleaning the second pipe, the second pipe is moved to the water suction port, and water is introduced into the waste tank from the second pipe through a vacuum pump; The water supply tank includes a box body and a cover matching the box body, the cover is provided with a sealing hole, the inner wall of the box body is provided with a valve core, the valve core is connected to a float matching the sealing hole through a valve stem, and when the water level reaches a set position, the float drives the valve core on the valve stem to seal the sealing hole of the cover.

3. The automatic bird's nest selection device based on visual recognition technology according to claim 1 is characterized in that: The inspection mold is a dry picking mold or a wet picking mold. The dry picking mold includes a placement plate, a connecting column and a support seat. The placement plate is connected to the support seat through the connecting column. The placement plate is a polypropylene plate. The support seat is a spherical solid rubber seat. The support seat is used to place the dry bird's nest and bird's nest to be inspected; the wet picking mold includes a placement plate, which is a polypropylene plate for placing the wet bird's nest material to be inspected.

4. A method for automatically selecting bird's nests based on visual recognition technology, characterized in that: The method for automatically selecting bird's nests needs to provide an automatic bird's nest selection device based on visual recognition technology as claimed in claim 1, and the method comprises: Step 1, select a picking and inspection mold according to the type of bird's nest material to be inspected, place the selected picking and inspection mold at the inspection port, and then place the bird's nest material on the picking and inspection mold; and install a picking cutter head matching the selected picking and inspection mold at the front end of the mechanical arm; Step 2: Divide the bird's nest material to be inspected into a plurality of inspection areas of equal area according to the lens field of view of the image collector; Step 3: Turn on the illumination light source for illumination, select a detection area, and the lens of the image collector collects a color image of the detection area through the hollow area, and transmits the color image to the electrical control box; Step 4: After receiving the color image, the electrical control box converts the color image into a grayscale image; Step 5: After filtering and enhancing the obtained grayscale image, the corresponding grayscale value is obtained and sent to the electrical control box; Step 6, the electrical control box determines the impurity situation by comparing the gray value with the set value. If the gray value does not exceed the set value, it is determined that the pixel points in the detection area are normal, and there is no impurity in the detection area, and the process returns to step 3 to select the next detection area for processing; if the gray value exceeds the set value, it is determined that the pixel points in the detection area are impurities, and the process goes to step 7; Step 7, the electrical control box obtains the coordinate position of the impurity, and controls the robot arm to move the picking cutter head to the coordinate position of the impurity through the operation module; Step 8: After the picking cutter head descends, it cuts off the impurities and sucks the impurities out to the waste box through the negative pressure of the vacuum pump; select the next detection area and repeat steps 3-8 until the impurities in all detection areas are removed.

5. The method for automatically selecting bird's nests based on visual recognition technology according to claim 4, characterized in that: The method includes: Step 9, the pressure sensor sends the pressure value to the electrical control box in real time, and the electrical control box determines whether the inspection mold is damaged based on the pressure value. If the pressure value does not exceed the set value, it means that no damage is caused and no processing is performed; if the pressure value exceeds the set value, it means that damage is caused, and the electrical control box controls the robotic arm through the operating module to move the picking head upward.

6. The method for automatically selecting bird's nests based on visual recognition technology according to claim 5, characterized in that: The method for automatically selecting bird's nests also needs to provide a water supply box, the lower fixed plate is provided with a water suction port, and the water supply box is connected to the water suction port through a water inlet pipe; the method comprises: Step 10: When the second pipe is blocked, move the second pipe to the water suction port, and introduce water from the second pipe into the waste tank through a vacuum pump.

7. The method for automatically selecting bird's nests based on visual recognition technology according to claim 6, characterized in that: The method includes: Step 11, the air pressure sensor sends the air pressure value to the electrical control box in real time, and the electrical control box determines whether the picking cutter head is blocked according to the air pressure value. If the air pressure value does not exceed the set value, it means that the picking cutter head is not blocked and no processing is performed; if the air pressure value exceeds the set value, it means that the picking cutter head is blocked, and the electrical control box controls the robot arm through the operation module to move the picking cutter head upward to the cutter head cleaning needle for cleaning; there is no order among step 9, step 10 and step 11.

Images