A visual recognition-based poultry egg plating system

Abstract

This invention relates to the field of automated poultry egg identification, sorting, and traying technology, and in particular to a vision-based poultry egg traying system. It includes: an egg-collecting area, an egg-placing area, an image detection unit, and an egg-collecting robot; wherein, the egg-collecting area provides a placement area for the eggs before they are picked up; wherein, the egg-placing area is used to place egg trays to hold the eggs; wherein, the image detection unit detects the eggs and provides the egg-collecting robot with the egg-grabbing position, the egg-placing position, and the position of the larger end of the egg; wherein, the egg-collecting robot picks up the eggs from the egg-collecting area and places them one by one into the egg-placing positions on the egg trays; by using the egg-collecting robot to pick up and place the eggs, the rolling of the eggs on the drive rollers can be avoided, reducing secondary contamination of the eggs during the traying process and reducing the labor intensity of manual labor.

Description

Technical Field

[0001] This invention relates to the field of automated poultry egg identification, sorting, and traying technology, and in particular to a poultry egg traying system based on visual recognition. Background Technology

[0002] In my country, there is a high demand for fresh poultry eggs (poultry eggs, duck eggs, etc.). Every day, farms need to collect fresh poultry eggs, remove unqualified eggs, put them into egg trays, and send them to the market in batches for sale.

[0003] For the packaging of poultry eggs, small farms usually use manual screening to remove eggs that are contaminated or broken. The qualified eggs are then packaged. To ensure the quality of the eggs, the larger end with the air cell should be placed upwards during packaging to prevent contamination during transportation and storage and to prolong the shelf life of the eggs.

[0004] In some larger farms, longer egg conveyor lines are used to collect poultry eggs in a centralized manner. During the egg conveying process, the eggs move along two drive rollers in the same direction. At this time, the eggs will move in the direction pointed to by the smaller end. When the smaller end moves to the end in the opposite direction to the predetermined direction, it will be blocked and turned around. Finally, after the direction is unified, the eggs are sent to the egg tray for loading.

[0005] Of the two existing methods, manual egg loading requires high speed and is labor-intensive when egg production is high. Mechanical egg sorting and turning, on the other hand, involves moving and tumbling the eggs along a long conveyor belt, increasing the likelihood of surface contamination and potential damage. Summary of the Invention

[0006] To solve the above problems, the technical solution adopted by the present invention is as follows:

[0007] This invention provides a visual recognition-based poultry egg plating system, comprising:

[0008] Egg retrieval area, egg placement area, image detection unit, and egg retrieval robotic arm;

[0009] The egg-collecting area provides a place for placing poultry eggs before they are picked up;

[0010] The egg-laying area is used to place egg trays that hold poultry eggs;

[0011] The image detection unit detects the eggs and provides the egg-retrieving robot with the egg-grabbing position, the egg-placement position, and the position of the large end of the egg.

[0012] The egg-collecting robotic arm grabs the eggs from the egg-collecting area and places them one by one into the egg placement positions on the egg tray.

[0013] The egg-retrieving robotic arm includes a controller that receives information from the image detection unit and controls the robotic arm to move between the egg-retrieving area and the egg-placing area, grasping and placing the eggs; and...

[0014] The controller, based on the egg placement position provided by the image detection unit, shifts the larger end of the egg above the egg placement position to outside the vertical projection outline of the egg placement position, and then places the egg down.

[0015] Furthermore, the image detection unit includes a camera component, an egg position detection module in the egg-collecting area, an egg end detection module, and an egg placement position detection module;

[0016] The camera component is used to capture real-time images of the egg retrieval area and the egg placement area;

[0017] The egg location detection module in the egg collection area detects the location of the eggs in the egg collection area based on real-time images and provides the location information of the eggs to the egg collection robot arm.

[0018] The egg end detection module detects the large and small ends of the eggs in the egg collection area and provides the location of the large end of the egg to the egg collection robot.

[0019] The egg placement detection module detects multiple egg placement positions on the egg tray and provides the egg placement information to the egg-retrieving robot.

[0020] Furthermore, the egg location detection module in the egg collection area pre-acquires the data model of the egg collection area and obtains the outline of the egg collection area based on real-time images;

[0021] Select a real-time image of one of the eggs in the egg collection area. Based on the location of the egg in the egg collection area in the real-time image, obtain the length ratio of the egg relative to the outline of the egg collection area. Compare this ratio with the data model of the egg collection area to obtain the location of the egg in the egg collection area.

[0022] Furthermore, the egg end detection module performs end detection on the egg after the egg position detection module in the egg collection area selects the egg;

[0023] During end detection, the outline of the egg is first obtained. The two points A1 and A2 that are farthest apart on the outline of the egg are selected, and a line is drawn with A1 and A2 as endpoints to obtain the central axis A1A2 of the egg.

[0024] Draw dividing lines B1B2, which are perpendicular to the central axis A1A2, pass through the midpoint of the central axis A1A2, and intersect with two points B1 and B2 on the outline of the egg respectively.

[0025] Then calculate the area of ​​the egg outlines on both sides of the dividing line B1B2. Record point A1 or A2 on the side with the larger area as the large end of the egg, and record point A2 or A1 on the side with the smaller area as the small end of the egg.

[0026] Furthermore, the method of selecting the longest point through continuous radiation is used to select the farthest points A1 and A2;

[0027] First, arbitrarily select a point I1 on the outline of the egg. Using point I1 as the origin, draw multiple straight lines radially within the outline of the egg. Select the longest straight line I1I2 within the outline of the egg.

[0028] Draw multiple straight lines radiating outwards from point I2 within the outline of the egg, and select the longest straight line I2I3.

[0029] Continue with the farthest point of radiation I n-1 Line I drawn with the origin n-1 I n Continue until the two points A1 and A2 that are furthest apart within the outline of the egg are selected.

[0030] Furthermore, the egg placement detection module pre-obtains the distribution of egg placement positions on the egg tray, as well as the distance information between egg placement positions on the egg tray;

[0031] The egg placement detection module pre-acquires the data model of the egg placement area and obtains the outline of the egg placement area based on real-time images;

[0032] Select the real-time image outline of one of the egg placement positions as the basic egg placement position. Based on the location of the basic egg placement position in the egg-laying area, obtain the length ratio of the basic egg placement position relative to the outline of the egg-laying area. Compare this ratio with the data model of the egg-laying area to obtain the location of the basic egg placement position.

[0033] The location information of the remaining egg placement positions is calculated by using the basic egg placement positions, the distribution of egg placement positions, and the distance information between egg placement positions on the egg tray.

[0034] Furthermore, selecting the real-time image contour of the egg placement location also includes selecting a reference egg placement location contour;

[0035] The reference egg placement position is the egg placement position on the egg tray other than the basic egg placement position;

[0036] And obtain the interval egg placement information between the basic egg placement position and the reference egg placement position;

[0037] The location information of the egg placement positions is adjusted by comparing the other egg placement positions calculated with the reference egg placement positions.

[0038] Furthermore, it also includes a non-conforming poultry egg placement area, which is used to place non-conforming poultry eggs;

[0039] The controller directs the egg-retrieving robot to deliver substandard eggs to the substandard egg placement area.

[0040] Furthermore, the image detection unit also includes an egg quality monitoring module, which acquires images of the egg surface through real-time images and compares the egg surface color and texture with qualified eggs.

[0041] The poultry egg quality monitoring module provides information on substandard poultry eggs to the egg-collecting robot, which then delivers the substandard poultry eggs to the substandard poultry egg placement area.

[0042] The camera assembly also includes a bottom image acquisition camera for eggs, which acquires images of the bottom of eggs and provides them to the egg quality monitoring module.

[0043] Furthermore, an egg tray placement direction detection component is provided at the bottom of the egg placement area, and the egg tray placement direction detection component includes two sets of position sensors;

[0044] The two sets of position sensors are spaced 1 / 2 the distance between the egg placement positions, both horizontally and vertically.

[0045] Furthermore, the larger end of the egg is offset to 1 / 3-1 / 4 of the length of the larger end and the length in the direction of the larger end of the egg beyond the vertical projection outline of the egg placement position;

[0046] The egg-retrieving robot places the egg when the bottom of the egg is 1-8 mm away from the outline of the egg placement position.

[0047] Furthermore, the egg-retrieving robotic arm also includes a robotic arm, a negative pressure suction cup, and a negative pressure assembly;

[0048] The negative pressure suction cup adsorbs and grasps the poultry eggs, and the negative pressure component provides adsorption power for the negative pressure suction cup;

[0049] The robotic arm drives the negative pressure suction cup to move.

[0050] The beneficial effects of this invention are reflected in:

[0051] The present invention provides a visual recognition-based poultry egg loading system, which acquires images of the egg picking area and the egg placing area through an image detection unit, and detects the position of the poultry eggs being picked up and placed, providing the egg picking robot with position information during the process of picking up and placing poultry eggs, as well as identifying the large end of the poultry eggs.

[0052] The egg-retrieving robot grabs the eggs and, when placing them down, makes the larger end of the egg extend beyond the vertical projection outline of the egg placement position. After the egg is placed down, the larger end of the egg is blocked by the outline of the egg placement position, while the smaller end of the egg flips downwards, and finally the larger end of the egg is placed in the egg placement position with the larger end facing upwards.

[0053] By using a robotic arm to grab and place eggs, the eggs can be prevented from rolling on the transmission rollers, reducing secondary contamination during the loading process and lowering the labor intensity of manual labor. Attached Figure Description

[0054] Figure 1 This is a schematic diagram of the poultry egg tray-loading system based on visual recognition provided by the present invention;

[0055] Figure 2 This is a schematic diagram of the egg-collecting area provided by the present invention;

[0056] Figure 3 A schematic diagram of an existing egg tray structure;

[0057] Figure 4 This is a schematic diagram showing the overlapping state of the egg trays;

[0058] Figure 5 This is a schematic diagram illustrating the process of an egg flipping over on its tray.

[0059] Figure 6 This is a block diagram of the image detection unit structure;

[0060] Figure 7 A schematic diagram of the central axis and dividing lines on the outline of a poultry egg;

[0061] Figure 8 A schematic diagram illustrating the process of taking the longest point of continuous radiation provided by the present invention;

[0062] Figure 9 for Figure 8 A diagram showing the longest line connecting the points in the diagram;

[0063] Figure 10 This is a schematic diagram showing the location of the position sensor in the egg-laying area;

[0064] Figure 11 A comparison image of two trays flipped 180°;

[0065] Figure 12This is a schematic diagram of the egg-retrieving robotic arm.

[0066] Figure label:

[0067] Egg collection area 1;

[0068] Egg placement area 2; Position sensor 21;

[0069] Camera component 31, bottom image acquisition camera 311 for poultry eggs;

[0070] 4. Egg-retrieving robotic hand; 41. Robotic arm; 42. Negative pressure suction cup; 43. Negative pressure component;

[0071] Area 5 for storing substandard poultry eggs;

[0072] Egg tray 6, poultry egg placement 61, basic poultry egg placement 62, reference poultry egg placement 63;

[0073] 7. Poultry eggs. Detailed Implementation

[0074] The present invention will now be described in further detail with reference to the accompanying drawings and embodiments thereof.

[0075] Example 1

[0076] Reference Figures 1-12 This invention provides a visual recognition-based poultry egg plating system, comprising:

[0077] Egg retrieval area 1, egg placement area 2, image detection unit and egg retrieval robotic arm 4;

[0078] The egg-collecting area 1 provides a place for the poultry eggs 7 before they are picked up;

[0079] The egg placement area 2 is used to place the egg tray 6 that holds the poultry eggs 7.

[0080] In this embodiment, the egg collection area 1 can be a flat area connected to the egg collection conveyor belt set under the poultry house, or the end of the egg collection conveyor belt, or a flat area. The poultry eggs 7 to be picked up by the egg collection robot 4 are all placed in the egg collection area 1.

[0081] Egg placement area 2 is similar to egg retrieval area 1. The egg trays 6 placed in egg placement area 2 are the commercially available egg trays 6, for example... Figure 4 The paper pulp egg tray 6 shown has several small holes evenly arranged on it, and each small hole can hold an egg 7. The small hole is the egg 7 placement position 61 in this embodiment.

[0082] After filling all the holes in an egg tray 6 with eggs 7, the tray is finished. Once an egg tray 6 is filled with eggs 7, it can be removed manually or mechanically, and an empty egg tray 6 can be placed in to continue filling. The filled egg trays 6 only need to be transferred to a vehicle and transported to the market, reducing the labor intensity.

[0083] The image detection unit detects the eggs 7 and provides the egg-grabbing robot 4 with the egg-grabbing position, the placement position of the eggs 7, and the position of the large end of the eggs.

[0084] The testing of poultry eggs can screen out broken and contaminated eggs. The positions where the poultry eggs are picked up and placed are the positions where the egg-picking robot 4 picks up and places the eggs in the corresponding egg-picking area 1 and egg-laying area 2. Identifying the position of the large end of the poultry egg can provide a positional reference for the flipping of the poultry egg during the egg-laying process.

[0085] The egg-collecting robot 4 grabs the poultry eggs in the egg-collecting area 1 and places them one by one into the poultry egg placement position 61 of the egg tray 6;

[0086] The egg-retrieving robotic arm 4 includes a controller that receives information from the image detection unit and controls the robotic arm to move between the egg-retrieving area 1 and the egg-placing area 2 to grasp and place the eggs; and...

[0087] The controller, based on the egg placement position 61 provided by the image detection unit, shifts the larger end of the egg above the egg placement position 61 to outside the vertical projection outline of the egg placement position 61, and then places the egg down.

[0088] like Figure 5 As shown, at this point, the larger end of the egg grasped by the egg-retrieving robot 4 is outside the egg placement position 61, i.e., the small hole for placing the egg. After the egg-retrieving robot 4 releases its grip on the egg, the egg falls vertically. The larger end of the egg first contacts the outline of the egg placement position 61, while the smaller end, located above the small hole, is not obstructed. After the larger end is obstructed, the smaller end continues to fall, causing the egg to flip vertically, with the smaller end falling downwards into the egg placement position 61. The egg, after being placed with its larger end facing upwards, avoids contamination during transportation and storage, and prolongs the egg's shelf life.

[0089] The automatic flipping of eggs with the larger end facing upwards during egg loading simplifies the movement of the egg-retrieving robot 4. Once the robot 4 moves the egg to the egg placement position 61 and the larger end is outside the outline of the position, it releases the egg. This eliminates the need for the robot 4 to flip the egg to the larger end before placing it in the position 61, effectively simplifying the structure and process control of the robot 4, reducing the space required for its movement, and lowering equipment costs.

[0090] Example 2

[0091] Reference Figure 6 Furthermore, the image detection unit includes a camera component 31, an egg position detection module in the egg-collecting area, an egg end detection module, and an egg placement position 61 detection module;

[0092] The camera component 31 is used to capture real-time images of the egg retrieval area and the egg placement area 2;

[0093] The camera assembly 31 may consist of one or more cameras, positioned above the egg-collecting area and the egg-laying area 2 (e.g., Figure 1 As shown in the figure, images of the egg collection area and the egg placement area 2 are collected in real time, providing a basis for position recognition in the process of loading poultry eggs onto trays.

[0094] The egg position detection module in the egg-collecting area detects the position of the eggs in the egg-collecting area based on real-time images and provides the position information of the eggs to the egg-collecting robot 4. The egg-collecting robot 4 uses the position information provided by the egg position detection module as the gripping position for the eggs.

[0095] The egg end detection module detects the large and small ends of the eggs in the egg-collecting area and provides the location of the large end of the egg to the egg-collecting robot 4. After identifying the large and small ends of the egg, the egg-collecting robot 4 places the egg according to the position information of the large end provided by the egg end detection module. When the egg is placed into the egg tray 6, the large end of the falling egg is blocked, and the small end flips down and falls into the egg placement position 61.

[0096] The egg placement position 61 detection module detects multiple egg placement positions 61 on the egg tray 6 and provides the information of the egg placement positions 61 on the egg tray 6 to the egg-retrieving robot 4. The egg-retrieving robot 4 places the eggs one by one into the egg tray 6 according to the information of the egg placement positions 61 provided by the egg placement position 61 detection module.

[0097] By using multiple position acquisition modules to simultaneously detect the position during the egg retrieval and tray loading process, the required position information for the egg retrieval robot 4 can be quickly provided.

[0098] Example 3

[0099] Furthermore, the egg location detection module in the egg collection area pre-acquires the data model of the egg collection area and obtains the outline of the egg collection area based on real-time images;

[0100] Select a real-time image of one of the eggs in the egg collection area. Based on the location of the egg in the egg collection area in the real-time image, obtain the length ratio of the egg relative to the outline of the egg collection area. Compare this ratio with the data model of the egg collection area to obtain the location of the egg in the egg collection area.

[0101] In this embodiment, a data model of the egg-retrieving area is pre-acquired as a reference for the egg-retrieving robot 4 during egg retrieval. The image outline of the egg-retrieving area and the data model of the egg-retrieving area can be compared and overlapped. The actual size data of the egg-retrieving area cannot be obtained from the image alone. However, after selecting the outline of an egg in the egg-retrieving area, the visual position of the egg in the egg-retrieving area can be obtained. Then, the position of the egg in the egg-retrieving area is mapped onto the data model of the egg-retrieving area for comparison, thereby obtaining the actual position information of the egg in the egg-retrieving area.

[0102] When the image acquisition viewpoint is tilted towards the egg-collecting area, the contour image of the tilted egg-collecting area can be restored to the shape corresponding to the data model during comparison, thus obtaining accurate egg location information and avoiding egg-collecting errors caused by camera installation errors.

[0103] It should be noted that the egg-retrieving area model data model described in this application includes the shape, size, and position of the egg-retrieving area relative to the motion base point of the egg-retrieving robot 4.

[0104] Example 4

[0105] Reference Figure 7 Furthermore, the egg end detection module performs end detection on the egg after the egg position detection module in the egg collection area selects the egg;

[0106] During end detection, the outline of the egg is first obtained. The two points A1 and A2 that are farthest apart on the outline of the egg are selected, and a line is drawn with A1 and A2 as endpoints to obtain the central axis A1A2 of the egg.

[0107] Draw dividing lines B1B2, which are perpendicular to the central axis A1A2, pass through the midpoint of the central axis A1A2, and intersect with two points B1 and B2 on the outline of the egg respectively.

[0108] Then calculate the area of ​​the egg outlines on both sides of the dividing line B1B2. Record point A1 or A2 on the side with the larger area as the large end of the egg, and record point A2 or A1 on the side with the smaller area as the small end of the egg.

[0109] Because poultry eggs are oval-shaped with a large end and a small end, when placed naturally on a flat surface, the middle of the egg contacts the surface, while the two ends point to the sides.

[0110] like Figure 7 As shown, the two endpoints A1 and A2 of the central axis A1A2 are located at the large end and the small end of the egg, respectively. The dividing line B1B2 drawn with the midpoint of the central axis as the passing point divides the outline of the egg into two regions along the length of the central axis. These two regions are located in the direction of the large end and the small end, respectively, so their areas must be different. The area of ​​the side with the large end is larger than that of the side with the small end.

[0111] Using point A1 or A2 as a marker for the large end, when grasping and placing the eggs in the egg placement position 61, the location of A1 or A2 can make the marker position of the large end of the egg clearer, that is, it can more accurately block the large end of the egg, and the process of the small end flipping and falling into the egg placement position 61 during the fall is also more accurate.

[0112] Example 5

[0113] Reference Figure 8 , Figure 9 Furthermore, the method of selecting the longest point through continuous radiation is adopted to select the farthest points A1 and A2;

[0114] First, arbitrarily select a point I1 on the outline of the egg. Using point I1 as the origin, draw multiple straight lines radially within the outline of the egg. Select the longest straight line I1I2 within the outline of the egg.

[0115] Draw multiple straight lines radiating outwards from point I2 within the outline of the egg, and select the longest straight line I2I3.

[0116] Continue with the farthest point of radiation I n-1 Line I drawn with the origin n-1 I n Continue until the two points A1 and A2 that are furthest apart within the outline of the egg are selected.

[0117] When selecting the longest point in the radial pattern, line I... n-1 I n The longest continuous line I within the outline of the egg is selected from multiple consecutive rays. n-1 I n Then, point I n-1 and point I n Eventually, the energy will converge towards the two furthest points of the egg, which are the large end and the small end of the egg, radiating continuously to point I. n-1 and point I nWhen the selected location is repeated from the previous radiation, the two points A1 and A2 that are furthest apart inside the egg can be obtained. Figure 9 (as described in I4 and I5).

[0118] A1 and A2 can usually be obtained in 3-5 attempts, while radiation point I n-1 and point I n When the selected location is repeated from the previous radiation, the accuracy can be appropriately reduced. It does not need to be completely coincident. As long as the distance is within a certain range, it can be considered as the two farthest points, so as to improve the identification speed of points A1 and A2.

[0119] Example 6

[0120] Reference Figure 3 Furthermore, the egg placement position 61 detection module pre-obtains the distribution of the egg placement positions 61 on the egg tray 6, as well as the distance information between the egg placement positions 61 on the egg tray 6;

[0121] The detection module for the egg placement position 61 pre-acquires the data model of the egg placement area and obtains the outline of the egg placement area based on the real-time image;

[0122] Select the real-time image outline of one of the egg placement positions 61, which is used as the basic egg placement position 62. Based on the location of the basic egg placement position 62 in the egg-laying area, obtain the length ratio of the basic egg placement position 62 relative to the outline of the egg-laying area, and compare it with the data model of the egg-laying area to obtain the location of the basic egg placement position 62.

[0123] The location information of the remaining egg placement positions 61 is calculated based on the distribution of the basic egg placement positions 62 and egg placement positions 61, and the distance information between the egg placement positions 61 on the egg tray 6.

[0124] When using real-time images or egg placement positions 61, if each egg placement position 61 on the egg tray 6 is identified by recognizing its outline, and since there are usually dozens of egg placement positions 61 on an egg tray 6, recognition errors are inevitable.

[0125] In this embodiment, the data model of the egg-collecting area and the egg tray 6 is obtained in advance using the same method as in Embodiment 3. Then, an egg placement position 61 on the egg tray 6, namely the basic egg placement position 62, is selected to calculate and identify the location information of the egg placement area where the egg placement position 61 is located. The distribution of the egg placement positions 61 on the egg tray 6 and the distance information between the egg placement positions 61 are known. After knowing the position of the basic egg placement position 62, the position information of the remaining egg placement positions 61 can be obtained by translating the horizontal and vertical directions by N times the distance between the egg placement positions 61. It is not necessary to identify the position information of the remaining egg placement positions 61 one by one through the contour, thereby improving the placement accuracy of the eggs and shortening the identification time of the egg placement positions 61.

[0126] like Figure 3 As shown, the basic egg placement position 62 is selected from one corner of the egg tray 6. After obtaining the basic egg placement position 62, the position information of the remaining egg placement positions 61 can be obtained by translating horizontally and vertically with this position as the origin.

[0127] Example 7

[0128] Reference Figure 3 Furthermore, selecting the real-time image contour of the egg placement position 61 also includes selecting the contour of the reference egg placement position 63;

[0129] The reference egg placement position 63 is the egg placement position 61 on the egg tray 6 other than the basic egg placement position 62;

[0130] And obtain information on the egg placement position 61 between the basic egg placement position 62 and the reference egg placement position 63;

[0131] By comparing the position information of the spaced egg placement positions 61 with the pre-acquired distribution and spacing of the egg placement positions 61 on the egg tray 6, the position information of the egg placement positions 61 is adjusted.

[0132] Egg tray 6 is placed on the egg placement area. If the horizontal placement angle of egg tray 6 is tilted from the expected placement angle when it is placed on the egg placement area, the tilt will have little impact on the egg placement position 61 which is closer to the basic egg placement area. However, if the distance to the basic egg placement position 62 is far away, there may be a deviation when placing the eggs, which may cause the eggs to not be placed properly into the egg placement position 61.

[0133] In this embodiment, in addition to identifying the basic egg placement position 62 on the egg tray 6, the identification of the reference egg placement position 63 is also added. The baseline placement position and the reference placement position, which are a certain distance apart on the real-time image, are compared with the distribution of the egg placement positions 61 on the egg tray 6 and the distance information between the egg placement positions 61 obtained in advance. This verifies whether the basic egg placement position 62 and the reference egg placement position 63 can correspond to the pre-obtained position. If the reference egg placement position 63 cannot correspond to the pre-obtained position, the coordinate point of the egg placement position 61 information is rotated and adjusted with the basic egg placement position 62 as the center so that the position of the reference egg placement position 63 can correspond. Then the position information is provided to the egg-retrieving robot 4.

[0134] like Figure 3 As shown, the reference egg placement position 63 and the basic egg placement position 62 are 5 egg placement positions 61 apart horizontally and 5 egg placement positions 61 apart vertically. If the reference image placement position obtained from the real-time image deviates from the pre-acquired position information during comparison, the position information of the egg placement position 61 to be output is adjusted so that the output position information corresponds to the position information of the real-time image.

[0135] Example 8

[0136] Reference Figure 2 Furthermore, it also includes a non-conforming poultry egg placement area 5, which is used to place non-conforming poultry eggs;

[0137] The controller controls the egg-retrieving robot 4 to deliver the substandard eggs to the substandard egg placement area 5.

[0138] Egg trays 6 in the egg collection area are used to hold qualified poultry eggs. Once the egg trays 6 are full, they are moved directly away, loaded onto trucks and transported to the market. However, unqualified poultry eggs, such as those with broken surfaces or those contaminated by feces and egg liquid from broken eggs, can easily accumulate in the egg collection area if they are not removed in time.

[0139] The designated non-compliant egg placement area 5 allows for the placement of non-compliant eggs. After identifying a non-compliant egg, the egg-retrieving robotic arm 4 can remove the egg and place it in the designated compliant egg placement area. This facilitates the subsequent unified processing of the eggs by staff and reduces interference from non-compliant eggs in the process of retrieving compliant eggs.

[0140] Example 9

[0141] The image detection unit also includes a poultry egg quality monitoring module, which acquires images of the surface of poultry eggs through real-time images and compares the color and texture of the poultry egg surface with those of qualified poultry eggs.

[0142] The controller provides information on substandard eggs to the egg-collecting robot 4 based on the comparison results from the egg quality monitoring module.

[0143] The camera assembly 31 also includes an egg bottom image acquisition camera 311, which acquires images of the bottom of the eggs and provides them to the egg quality monitoring module.

[0144] A qualified poultry egg has a smooth surface and a uniform color, while a cracked or contaminated poultry egg will have prominent lines and color areas that do not match the surface color. These poultry eggs are unqualified. After identifying these unqualified poultry eggs, their location information can be directly provided to the egg-retrieving robot 4.

[0145] Poultry eggs are placed on the egg collection area, and the camera of the image acquisition component is generally set above the egg collection area. When the poultry eggs are static, it is difficult to capture whether the side of the poultry egg that is in contact with the egg collection area is qualified. The poultry egg bottom image acquisition camera 311 provided in this embodiment can capture the image of the bottom of the poultry egg after the egg collection robot 4 grabs and lifts the poultry egg. The poultry egg quality inspection module can detect the bottom of the poultry egg through the poultry egg bottom image acquisition camera 311. If the bottom of the poultry egg is unqualified, it is placed in the unqualified poultry egg placement area 5.

[0146] Example 10

[0147] Reference Figure 10 Furthermore, an egg tray placement direction detection component is provided at the bottom of the egg placement area, and the egg tray 6 direction detection component includes two sets of position sensors 21;

[0148] The two sets of position sensors 21 are spaced 1 / 2 the distance between the egg placement positions 61 in both the horizontal and vertical directions.

[0149] To facilitate egg transportation, egg trays 6 filled with eggs are typically stacked. However, when stacked, the bottom protrusion of the egg placement slot 61 (corresponding to the small hole) of the upper egg tray 6 does not directly rest on the eggs in the lower egg placement slot 61, as this could easily crush the eggs. Instead, a support pillar protrudes between the two small holes on one side of the lower opening. This pillar contacts the bottom of the small holes, providing support to the upper egg tray 6 and effectively protecting the eggs when stacked. Figure 4 , Figure 11 As shown.

[0150] To facilitate manufacturing, the two adjacent sides of the egg tray 6 are extended by half the length of the egg placement area 61. When overlapping eggs, the upper egg tray 6 can be rotated 180° to overlap with the lower egg tray 6, eliminating the need to manufacture multiple models of egg tray 6. After overlapping, the boundaries of the upper and lower egg trays 6 can also completely coincide. Figure 4 , Figure 11 As shown.

[0151] When placing an empty egg tray 6 on the egg-laying area, the two egg trays 6 placed one in front of the other can be positioned 180° apart. When they are filled with eggs and stacked together, the bottom of the egg holes of the egg tray 6 can be placed in contact with the support column.

[0152] The bottom of the small hole in the egg tray 6 is in contact with the surface of the egg placement area. When the egg tray is placed on the position sensor 21, the position sensor 21 can detect it. If the position sensor 21 is blocked at different positions, it means that the egg tray 6 is placed at different angles. The placement direction of the tray can be detected by two sets of position sensors 21, so as to obtain the approximate placement position of the egg tray 6 in the egg retrieval area in advance. When obtaining the image outline of the location of the poultry egg placement area, the location information of the poultry egg placement area can be obtained faster and more accurately based on the approximate location.

[0153] In addition, the positions of the two egg trays 6 placed one after the other can also be verified by the two sets of position sensors 21. If the same position sensor 21 is blocked in both placements of the egg tray 6, an alarm can be issued to indicate that the tray was placed in the wrong direction the second time.

[0154] Example 11

[0155] Furthermore, the larger end of the egg is offset to 1 / 3-1 / 4 of the length of the larger end and the direction of the larger end of the egg beyond the vertical projection outline of the egg placement position 61.

[0156] In this embodiment, the length of the large end of the egg and the length of the head end are the same as the length of the egg's axis.

[0157] If the length of the larger end of the egg that deviates from the vertical projection outline of the egg placement position 61 is less than 1 / 4 of the length of the egg's axis, then the smaller end of the egg is likely to come into contact with the side wall of the egg placement position 61. When it falls, it is blocked by the side wall of the egg placement position 61. As a result, when the egg is placed, the smaller end of the egg cannot fall completely into the middle of the egg placement position 61, and the larger end at the top is also skewed.

[0158] If the length of the larger end of the egg that deviates from the vertical projection outline of the egg placement position 61 is greater than 1 / 3 of the length of the egg's axis, it will affect the speed at which the egg falls. In some cases, the larger end may be heavier, causing the egg to roll after landing on the egg tray 6 and fail to fall into the egg placement position 61.

[0159] The egg-retrieving robotic arm 4 places the egg when the bottom of the egg is 1-8 mm away from the outline of the egg placement position 61. This avoids the bottom of the egg from contacting the surface of the egg tray 6 during the egg-moving process, or the egg from falling too high, which could cause the egg to break.

[0160] Example 12

[0161] Reference Figure 12 Furthermore, the egg-retrieving robotic arm 4 also includes a robotic arm 41, a negative pressure suction cup 42, and a negative pressure component 43;

[0162] The negative pressure suction cup 42 adsorbs and grasps the poultry eggs, and the negative pressure component 43 provides adsorption power for the negative pressure suction cup 42.

[0163] The robotic arm 41 drives the negative pressure suction cup 42 to move.

[0164] If a mechanical claw is used to grasp poultry eggs, a more complex mechanical claw control structure is required. In addition, when the claw structure grasps poultry eggs, the claws come together to clamp the eggs. The orientation of the large and small ends of the poultry eggs is easily rotated when they are clamped, which makes the orientation information of the large end of the poultry eggs obtained before laying the eggs inaccurate, resulting in poor egg flipping effect.

[0165] The negative pressure suction cup 42 is used to grasp the poultry egg, directly adsorbing onto the upper surface of the egg without changing the orientation of the large and small ends of the egg. When the egg is placed down, the suction cup will not interfere with the egg tray 6. The negative pressure component 43 provides negative pressure, and the control structure and control steps are relatively simple.

[0166] Preferably, in this embodiment, the negative pressure component 43 can be a small negative pressure pump or a cylinder with a piston inside to adsorb and release the poultry eggs.

[0167] In the description of the embodiments of the present invention, it should be understood that the terms "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "center," "top," "bottom," "top," "bottom," "inner," "outer," "inner side," and "outer side," etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the accompanying drawings and are only for the purpose of describing the present invention and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. "Inner side" refers to the interior or enclosed area or space. "Outer perimeter" refers to the area surrounding a specific component or specific area.

[0168] In the description of this invention, unless otherwise stated, "a plurality of" means two or more.

[0169] In the description of the embodiments of the present invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installation," "connection," "joining," and "assembly" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in the present invention based on the specific circumstances.

[0170] In the description of embodiments of the present invention, specific features, structures, materials or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0171] In the description of the embodiments of the present invention, it should be understood that "-" and "~" represent a range of two numerical values, and this range includes the endpoints. For example, "AB" represents a range greater than or equal to A and less than or equal to B. "A~B" represents a range greater than or equal to A and less than or equal to B.

[0172] In the description of embodiments of the present invention, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, and B existing alone. Additionally, the character " / " in this document generally indicates that the preceding and following related objects have an "or" relationship.

[0173] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A visual recognition-based poultry egg plating system, characterized in that, include: Egg retrieval area, egg placement area, image detection unit, and egg retrieval robotic arm; The egg-collecting area provides a place for placing poultry eggs before they are picked up; The egg-laying area is used to place egg trays that hold poultry eggs; The image detection unit detects the eggs and provides the egg-retrieving robot with the egg-grabbing position, the egg-placement position, and the position of the large end of the egg. The egg-collecting robotic arm grabs the eggs from the egg-collecting area and places them one by one into the egg placement positions on the egg tray. The egg-retrieving robotic arm includes a controller that receives information from the image detection unit and controls the robotic arm to move between the egg-retrieving area and the egg-placing area, grasping and placing the eggs; and... The controller, based on the egg placement position provided by the image detection unit, shifts the larger end of the egg above the egg placement position to outside the vertical projection outline of the egg placement position, and then places the egg down. The egg's large end is offset from the vertical projection outline of the egg placement position by 1 / 3 to 1 / 4 of the length of the large end and the length in the direction of the large end; the egg-retrieving robot puts down the egg when the bottom of the egg is 1-8 mm away from the outline of the egg placement position.

2. The poultry egg plating system based on visual recognition according to claim 1, characterized in that: The image detection unit includes a camera component, an egg position detection module in the egg-collecting area, an egg end detection module, and an egg placement position detection module. The camera component is used to capture real-time images of the egg retrieval area and the egg placement area; The egg location detection module in the egg collection area detects the location of the eggs in the egg collection area based on real-time images and provides the location information of the eggs to the egg collection robot arm. The egg end detection module detects the large and small ends of the eggs in the egg collection area and provides the location of the large end of the egg to the egg collection robot. The egg placement detection module detects multiple egg placement positions on the egg tray and provides the egg placement information to the egg-retrieving robot.

3. The visual recognition-based poultry egg traying system according to claim 2, characterized in that: The egg location detection module in the egg collection area pre-acquires the data model of the egg collection area and obtains the outline of the egg collection area based on real-time images; Select a real-time image of one of the eggs in the egg collection area. Based on the location of the egg in the egg collection area in the real-time image, obtain the length ratio of the egg relative to the outline of the egg collection area. Compare this ratio with the data model of the egg collection area to obtain the location of the egg in the egg collection area.

4. The visual recognition-based poultry egg traying system according to claim 3, characterized in that: The egg end detection module performs end detection on the egg after the egg position detection module in the egg collection area selects the egg. During end detection, the outline of the egg is first obtained. The two points A1 and A2 that are farthest apart on the outline of the egg are selected, and a line is drawn with A1 and A2 as endpoints to obtain the central axis A1A2 of the egg. Draw dividing lines B1B2, which are perpendicular to the central axis A1A2, pass through the midpoint of the central axis A1A2, and intersect with two points B1 and B2 on the outline of the egg respectively. Then calculate the area of ​​the egg outlines on both sides of the dividing line B1B2. Record point A1 or A2 on the side with the larger area as the large end of the egg, and record point A2 or A1 on the side with the smaller area as the small end of the egg.

5. The visual recognition-based poultry egg traying system according to claim 4, characterized in that: The method of selecting the longest point through continuous radiation is used to select the farthest points A1 and A2; First, arbitrarily select a point I1 on the outline of the egg. Using point I1 as the origin, draw multiple straight lines radially within the outline of the egg. Select the longest straight line I1I2 within the outline of the egg. Draw multiple straight lines radiating outwards from point I2 within the outline of the egg, and select the longest straight line I2I3. Continue with the farthest point of radiation I n-1 Line I drawn with the origin n-1 I n Continue until the two points A1 and A2 that are furthest apart within the outline of the egg are selected.

6. The visual recognition-based poultry egg traying system according to claim 2, characterized in that: The egg placement detection module pre-acquires the distribution of egg placement positions on the egg tray, as well as the distance information between egg placement positions on the egg tray; The egg placement detection module pre-acquires the data model of the egg placement area and obtains the outline of the egg placement area based on real-time images; Select the real-time image outline of one of the egg placement positions as the basic egg placement position. Based on the location of the basic egg placement position in the egg-laying area, obtain the length ratio of the basic egg placement position relative to the outline of the egg-laying area. Compare this ratio with the data model of the egg-laying area to obtain the location of the basic egg placement position. The location information of the remaining egg placement positions is calculated by using the basic egg placement positions, the distribution of egg placement positions, and the distance information between egg placement positions on the egg tray.

7. The visual recognition-based poultry egg traying system according to claim 6, characterized in that: Selecting the real-time image contour of the egg placement location also includes selecting a reference egg placement location contour; The reference egg placement position is the egg placement position on the egg tray other than the basic egg placement position; And obtain the interval egg placement information between the basic egg placement position and the reference egg placement position; The location information of the egg placement positions is adjusted by comparing the other egg placement positions calculated with the reference egg placement positions.

8. The poultry egg plating system based on visual recognition according to claim 2, characterized in that: It also includes a designated area for storing substandard poultry eggs, which is used to store substandard poultry eggs. The controller directs the egg-retrieving robot to deliver substandard eggs to the substandard egg placement area.

9. The visual recognition-based poultry egg traying system according to claim 8, characterized in that: The image detection unit also includes a poultry egg quality monitoring module, which acquires images of the surface of poultry eggs through real-time images and compares the color and texture of the poultry egg surface with those of qualified poultry eggs. The controller provides information on substandard eggs to the egg-collecting robot based on the comparison results from the egg quality monitoring module. The camera assembly also includes a bottom image acquisition camera for eggs, which acquires images of the bottom of eggs and provides them to the egg quality monitoring module.

10. The poultry egg plating system based on visual recognition according to claim 1, characterized in that: The bottom of the egg placement area is equipped with an egg tray placement direction detection component, which includes two sets of position sensors. The two sets of position sensors are spaced 1 / 2 the distance between the egg placement positions, both horizontally and vertically.

11. The poultry egg tray-loading system based on visual recognition according to claim 1, characterized in that: The egg-retrieving robotic arm also includes a robotic arm, a negative pressure suction cup, and a negative pressure assembly; The negative pressure suction cup adsorbs and grasps the poultry eggs, and the negative pressure component provides adsorption power for the negative pressure suction cup; The robotic arm drives the negative pressure suction cup to move.

Images