Device and method for automatically measuring egg-laying performance of individual geese in small groups of families and cages

Abstract

The present invention discloses an automatic determination device and method for the egg-laying performance of individual geese in a small group of family cages. The device comprises an egg-laying detection device, a coding area device, a weighing area device, and a combined goose cage. Through the matched egg detection module, coding module, information storage module, information analysis module and corresponding mark detection algorithm, area division algorithm, target continuous counting algorithm, egg goose matching algorithm, event triggering algorithm, egg-laying detection, egg-laying performance detection, egg-laying collection and other related tasks are performed, which effectively solves the problems of waste of resource space, high cost of separate cage breeding facilities and management, and improves the degree of automation.

Description

Technical Field

[0001] The invention relates to the field of poultry breeding, and in particular to a device and method for automatically measuring the egg-laying performance of individual geese raised in cages in a small family group. Background Art

[0002] At present, there are two main modes of goose farming: caged and non-caged. Compared with the non-caged mode, caged goose farming is helpful for measuring the egg-laying performance of individual geese and improving the efficiency of goose breeding. In addition, breeders can better control the feed intake and feeding time of geese, which helps to ensure that the geese get proper nutrition and effectively manage feed costs; at the same time, breeders can more easily clean the goose house, improve the goose breeding environment, strengthen disease prevention and control, etc., so that high-yield geese can be selected efficiently and animal welfare can be improved at the same time.

[0003] There are two modes of goose cage farming: one-cage-one-farming and small-group cage farming for the family. The traditional one-cage-one-farming model has great advantages for breeding high-yield geese, but it wastes resources and space, and the costs of individual cage breeding facilities and management are high. In addition, geese are gregarious animals, and solitary confinement may limit their normal social behavior and affect their mental health and production performance.

[0004] With the development of goose farming industry, the scale of farms is getting higher and higher, but the degree of automation is low. Most farms collect goose eggs and measure the egg-laying performance of geese manually in the traditional way. Obviously, the traditional method can no longer meet the needs of large-scale farms to record egg-laying information and breed high-yield geese.

[0005] In order to solve the above problems, researchers have designed egg-laying detection methods and developed corresponding devices, but there are still some problems. The methods for measuring the egg-laying performance of individual breeding geese are different for different breeding modes. In the prior art, an RFID-based cage breeding poultry individual egg-laying recording system is proposed. The breeder scans the cage electronic tags of the collected eggs through a portable data acquisition device. This method requires additional technical training for the breeder and the inefficiency of manual inspections. In another prior art, an automated egg-laying box is designed for the egg-laying performance measurement of non-cage geese, but the cost of an egg-laying box is too expensive. In addition, the geese entering the egg-laying box will cause the door of the egg-laying box to close, which may cause the geese to have a stress response. Summary of the invention

[0006] In view of the above-mentioned deficiencies in the prior art, the present invention provides a device and method for automatically measuring the egg-laying performance of individual geese in a small group of cages.

[0007] In order to achieve the above-mentioned object of the invention, the technical solution adopted by the present invention is:

[0008] 1. A method for automatically measuring the egg-laying performance of individual geese in a small group of cages, characterized in that it comprises the following steps:

[0009] S1, using an egg-laying detection device to detect whether eggs are laid, and to locate information on newly laid goose eggs, specifically comprising the following steps:

[0010] S11, the host computer of the egg-laying detection module uses RTSP streaming and multi-threading technology to achieve real-time preview and streaming of multiple cameras;

[0011] S12, using a region division algorithm to obtain an accurate detection region of goose eggs;

[0012] S13, using the target continuous counting algorithm and the target detection algorithm to detect the egg-laying situation in the area and the identification information of the breeding geese in real time, wherein the target continuous counting algorithm is specifically as follows:

[0013] S131, treating the number of goose egg targets detected in multiple consecutive frames of images as a queue, where the queue elements represent the number of goose egg targets in each frame of image;

[0014] S132, when a new detection frame image arrives, the number of new goose egg targets is added to the end of the queue, and the first element in the queue is deleted;

[0015] S133, counting the element with the largest number of occurrences, and taking the element as the number of goose eggs in the queue length plus 1 time frame;

[0016] S14, using the egg-goose matching algorithm to match the newly appeared goose eggs, and save the coded egg code to the matching queue, wherein the specific method of information matching is:

[0017] S141, when the queue length in the area obtained by the target persistence counting method is equal to the number of goose eggs counted at the end of the time frame plus the number of goose eggs counted at the end of the time frame, it indicates that a new goose egg has appeared in the area, and the center coordinates of the new goose egg at this time are used to calculate the distance between the goose egg and other geese through the Euclidean distance;

[0018] S142, based on the distance between the new goose egg and other geese calculated in S141, starting from the first goose, checking whether the goose's lay-off time exceeds the set time threshold; if not, continue to judge the lay-off time of the next goose until a goose egg whose lay-off time exceeds the set time threshold is matched and the matching is completed;

[0019] S15, when the goose egg rolls down to the conveyor belt and passes through the event triggering algorithm, the first data in the matching queue is popped out of the queue and uploaded to the coding queue for storage;

[0020] S2. The conveyor belt will move forward at a constant speed of 1 cm / s and print the goose eggs. During the movement, if new goose eggs roll down, the queue will be sorted according to the distance priority to the printing sensor.

[0021] Furthermore, the region division algorithm is used in S12 to obtain the accurate detection area of ​​the goose eggs by using an infrared camera to obtain any frame of image in the goose cage, and the four vertex coordinates of the goose cage in the illuminated area constitute a quadrilateral to represent the monitoring area.

[0022] Furthermore, the S15 specifically includes the following steps:

[0023] S151, obtaining the center coordinates of the goose egg target frame in the detection area by the target detection algorithm, and counting the number of goose eggs in the current frame in the area by the target persistence technology algorithm;

[0024] S152, judging whether the number of goose eggs in the current frame is greater than the number of goose eggs counted at the last moment, if so, assigning the number of goose eggs in the current frame to the number of goose eggs counted at the last moment, and at the same time assigning the goose eggs to matching mother goose tags and saving them in a matching queue, and uploading them to a server queue;

[0025] S153, creating a child thread to detect whether the data in the server queue and the inkjet printing queue is increased. If there is new data, the data is uploaded to the priority inkjet printing queue of the front-end server and the inkjet printing module host computer.

[0026] A device for automatically measuring the egg-laying performance of individual geese in a small family group cage, comprising an egg-laying detection device, a coding area device, a weighing area device, and a combined goose cage, wherein:

[0027] The egg-laying detection device includes a plurality of infrared cameras and an egg-laying detection module host computer corresponding to the breeding cage, wherein the infrared camera is deployed on a hanging rail and communicates with the egg-laying detection module host computer through POE; the egg-laying detection module host computer uses RTSP streaming and multi-threading technology to realize real-time preview and streaming of multiple cameras;

[0028] The coding area device includes a coding machine, a coding machine electric induction switch and a coding module host computer. When the coding machine electric induction switch senses that a goose egg passes by, it communicates with the coding module host computer through RS-232 or RS-485. After the coding module host computer receives the trigger signal, it produces corresponding print data and sends the data to the coding machine through RS-232 or RS-485.

[0029] The combined goose cage comprises a plurality of goose cages of the same size, a feed trough, an automatic drinking water pipe, a branch egg conveyor belt and an egg collecting conveyor belt. The bottom plate of each goose cage forms a fixed angle with the egg collecting conveyor belt. When the goose lays eggs, the eggs automatically slide from the goose cage to the egg conveyor belt, and the egg conveyor belt moves forward at a fixed speed. Each goose cage houses a male goose and the same number of female geese, and each female goose wears a different collar for distinction.

[0030] The present invention has the following beneficial effects:

[0031] (1) Five female geese and one male goose are kept in one cage, and different collars are worn on the female geese to distinguish them. Compared with raising them in a single cage, this can greatly save the cost of making cages and the space they occupy. It also makes it easier for breeders to clean the goose house, improve the goose breeding environment, and strengthen disease prevention and control. This allows for efficient selection of high-yield geese and improves animal welfare.

[0032] (2) An improved deep learning network was used to further solve the problem of interference of environmental information in complex environments on the feature extraction of breeding geese and goose eggs and to improve the detection efficiency.

[0033] (3) The target continuous counting algorithm solves the misjudgment caused by recognition problems and greatly improves the accuracy of information.

[0034] (4) Through the egg-goose matching algorithm, the information matching problem of geese and goose eggs is automatically identified, which greatly reduces the need for manual observation of cameras to determine egg-laying conditions.

[0035] (5) The area division algorithm can prevent the complex environment outside the infrared camera detection area from affecting the detection effect, greatly reducing the impact of abnormal situations on the system.

[0036] (6) The accuracy of the coding data can be ensured by the event trigger line algorithm. Only when the goose egg rolls onto the conveyor belt within the area division and the center point passes through the event trigger line, will the matching information of the goose egg and the goose be uploaded to the priority coding queue in the coding module host computer.

[0037] (7) Using the distance priority coding queue can ensure that when the goose eggs at an uncertain position and time fall onto the conveyor belt during its movement, they are correctly inserted into the coding queue, thus ensuring the accuracy of the coding.

[0038] (8) The eggs are transported to the egg collection conveyor belt through the branch conveyor belt for unified coding, which greatly reduces the cost of the inkjet printer and improves the automation efficiency.

[0039] (9) Through the communication strategy of the egg-laying detection module, the coding module, the information storage module, and the information analysis module, it is possible to raise multiple geese in one cage, so that the entire detection process can be automated in a streamlined manner, greatly improving the determination of the egg-laying performance of individual breeding geese and reducing labor costs.

[0040] (10) The platform for the management of egg-laying data of individual geese in cages of small groups of family-bred geese includes waterfowl egg-laying record reports, group breeding goose egg-laying record reports, user feedback reports, warning goose egg-laying performance reports, alarm equipment reports, and operation management functions. It greatly facilitates the management personnel to select high-yield and low-yield geese and deal with early warnings in a timely manner, and at the same time facilitates the unified management of farm resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 It is a system module diagram of the present invention.

[0042] Figure 2 It is a three-dimensional schematic diagram of the complete structure of the present invention.

[0043] Figure 3 It is a schematic diagram of the coding load-bearing collection platform of the present invention.

[0044] Figure 4 It is a three-dimensional schematic diagram of a single cage structure of the present invention.

[0045] Figure 5 It is a target detection deep learning model designed by the present invention based on the characteristics of goose and goose egg data.

[0046] Figure 6 It is a schematic diagram of the region partitioning algorithm of the present invention.

[0047] Figure 7 It is the principle diagram of the target continuous counting algorithm of the present invention.

[0048] Figure 8 The present invention is a flow chart of a method for measuring the egg-laying performance of individual caged breeding geese. DETAILED DESCRIPTION

[0049] The specific implementation modes of the present invention are described below so that those skilled in the art can understand the present invention. However, it should be clear that the present invention is not limited to the scope of the specific implementation modes. For those of ordinary skill in the art, as long as various changes are within the spirit and scope of the present invention as defined and determined by the attached claims, these changes are obvious, and all inventions and creations utilizing the concept of the present invention are protected.

[0050] like Figure 1 As shown, the device for measuring the egg-laying performance of individual geese raised in cages in a small group of families is composed of an egg-laying detection device, a coding area device, a weighing area device, and a combined goose cage.

[0051] Egg laying detection device:

[0052] The egg-laying detection device consists of infrared cameras corresponding to the number of breeding cages and an egg-laying detection module host computer corresponding to the row of breeding cages. The infrared camera is deployed on the hanging rail and communicates with the egg-laying detection module host computer through POE. The infrared camera is 2.2m away from the breeding cage, and the overhead shooting range just includes one breeding cage; the egg-laying detection module host computer uses RTSP streaming and multi-threading technology to realize real-time preview and streaming of multiple cameras.

[0053] Coding area device:

[0054] The inkjet printer area device consists of an inkjet printer, an electro-sensing switch of the inkjet printer, and an upper computer of the inkjet printer module. When the electro-sensing switch of the inkjet printer senses that goose eggs are passing by, it communicates with the upper computer of the inkjet printer module through RS-232 or RS-485. After receiving the trigger signal, the upper computer of the inkjet printer module generates the corresponding print data and sends the data to the inkjet printer through RS-232 or RS-485.

[0055] Combined goose cage:

[0056] The combined goose cage is composed of multiple goose cages of the same size, feed troughs, automatic drinking water pipes, split egg conveyor belts and egg collection conveyor belts. Each goose cage is 2.1 meters long, 1 meter wide and 0.9 meters high. The bottom plate forms an angle of 5 degrees with the egg conveyor belt. When the goose lays eggs, the eggs automatically slide from the goose cage to the egg conveyor belt, which moves at a constant speed of 1 cm / s. Each goose cage houses five female geese and one male goose. In order to distinguish the five female geese, the female geese are given collars of different styles, while the male geese do not wear collars.

[0057] like Figure 1 Shown is a system module diagram of the present invention, where the caged goose individual egg-laying performance measurement system consists of an egg-laying detection module, a coding module, an information storage module, and an information analysis module.

[0058] Egg-laying detection module: The main technical points include target detection algorithm, area division algorithm, target continuous counting algorithm, egg-laying goose matching algorithm, and event triggering algorithm;

[0059] Target detection algorithm: The identity of the goose in the cage is distinguished based on the different shapes of the collars they wear, so accurate recognition of goose collars is particularly important in complex environments. At the same time, an egg-laying module host computer needs to detect multiple pictures at the same time, so the inference speed of the model needs to be improved. For this purpose, an improved YOLOv8 network YOLOv8-CEG is designed.

[0060] Region division algorithm: Since an infrared camera detects the egg-laying situation of a breeding cage, but the illumination angle of each camera is not necessarily the same, in order to ensure the accurate drawing of the detection area and to prevent the complex environment outside the infrared camera detection area from affecting the detection effect, the region division technology is adopted. The specific description is as follows: Get a frame of image, and use the quadrilateral formed by the four vertices of the cage in the illumination area to represent the detection area (Detection Area, DA) as follows Figure 5 As shown, in the rectangular coordinate system, it is known that , , , With the coordinates of the two points, we can use formulas (1) and (2) to derive the equation of the line passing through the two points. By obtaining four equations of the line, we can get a DA.

[0061]

[0062] Specifically, each DA contains parameters such as area (ID), number of goose eggs in the current frame (Current_Egg), maximum number of goose eggs (Max_EggCount), egg goose matching data queue (Mate_Queue), queue for storing server data (Server_Queue), queue for storing inkjet data (Printer_Queue), and event trigger line for determining whether goose eggs have rolled onto the conveyor belt.

[0063] More specifically, the data structure stored in Mate_Queue is the egg code (Egg_Code), which is used to store the Egg_Code that is successfully matched in DA, and when the center point of the goose egg rolls down to the conveyor belt, the data pops up and is uploaded to Printer_Queue after passing the event trigger line; the data structure stored in Server_Queue is the egg-laying goose number, the cage number, the egg-laying time, the egg code number, and the egg-laying moment picture; the data structure stored in Printer_Queue is the egg code, the time when the egg rolls down to the conveyor belt, and the distance between the egg and the inkjet printer. Server_Queue and Printer_Queue are used for communication between threads. When the upload thread detects data in the two queues, it uploads the data to the front-end server and the upper computer of the inkjet module to ensure the efficient execution of the main thread.

[0064] Target persistence counting algorithm: During real-time detection, due to the system error of the target detection algorithm itself, the motion blur caused by the camera or target movement will affect the judgment signal of the goose egg laying information. In order to reduce the jitter caused by the system error, the target persistence counting algorithm is used to eliminate instantaneous false detection or interference and increase the accuracy of the judgment signal.

[0065] Specifically, the principle of the target persistence algorithm is as follows Figure 6 As shown in the figure, the number of targets detected in N consecutive frames is regarded as a queue with a queue length of N. The queue elements represent the number of targets in each frame. With the arrival of new frames, the new number of targets is added to the end of the queue, and the first element in the queue is deleted to keep the queue length unchanged. Then, we count the data with the most occurrences in the queue, that is, the value with the largest number of targets. We use this value as the number of goose eggs (Current_Egg) in the frame at time N+1. In order to prevent N from being too large, causing delays in goose egg counting, and thus affecting the determination of the egg-laying performance of breeding geese, it is estimated that N is selected as 4, and the input frequency of the video is 2 frames / s, that is, Current_Egg is counted once at the end of 2 seconds.

[0066] Egg goose matching algorithm: Since five female geese and one male goose are raised in a small group in a cage, the target detection algorithm can obtain the center point coordinates of the five female geese ( )(i=1,2,3,4,5), when Current_Egg>Max_Eggcount in the area is obtained through the target continuous counting algorithm, it means that a new goose egg appears in DA. At this time, the center point coordinates (x, y) of the new goose egg are calculated through the Euclidean distance (3) to get the distance between the goose egg and other geese (i=1,2,3,4,5).

[0067] (3)

[0068] Specifically, get the distance between the goose eggs and other geese , then from Start to detect whether the goose's lay time exceeds the set time threshold (T). If it does not exceed the time threshold (T), continue to determine the next goose's lay time. This is mainly to ensure that the goose will not lay eggs continuously in a short period of time, and further ensure the correctness of the egg-geese matching.

[0069] Event trigger algorithm: In order to prevent the goose eggs from not falling onto the conveyor belt in time after the goose eggs are laid, and the conveyor belt moves at a constant speed of 1cm / s, if the goose egg laying data is uploaded to the priority coding queue as soon as the goose eggs are detected, it will inevitably lead to coding errors. Therefore, an event trigger line is set at the exit of each egg-laying cage. In the rectangular coordinate system, it is known that , , , The coordinates of the two points can be obtained according to formulas (1) and (2), as follows: Figure 4As shown by the dotted line; when the goose eggs roll onto the conveyor belt and the center point passes through the event trigger line, the breeding goose egg laying data in Mate_Queue is uploaded to Printer_Queue, and the upload thread uploads the data to the upper computer of the inkjet module.

[0070] Specifically, the target detection algorithm first obtains the center coordinates (x, y) of the goose egg target frame in DA, and the target persistence counting algorithm counts the number of goose eggs in the current frame in DA (Current_Egg), and detects whether Current_Egg is greater than Max_EggCount. If it is greater, Current_Egg is assigned to Max_EggCount. At the same time, the egg-goose matching algorithm is used to obtain the mother goose tag to which the goose egg belongs and save it in Mate_Queue, and the data to be uploaded to the server is saved in Server_Queue. Finally, when the goose egg rolls down to the conveyor belt, the center point passes through the event trigger line, and the Mate_Queue data is popped out and uploaded to Printer_Queue. In order to ensure the smooth progress of the main thread, a child thread is created to detect whether the data in Server_Queue and Printer_Queue is increased in real time. If there is new data, the data is uploaded to the priority coding queue of the front-end server and the coding module host computer.

[0071] Specifically, the reason for setting Current_Egg and Max_EggCount is that after the goose lays an egg, it may be blocked by the goose, causing Current_Egg = 0, rather than the egg rolling onto the conveyor belt causing Current_Egg = 0. Therefore, Max_EggCount is set to indicate the maximum number of goose eggs in the current DA. Only when the goose egg rolls onto the conveyor belt and the center point passes through the ETL, will Max_EggCount be reduced by 1.

[0072] Specifically, the priority inkjet queue is used to ensure the correctness of the goose egg inkjet information. The conveyor belt will move forward at a constant speed of 1cm / s. During the movement, if a new goose egg rolls down, the queue will be sorted according to the distance priority to the inkjet sensor. In each time step, the goose egg nodes in the priority inkjet queue are traversed and their distances are updated according to the movement of the conveyor belt. The conveyor belt moves a unit distance in each time step. For each goose egg node, its distance is subtracted by one unit distance. At the same time, in each time step, check whether the distance of the first node in the priority queue to the inkjet printer is equal to zero, indicating that the goose egg has reached the position of the inkjet printer, then pop up the goose egg information from the queue and prepare for printing.

[0073] The platform for the management of the egg-laying data of individual geese in cages of small groups of families. It includes waterfowl egg-laying record reports, group breeding goose egg-laying record reports, user feedback reports, early warning goose egg-laying performance reports, alarm equipment reports, and operation management functions.

[0074] like Figure 2 The figure shows a complete three-dimensional schematic diagram of the structure of the present invention. The infrared camera used in the camera ensures that the goose collar and egg-laying situation can be accurately seen in the night environment, and the goose's rest at night is not affected. An infrared camera shoots two cages from above, and the height of the camera and the cage is set to 2.2m. Each goose cage is 2.1 meters long, 1 meter wide, and 0.9m high. The bottom plate forms a certain angle with the egg conveyor belt. When the goose lays eggs, the goose eggs automatically slide onto the egg conveyor belt. Five female geese and one male goose are kept in each goose cage. In order to distinguish the five female geese, the female geese are equipped with collars of different styles, while the male geese do not wear collars. In order to increase the accuracy of recognition, the five labels are marked with numbers 1, 2, 3, 4, and 5 respectively. More visual features are provided to improve the deep learning model to better learn and distinguish the identity characteristics of each female goose, thereby improving the accuracy of recognition.

[0075] The specific process is as follows:

[0076] S11: The host computer of the egg-laying detection module uses RTSP streaming and multi-threading technology to achieve real-time preview and streaming of multiple cameras;

[0077] S12: Use the region division algorithm to obtain an accurate detection area;

[0078] S13: Using the target continuous counting algorithm and the target detection algorithm to detect the egg-laying situation in the area and the identification information of the breeding geese in real time;

[0079] S14: Use the goose egg matching algorithm to match the newly emerged goose eggs and save the egg coding to Mate_Queue;

[0080] S15: When the goose egg rolls down into the path of the conveyor belt and passes through the event triggering algorithm, the first data in the Mate_Queue queue is popped out of the queue and uploaded to the Printer_Queue queue for storage.

[0081] Since the identity of the goose in the cage is distinguished by the different shapes, black and white colors, and digital labels of the collars, it is particularly important to accurately identify the goose collars in complex environments. At the same time, an egg-laying module host computer needs to detect multiple images at the same time and upload the detection data, so the inference speed of the model needs to be improved. For this purpose, an improved YOLOv8 network YOLOv8-CEG is designed. Figure 3As shown in the figure, several advanced convolutional network modules (ConvNeXtV2, EMA, GSConv modules) are combined to enhance the accuracy of collar segmentation in complex environments of YOLOv8. In the backbone network, ConvNeXt V2 is used to replace the C2F module in the backbone. Compared with the original C2F module in the original YOLOv8, ConvNeXtV2 shows excellent feature extraction and recognition capabilities with a comparable number of model parameters. In addition, the lightweight nature of ConvNeXt V2 makes it suitable for resource-limited settings. The EMA module is introduced in the backbone to increase the extraction of contextual information and improve the network's ability to recognize multi-scale and multi-shape targets. GSConv is used on the Neck side instead of ordinary convolution operations to reduce the computational complexity and inference time of the model while maintaining its accuracy.

[0082] Using YOLOv8-CEG, Goose and Egg can be accurately identified and different features on the collar can be successfully segmented. The algorithm can distinguish the target object from the complex background that may be highly similar or geometrically overlapped. Incorporating contextual information into the image is used to distinguish the morphological features of different collars and improve the overall performance of the algorithm.

[0083] The model training and evaluation method includes the following steps:

[0084] Photos of geese and goose eggs wearing different collars in cages were collected, and the Labelme annotation tool was used to annotate the images (the rectangular labels of geese and goose eggs were Goose and egg, the male goose was labeled Goose0, and the five female geese were labeled Goose1, Goose20, Goose3, Goose4, and Goose5 respectively). The dataset was divided into training set, test set, and validation set according to the set ratio of 7:2:1; the image sizes of the training set, test set, and validation set were all adjusted to 1440×1440, and YOLO-CEG was used for training, with 200 training rounds; after each round of training, the validation set was used to verify the best network training weights; the test set was used to judge whether the model has good generalization ability.

[0085] Coding information

[0086] The conveyor belt will move forward at a constant speed of 1 cm / s. During the movement, if new goose eggs roll down, they will be sorted in the queue according to the distance priority to the inkjet sensor.

[0087] The specific process is as follows:

[0088] S21: In each time step, the data in Printer_Queue is obtained and saved in the priority inkjet printing queue;

[0089] S22: In each time step, traverse the goose egg nodes in the priority coding queue and update their distances according to the movement of the conveyor belt;

[0090] S23: When the electro-sensing switch of the inkjet printer detects the presence of goose eggs, it communicates with the upper computer of the inkjet printer module through RS-232 or RS-485. After receiving the trigger signal, the upper computer of the inkjet printer module produces the corresponding printing data and sends the data to the inkjet printer through RS-232 or RS-485.

[0091] S24: The inkjet printer prints the goose identity information onto the goose eggs, transmits a printing completion signal to the inkjet module host computer, and updates the priority queue information in the inkjet module host computer.

[0092] The platform for the management of the egg-laying data of individual geese in cages of small groups of families. It includes waterfowl egg-laying record reports, group breeding goose egg-laying record reports, user feedback reports, early warning goose egg-laying performance reports, alarm equipment reports, and operation management functions.

[0093] Waterfowl egg-laying record report: including goose code, egg-laying performance, cage name, number of eggs laid today, cumulative number of eggs laid, time of this egg-laying, time interval between egg-laying stops, whether it is in the egg-laying period, egg-spraying code, and pictures of the egg-laying moment.

[0094] Group breeding goose egg-laying cycle report: including group egg-laying time distribution diagram, group cumulative egg-laying number, group average egg-laying number.

[0095] User feedback report: Users can use the mini program to upload goose illness information, equipment damage information, goose egg laying information, and manually correct the correct ownership of eggs.

[0096] Early warning report on goose egg-laying performance: It determines whether the number of eggs laid by breeding geese is far from the average number of eggs laid by the group. If it is far from the average number of eggs laid by the group, an early warning report will be generated. The early warning report includes the breeding goose code, egg-laying performance, cage number, egg-laying number, and egg-laying time distribution diagram.

[0097] Equipment early warning report: including automatic alarms for hardware devices such as infrared camera disconnection and host damage.

[0098] Operation management: including threshold management, farm management, equipment management, and waterfowl archives, which are provided to administrators to facilitate unified management of farm resources; threshold management mainly involves managers setting time thresholds (T) based on actual goose production suspension time;

[0099] Device for measuring egg-laying performance of individual breeding geese in cages

[0100] The whole device consists of the following structure. Figure 5The figure is a schematic diagram of the specific implementation structure of the device of the present invention. In the figure, the device consists of 1, a multifunctional cabinet; 2, an infrared camera; 3, an egg conveyor belt; 4, an egg collecting conveyor belt; 5, a PLC conveyor belt controller; 6, an egg picking machine tail seat; 7, an egg collecting plate; 8, a weighing table; 9, an inkjet printer; 10, a breeding cage; 11, a feed trough; 12, a goose collar; and 13, a 5° inclined egg sliding bottom plate.

[0101] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0102] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0103] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0104] The present invention uses specific embodiments to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

[0105] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the principles of the present invention, and should be understood that the protection scope of the present invention is not limited to such specific statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. A method for automatically determining the egg-laying performance of individual geese in a small group of cages, characterized in that: The steps include: S1, using an egg-laying detection device to detect whether eggs are laid, and to locate information on newly laid goose eggs, specifically comprising the following steps: S11, the host computer of the egg-laying detection module uses RTSP streaming and multi-threading technology to achieve real-time preview and streaming of multiple cameras; S12, using a region division algorithm to obtain an accurate detection region of goose eggs; S13, using the target continuous counting algorithm and the target detection algorithm to detect the egg-laying situation in the area in real time and to identify the identity information of the breeding geese, wherein the specific method of the target continuous counting algorithm is: S131, treating the number of goose egg targets detected in multiple consecutive frames of images as a queue, where the queue elements represent the number of goose egg targets in each frame of image; S132, when a new detection frame image arrives, the number of new goose egg targets is added to the end of the queue, and the first element in the queue is deleted; S133, counting the element with the largest number of occurrences, and taking the element as the number of goose eggs in the queue length plus 1 time frame; S14, using the egg-goose matching algorithm to match the newly appeared goose eggs, and save the coded egg code to the matching queue, wherein the specific method of information matching is: S141, when the number of goose eggs in the time frame of the queue length plus 1 in the area obtained by the target persistence counting algorithm is greater than the maximum number of goose eggs, it indicates that a new goose egg has appeared in the area, and the center coordinates of the new goose egg at this time are used to calculate the distance between the new goose egg and other geese through the Euclidean distance; S142, based on the distance between the new goose egg and other geese calculated in S141, starting from the first goose, checking whether the goose's lay-off time exceeds the set time threshold; if not, continue to judge the lay-off time of the next goose until a goose whose lay-off time exceeds the set time threshold is matched and the matching is completed; S15, when the goose egg rolls down the conveyor belt and passes through the event triggering algorithm, the first data in the matching queue is popped out of the queue and uploaded to the coding queue for storage, including the following steps: S151, obtaining the center coordinates of the goose egg target frame in the detection area by the target detection algorithm, and counting the number of goose eggs in the current frame in the area by the target persistence counting algorithm; S152, judging whether the number of goose eggs in the current frame is greater than the number of goose eggs counted at the last moment, if so, assigning the number of goose eggs in the current frame to the number of goose eggs counted at the last moment, and at the same time assigning the goose eggs to matching mother goose tags and saving them in a matching queue, and uploading them to a server queue; S153, creating a sub-thread to detect the server queue and the inkjet printing queue in real time and determine whether the data has increased. If there is an increase, the data is uploaded to the priority inkjet printing queue of the front-end server and the inkjet printing module host computer; S2. The conveyor belt will move forward at a constant speed of 1 cm / s and print the goose eggs. During the movement, if new goose eggs roll down, the queue will be sorted according to the distance priority to the printing sensor.

2. The method for automatically determining the egg-laying performance of individual geese in a small group of cages according to claim 1 is characterized in that: The method of using the area division algorithm in S12 to obtain the accurate detection area of ​​the goose eggs is: using an infrared camera to obtain any frame of image in the goose cage, and using the coordinates of the four vertices of the goose cage in the illuminated area to form a quadrilateral to represent the monitoring area.

3. A device for automatically measuring the egg-laying performance of individual geese in a small group of cages based on any one of claims 1-2, characterized in that: It includes egg laying detection device, coding area device, weighing area device and combined goose cage, including: The egg-laying detection device includes a plurality of infrared cameras and an egg-laying detection module host computer corresponding to the goose cage, wherein the infrared camera is deployed on a hanging rail and communicates with the egg-laying detection module host computer through POE; the egg-laying detection module host computer uses RTSP streaming and multi-threading technology to realize real-time preview and streaming of multiple cameras; The coding area device includes a coding machine, a coding machine electric induction switch and a coding module host computer. When the coding machine electric induction switch senses that a goose egg passes by, it communicates with the coding module host computer through RS-232 or RS-485. After the coding module host computer receives the trigger signal, it produces corresponding print data and sends the data to the coding machine through RS-232 or RS-485. The combined goose cage comprises a plurality of goose cages of the same size, a feed trough, an automatic drinking water pipe, a branch egg conveyor belt and an egg collecting conveyor belt. The bottom plate of each goose cage forms a fixed angle with the branch egg conveyor belt. When the goose lays eggs, the eggs automatically slide from the goose cage to the branch egg conveyor belt. The branch egg conveyor belt moves forward at a fixed speed. A male goose and the same number of female geese are kept in each goose cage. Each female goose wears a different collar for distinction.

Images