An intelligent method and system for detecting and cleaning contaminants from caged waterfowl

Abstract

The present invention discloses an intelligent sewage detection and feather cleaning method and system for caged waterfowl, which relates to the field of waterfowl breeding. The method includes: moving a feeding cart with a core processing module, a high-definition camera and a nozzle, and obtaining the two-dimensional code label and waterfowl data in the breeding cage through the high-definition camera; removing the breeding cage net in the image based on a defense algorithm through the core processing module to restore the occluded part; using an improved MDA-DepLabV3+ model to detect the pollution degree of the two-dimensional code label and the waterfowl feathers, and obtaining the proportion of the stain area by segmenting the waterfowl area and the stain area; comparing the proportion of the stain area with a set threshold through the core processing module to obtain a cleaning instruction, and starting the telescopic rod to move the nozzle to the top of the breeding cage according to the cleaning instruction; using the nozzle to execute the corresponding cleaning instruction. The present invention solves the problems that the existing methods are prone to cause stress to waterfowl, lack intelligent recognition algorithms to detect the pollution degree of waterfowl feathers, and lack intelligent regulation.

Description

Technical Field

[0001] The present invention relates to the field of waterfowl breeding, and particularly to an intelligent sewage detection and feather cleaning method and system for caged waterfowl. Background Art

[0002] At present, there are mainly two modes of waterfowl breeding: caged and uncaged. Compared with the uncaged mode, the caged mode of waterfowl can help breeders better control the feed intake and feeding time of waterfowl, thus ensuring that waterfowl obtain appropriate nutrition and effectively managing feed costs. In addition, the caged mode is convenient for cleaning the farm building, helps to improve the breeding environment, avoid water resource pollution, strengthen disease prevention and control, and improve animal welfare. In recent years, with the increasing intensity of basic farmland protection, traditional breeding industries have been affected. Many farmers' traditional breeding modes, especially the ground flat breeding mode with a large area and low land use efficiency, have gradually shifted to the stacked caged breeding mode with higher density and land use rate.

[0003] The existing methods for cleaning the wings of waterfowl in large-scale flat breeding mainly rely on manual labor or simple spraying. Regular spraying of water in traditional waterfowl breeding is labor-intensive, may also pollute natural water resources such as rivers, and requires regular and timed manual operations. Moreover, it is difficult to adapt to different specifications of welfare caged breeding modes for waterfowl. With the development of the waterfowl breeding industry, the scale of farms is getting larger and larger, but the degree of automation is relatively low. Therefore, it is particularly important to improve the breeding environment and animal welfare while increasing breeding efficiency.

[0004] Waterfowl are naturally fond of water. Although the traditional caged breeding mode has many advantages in feeding management, it brings challenges to the cleaning of waterfowl feathers. In the caged breeding mode, due to space limitations and the special nature of waterfowl feathers, existing cleaning methods are difficult to meet the cleaning needs of waterfowl of different specifications, thus affecting the welfare and health of waterfowl. In particular, traditional feather cleaning methods mostly rely on manual labor or simple spraying, which are not only inefficient but also prone to water resource waste and pollution. Therefore, how to improve the feather cleaning effect in the caged breeding mode without affecting the welfare of waterfowl has become an urgent problem to be solved.

[0005] To solve the above problems, at present, some researchers have designed automatic feather cleaning methods and developed corresponding devices, but there are still some problems. The methods for cleaning waterfowl feathers are different for different breeding modes (flat breeding or caged breeding).

[0006] CN209234668U discloses an automatic shower for free-range waterfowl, which can allow waterfowl to take a shower freely and automatically turn off after leaving, thereby improving the ability of waterfowl to adapt to dry farming environment and improving production performance such as down quality. However, this method requires the breeder to drive the waterfowl into the shower room, which is likely to cause stress reactions in the waterfowl. CN112655592A discloses a cage farming system for laying ducks with non-surface combination intermittent spraying, which can achieve regular feather cleaning of waterfowl under the cage farming mode, but lacks an intelligent recognition algorithm to detect whether the feathers of waterfowl need to be cleaned, and also lacks an intelligent control system to dynamically adjust the atomization intensity, time and water temperature according to the individual state of waterfowl. Summary of the Invention

[0007] In view of the above deficiencies in the prior art, a cage farming waterfowl intelligent pollution detection and feather cleaning method and system provided by the present invention can dynamically adjust the atomization intensity, spraying time and water temperature according to the individual state of waterfowl, minimize the waste of water resources and pollution emissions during the cleaning process, meet the physiological needs of waterfowl at the same time, ensure the stability of its production performance and reproductive performance, comprehensively improve animal welfare, and achieve the efficient utilization of resources.

[0008] In order to achieve the above invention purpose, the technical solution adopted by the present invention is: a cage farming waterfowl intelligent pollution detection and feather cleaning method, including the following steps:

[0009] S1: Move the feeding cart with a core processing module, a high-definition camera and a nozzle, and obtain the two-dimensional code label and waterfowl data in the breeding cage through the high-definition camera;

[0010] S2: Remove the breeding cage net in the image based on the defense algorithm through the core processing module, and restore the blocked part;

[0011] S3: Based on the restored image, use the improved MDA-DepLabV3+ model to detect the two-dimensional code label and the degree of waterfowl feather pollution, and obtain the proportion of the stain area by segmenting the waterfowl area and the stain area;

[0012] S4: Obtain the cleaning instruction by comparing the proportion of the stain area with the set threshold through the core processing module, and start the telescopic rod to move the nozzle to the top of the breeding cage according to the cleaning instruction;

[0013] S5: Execute the corresponding cleaning instruction by using the nozzle, and real-time monitor the relevant data through the flow sensor and temperature sensor equipped at the nozzle, and feedback the relevant data to the core processing module;

[0014] S6: Judge whether the feeding cart has moved to the end of the breeding cage. If so, enter step S7; otherwise, return to step S1 to continue moving the feeding cart;

[0015] S7: Start the manure water conveyor belt to remove waterfowl feces, and use the manure water separation device to dehydrate the collected wet manure and then transfer it to the fermentation tank for fermentation treatment to convert it into organic fertilizer;

[0016] S8: Use the intelligent detection of waterfowl feather stains and the adaptive feather cleaning management platform to uniformly manage and display the breeding farm.

[0017] Furthermore, in the S1, the core processing module adopts a parallel scheduling mechanism of multi-threading and thread pool as the control strategy;

[0018] The parallel scheduling mechanism of the multi-threading and thread pool is specifically as follows: The core processing module assigns a control thread to each high-definition camera and the corresponding nozzle control unit, including the camera thread and the nozzle control thread, and the camera thread and the nozzle control thread are uniformly managed by the thread pool built in the core processing module.

[0019] Furthermore, the defense algorithm in the S2 performs the following operations:

[0020] S21: Perform semantic segmentation on the original image to obtain a semantic map through the Unet network. The formula is:

[0021]

[0022] where, is the obtained semantic map, is the Unet network, is the original image;

[0023] S22: Perform bitwise operation on the original image according to the semantic map. Through mask operation, the part marked as the breeding cage net area in the semantic map is removed from the original image to obtain the image after removing the breeding cage net. The formula is:

[0024]

[0025] where, is the image after removing the breeding cage net, and are pixel points;

[0026] S23: Based on the image after removing the breeding cage net, use the Pix2pixHD network to restore the image after removing the breeding cage net to a real image and restore the occluded part. The formula is:

[0027]

[0028] where, is the restored image, is the Pix2pixHD network.

[0029] Furthermore, the step S3 includes the following sub-steps:

[0030] S31: Based on the restored image, use the improved MDA-DepLabV3+ model to identify the QR code label, and use the Pyzbar library to decode it to extract the aquaculture cage number;

[0031] S32: Use the improved MDA-DepLabV3+ model to segment the waterfowl area and the stain area, and obtain the proportion of the stain area. The formula is:

[0032]

[0033] where, is the proportion of the stain area, is the area of the stain area, is the total area of the waterfowl area.

[0034] Furthermore, the improved MDA-DepLabV3+ model uses depthwise separable convolution DSConv to replace the traditional convolution Conv, uses the MobileNetV2 network to replace the original backbone network, and introduces the dynamic atrous spatial pyramid pooling ASPP module.

[0035] Furthermore, the intelligent detection of waterfowl feather stains and the adaptive feather cleaning management platform in step S8 includes:

[0036] Waterfowl feather stain cleaning report: The waterfowl feather stain cleaning report is used to record the cleaning operation information of waterfowl feathers each time, including waterfowl code, cleaning time, pollution degree, aquaculture cage number where it is located, cleaning water flow rate, spraying time, spraying water temperature, water consumption, energy consumption, cleaning effect evaluation, and relevant pictures generated during the cleaning process;

[0037] Group waterfowl cleaning performance report: The group waterfowl cleaning performance report is used to display the cleaning effect of the waterfowl group, including cleaning time distribution, average water flow rate, cleaning time, and group water resource usage, and provides trend analysis of cleaning efficiency and effect to help optimize the group cleaning plan;

[0038] User feedback report: The user feedback report is used to upload feedback information through a small program, including information on incomplete feather cleaning and equipment failure information. The platform automatically generates a feedback report to help the administrator handle problems and improve the system service quality;

[0039] Equipment alarm report: The equipment alarm report is used to monitor the equipment status in real time, automatically generate an equipment failure alarm report, and remind the administrator to repair it in time, including equipment type, fault description, and occurrence time;

[0040] Operation Management Module: The operation management module is used to provide operation management functions, including threshold management, equipment management, house management, and waterfowl file management;

[0041] Report and Visualization Display Platform: The report and visualization display platform has the function of generating various reports, supports detailed statistics and analysis of cleaned data, equipment status, and warning information, and provides various visualization display methods.

[0042] The technical solution adopted by the present invention is also: A system for an intelligent sewage detection and feather cleaning method for caged waterfowl. The system includes multiple breeding cages with the same structure, and the multiple breeding cages are connected in parallel. A drinking water pipe, a QR code label, a feeding trough, an egg collection trough, a fecal water conveyor belt, and rollers are arranged on the breeding cages. The drinking water pipe is located inside the breeding cage, the feeding trough is located outside the breeding cage, the QR code label is pasted on the feeding trough, the egg collection trough is located below the feeding trough, the fecal water conveyor belt is located below each breeding cage, and the rollers are located at the tail of the fecal water conveyor belt;

[0043] The system also includes a feed truck. The feed truck is located directly in front of the feeding troughs of the multi-layer breeding cages. A core processing module is mounted in the middle of the feed truck. The core processing module is connected to multiple high-definition cameras and multiple nozzles. The nozzles are connected to the feed truck through telescopic rods.

[0044] Further, the telescopic rod is driven by an electric push rod or a pneumatic cylinder. A water quality filtering unit is integrated inside the nozzle, including a replaceable filter element and a water pressure stabilizer.

[0045] Further, both sides of the fecal water conveyor belt are rolled up to form a sealed enclosure structure. The fecal water conveyor belt is transported by a motor drive. A manure scraping plate and an adjustable roller are installed at the end of the fecal water conveyor belt, which can press out the excess water in the wet manure.

[0046] The beneficial effects of the present invention are:

[0047] (1) Multi-layer caged waterfowl feather stain intelligent detection and adaptive feather cleaning device: It includes a collaborative deployment strategy of multiple devices such as a core processing module, a QR code label, a waterfowl feather intelligent detection device, an adaptive feather cleaning device, a feed truck, and fecal water treatment to ensure the accuracy and efficiency in the detection and cleaning process of caged waterfowl feather stains.

[0048] (2) Retractable and anti-blocking nozzle design: The nozzle is connected by a telescopic rod, and the front and back movement of the nozzle can be controlled. The nozzle moves left and right along with the feed truck, which can make the cleaning more comprehensive and cost-saving. A water quality filtering system is equipped at the nozzle, effectively preventing the deposition of water scale and impurities, avoiding nozzle blockage, and improving the stability and service life of the equipment.

[0049] (3) Data monitoring and feedback: Flow sensors and temperature sensors are equipped at the nozzles, enabling the system to record water flow and temperature.

[0050] (4) Modular design: This system adopts a modular design, which can flexibly adjust the configurations of cameras and nozzles according to the actual scale and environmental requirements of the breeding farm. Through a core processing module, combined with multi-threading and thread pool technologies, it realizes the simultaneous control of multiple cameras and nozzles, and can perform personalized and autonomous decision-making cleaning on waterfowl. In addition, the system has good scalability, can adapt to breeding environments of different scales, and is convenient for later upgrading and maintenance.

[0051] (5) Intelligent removal of breeding cage nets: Based on the multi-stage defense method generative adversarial network combining the Unet semantic segmentation network and Pix2pixHD, it improves the accuracy of removing the occlusion of cage nets. Especially in complex images, it can effectively remove the cage nets and restore the image details of the occluded parts, providing an accurate basis for subsequent analysis of waterfowl feather stains.

[0052] (6) Intelligent identification of pollution: The present invention improves DepLabV3+ to obtain MDA-DepLabV3+ suitable for this scenario, which improves the accuracy of identifying two-dimensional code tags and waterfowl feather stains. Especially in complex environments, it can accurately judge the pollution degree of waterfowl feathers.

[0053] (7) Adaptive cleaning control: According to the real-time monitored pollution level, the system can dynamically adjust water flow, spraying time and water temperature, and implement personalized cleaning schemes, so that waterfowl with different pollution degrees can obtain the best cleaning effect, improving the cleaning effect and resource utilization rate.

[0054] (8) Separation and collection of waterfowl feces and water: The roller separation technology adopted in the present invention efficiently separates waterfowl feces from the cleaning water. The collected feces can be centrally processed and converted into organic fertilizers, while the cleaning water can be recycled or enter the sewage treatment system after preliminary filtration, realizing the recycling of water resources and environmental protection.

[0055] (9) Through the communication strategy of the waterfowl feather intelligent detection module, adaptive feather cleaning module, information storage module, and information analysis module, the cleaning efficiency is greatly improved, the labor cost is reduced, the health management level of waterfowl is improved, and at the same time, the accuracy and traceability of the cleaning process are ensured.

[0056] (10) Intelligent data recording and management: The system records the waterfowl pollution degree, cleaning frequency, water consumption and energy consumption data in real time, and uploads them to the background database, supporting data analysis and optimizing cleaning strategies to improve management efficiency. Description of the Drawings

[0057] Figure 1This is a flowchart of an intelligent method for detecting and cleaning the feathers of caged waterfowl according to the present invention.

[0058] Figure 2 This is a module diagram of an intelligent method for detecting and cleaning the feathers of caged waterfowl according to the present invention.

[0059] Figure 3 This is a top view schematic diagram of the system structure of an intelligent method for detecting and cleaning the feathers of caged waterfowl according to the present invention.

[0060] Figure 4 This is a side view schematic diagram of the system structure of an intelligent method for detecting and cleaning the feathers of caged waterfowl according to the present invention.

[0061] Among them: 1. High-definition camera; 2. Feeding cart; 3. Nozzle; 4. Telescopic rod; 5. Breeding cage; 6. QR code label; 7. Feeding trough; 8. Egg collection trough; 9. Core processing module; 10. Manure and water conveyor belt; 11. Roller; 12. Drinking water pipe. Specific implementation method

[0062] The present invention will be further described below in conjunction with the accompanying drawings and specific embodiments.

[0063] Example 1, as Figure 1 and Figure 2 shown, an intelligent method for detecting and cleaning the feathers of caged waterfowl includes the following steps:

[0064] S1: Move the feeding cart with a core processing module, a high-definition camera and a nozzle, and obtain the QR code label and waterfowl data in the breeding cage through the high-definition camera;

[0065] S2: Remove the breeding cage net in the image based on the defense algorithm through the core processing module, and restore the occluded part;

[0066] S3: Based on the restored image, use the improved MDA-DepLabV3+ model to detect the pollution degree of the QR code label and the waterfowl feathers, and obtain the proportion of the stain area by segmenting the waterfowl area and the stain area;

[0067] S4: Obtain the cleaning instruction by comparing the proportion of the stain area with the set threshold through the core processing module, and start the telescopic rod to move the nozzle to the top of the breeding cage according to the cleaning instruction;

[0068] S5: Execute the corresponding cleaning instruction by using the nozzle, and real-time monitor the relevant data through the flow sensor and temperature sensor equipped at the nozzle, and feedback the relevant data to the core processing module;

[0069] S6: Judge whether the feeding cart has moved to the end of the breeding cage. If so, enter step S7; otherwise, return to step S1 to continue moving the feeding cart;

[0070] S7: Start the manure water conveyor belt to remove waterfowl feces, and use the manure water separation device to dehydrate the collected wet manure and then transfer it to the fermentation tank for fermentation treatment to convert it into organic fertilizer;

[0071] S8: Use the intelligent detection of waterfowl feather stains and the adaptive feather cleaning management platform to uniformly manage and display the breeding farm.

[0072] In S1, the core processing module uses the parallel scheduling mechanism of multi-threading and thread pool as the control strategy to ensure that the system still has high responsiveness and high stability when facing multi-channel input and output tasks;

[0073] The parallel scheduling mechanism of the multi-threading and thread pool is specifically as follows: The core processing module assigns a control thread to each high-definition camera and the corresponding sprinkler control unit, including the camera thread and the sprinkler control thread. The camera thread and the sprinkler control thread are uniformly managed by the thread pool built in the core processing module. Among them, the camera thread is used to process image acquisition, feather stain recognition and positioning, and the sprinkler control thread is used to control the spraying action (water pressure, water volume, time, etc.) according to the recognition result.

[0074] This mechanism can not only process the cleaning processes of multiple waterfowl in parallel to avoid the delay caused by single-threaded serial execution, but also support real-time scheduling and thread load balancing, greatly improving the stability and processing efficiency of the system.

[0075] In order to prevent the occlusion problem of the breeding cage net from affecting the subsequent judgment of waterfowl feather stains, the present invention proposes a defense algorithm for removing the breeding cage net in the image. According to the proposed defense algorithm, the image is processed through three steps to remove the influence of the breeding cage net in the image and restore the clear image of the object behind the breeding cage net. Specifically, Unet is used for semantic segmentation, bitwise operation is used to remove the breeding cage net area, and Pix2pixHD generative adversarial network is used to restore the occluded part.

[0076] In S2, the defense algorithm performs the following operations:

[0077] S21: Perform semantic segmentation on the original image to obtain a semantic map through the Unet network. The formula is:

[0078]

[0079] Among them, is the obtained semantic map, is the Unet network, is the original image;

[0080] S22: Perform a bitwise operation on the original image according to the semantic map. By performing a masking operation, remove the part marked as the aquaculture cage net area in the semantic map from the original image to obtain an image after removing the aquaculture cage net. The formula is:

[0081]

[0082] where is the image after removing the aquaculture cage net, and are pixel points;

[0083] represents the aquaculture cage area, represents the non-aquaculture cage area;

[0084] S23: Based on the image after removing the aquaculture cage net, use the Pix2pixHD network to restore the image after removing the aquaculture cage net to a real image and restore the occluded part. The formula is:

[0085]

[0086] where is the restored image, is the Pix2pixHD network.

[0087] The core of image restoration is to use a generative adversarial network (GAN), specifically Pix2pixHD. Pix2pixHD is an image-to-image translation network suitable for tasks such as image inpainting and denoising. It restores the image after removing the aquaculture cage net to a real image through the adversarial training of a generator and a discriminator. Among them, the generator attempts to generate realistic images, and the discriminator evaluates the authenticity of the generated images. Through multiple iterations of training, the generator learns how to repair the missing parts in the image, making the restored image close to the real state.

[0088] The S3 includes the following sub-steps:

[0089] S31: Based on the restored image, use the improved MDA-DepLabV3+ model to identify the QR code label and use the Pyzbar library to decode it to extract the aquaculture cage number;

[0090] S32: Use the improved MDA-DepLabV3+ model to segment the waterfowl area and the stain area, and obtain the stain area ratio. The formula is:

[0091]

[0092] where is the stain area ratio, is the area of the stain area, is the total area of the waterfowl area.

[0093] The improved MDA-DepLabV3+ model replaces the traditional convolution Conv with depthwise separable convolution DSConv, replaces the original backbone network with the MobileNetV2 network, and introduces a dynamic atrous spatial pyramid pooling ASPP module.

[0094] Among them, by using depthwise separable convolution, the amount of computation and the number of parameters can be significantly reduced, thereby improving the inference speed and reducing the device burden. As a lightweight convolutional neural network, MobileNetV2's low computational complexity and high efficiency are very suitable for real-time processing in resource-constrained environments and adapting to embedded devices in the breeding environment. The ASPP module effectively captures multi-scale information in the image, helps the model identify stain areas of different scales, especially significantly improves the details in the feather area, and can identify smaller contaminated areas.

[0095] The training and evaluation of the MDA-DepLabV3+ model include the following steps:

[0096] First, collect image data containing QR code labels and stains on waterfowl in different scenarios, and use the Labelme annotation tool to accurately annotate the images. Then, divide the dataset into a training set, a validation set, and a test set according to the ratio of 7:2:1. To ensure the training effect, the size of all images is uniformly adjusted to 640x640 pixels. During the training process, use MDA-DepLabV3+ as the model architecture and conduct 200 rounds of training. After each round of training, use the validation set to evaluate the model, and save the best network training weights according to the validation results. Finally, evaluate the generalization ability of the model through the test set to ensure the stable performance of the model in actual applications and avoid overfitting. This training process helps to improve the accuracy and robustness of the model in the stain detection task. At the same time, through the evaluation of the validation set and the test set, it is ensured that the model has good generalization ability and can work stably in different waterfowl feather stain scenarios.

[0097] In this embodiment, the high-definition camera communicates with the core processing module through USB, and is used to detect the QR code label and the pollution degree of the waterfowl feathers in the corresponding breeding cage. The specific process is as follows:

[0098] First, it is necessary to accurately identify the QR code label and the stain degree of the waterfowl in a complex scenario. In addition, a core processing module needs to detect multiple pictures at the same time and complete the upload of the detection data. Therefore, the inference speed of the model needs to be improved. According to the requirements, the improved model MDA-DepLabV3+ is obtained to identify the QR code label and the waterfowl.

[0099] With the improved model, first, the QR code label is recognized, and the QR code is decoded through the Pyzbar library in the Python program to extract the number of the breeding cage. According to the result of QR code recognition, the program will save the corresponding cage number and detect the waterfowl identified in the cage. Next, the pollution degree of the waterfowl feathers will be evaluated.

[0100] After the improved MDA-DepLabV3+ model segments the waterfowl area and the stain area, to ensure the accuracy of recognition, the feeding cart will stay at each cage position for a period of time. After extracting the stain area, the proportion of the stain area is calculated.

[0101] Then, based on the stain proportion threshold defined in the operation management of the intelligent detection and adaptive feather cleaning management system platform for waterfowl feather stains 、 and , the waterfowl stains are divided into the following three categories:

[0102]

[0103] To ensure the accuracy of the cleaning process and the health management of waterfowl, the core processing module will generate cleaning instructions according to the pollution degree of each waterfowl. C The cleaning instructions include: starting the telescopic rod to extend the nozzle into the top of the cage, and selecting appropriate water flow pressure, cleaning time, and water flow temperature. This step ensures that the core processing module can transmit the correct pollution information to the adaptive feather cleaning device (including water flow sensor, temperature sensor, and telescopic rod) according to the result of stain detection.

[0104] Next, the core processing module generates different control signals according to the received pollution level information. Specifically, the pollution level information determines the cleaning operation parameters of the adaptive feather cleaning device:

[0105] Mild pollution: Instruct the adaptive feather cleaning device to start spraying with low water pressure for a short time, and keep the water temperature at room temperature (e.g., 20 - 25°C).

[0106] Moderate pollution: Instruct medium water pressure and moderate spraying time, and moderately increase the water temperature (e.g., 25 - 30°C).

[0107] Severe pollution: Instruct high water pressure and long-time spraying, and increase the water temperature (e.g., 30 - 40°C) to enhance the cleaning effect.

[0108] During this process, the cleaning instructions not only include the adjustment of water flow, spraying time, and water temperature, but also the operation of starting the telescopic rod. The core processing module will start the telescopic rod through the instruction to extend the nozzle into the top of the cage to ensure that the feather area of the waterfowl can be covered for cleaning.

[0109] The adaptive feather cleaning device monitors the spraying water volume and water temperature in real time through a flow sensor and a temperature sensor, and feeds back the relevant data to the core processing module. The system records data such as the pollution degree of waterfowl, cleaning frequency, water consumption, and energy consumption, and uploads these data to the background database for storage and analysis. Through data analysis, the system can optimize the cleaning strategy, improve the overall management efficiency, and ensure the high efficiency and accuracy of the waterfowl feather cleaning process.

[0110] The intelligent detection of waterfowl feather stains and the adaptive feather cleaning management platform in S8 includes:

[0111] Waterfowl feather stain cleaning report: The waterfowl feather stain cleaning report is used to record the cleaning operation information of waterfowl feathers each time, including waterfowl code, cleaning time, pollution degree, breeding cage number, cleaning water flow, spraying time, spraying water temperature, water consumption, energy consumption, cleaning effect evaluation, and relevant pictures generated during the cleaning process. These data can help the administrator monitor the cleaning situation of each waterfowl in real time, evaluate the cleaning effect, and provide a basis for subsequent cleaning strategies;

[0112] Group waterfowl cleaning performance report: The group waterfowl cleaning performance report is used to display the cleaning effect of the waterfowl group, including cleaning time distribution, average water flow, cleaning time, and group water resource consumption, and provides trend analysis of cleaning efficiency and effect to help optimize the group cleaning plan;

[0113] User feedback report: The user feedback report is used to upload feedback information through a small program, including information on incomplete feather cleaning and equipment failure information. The platform automatically generates a feedback report to help the administrator handle problems and improve the system service quality;

[0114] Equipment alarm report: The equipment alarm report is used to monitor the equipment status in real time and automatically generate an equipment failure alarm report to remind the administrator to repair in time, including equipment type, fault description, and occurrence time;

[0115] Operation management module: The operation management module is used to provide operation management functions, including threshold management, equipment management, farm building management, and waterfowl file management. Managers can set various management thresholds (such as water flow threshold, cleaning time threshold, temperature threshold, etc.) according to actual operation needs. The farm building management function helps the administrator manage the farm in zones, set cleaning parameters and equipment for different areas, and ensure that waterfowl in different areas get the most suitable cleaning conditions. At the same time, the waterfowl file management function can record the cleaning history, health status, etc. of each waterfowl, which is convenient for the administrator to conduct long-term tracking and management;

[0116] Report and Visualization Platform: The report and visualization platform has the function of generating various reports, supports detailed statistics and analysis of data cleaning, equipment status, and warning information, and provides various visualization methods to facilitate users to view and interpret data and grasp the system operation status in real time.

[0117] Embodiment 2, as Figure 3 and Figure 4 shown, a system for an intelligent sewage detection and feather cleaning method for caged waterfowl. The system includes multiple breeding cages 5 with the same structure, and the multiple breeding cages 5 are connected in parallel. A drinking water pipe 12, a QR code label 6, a feeding trough 7, an egg collection trough 8, a manure and water conveyor belt 10, and a roller 11 are arranged on the breeding cage 5. The drinking water pipe 12 is located inside the breeding cage 5, the feeding trough 7 is located outside the breeding cage 5, the QR code label 6 is pasted on the feeding trough 7, the egg collection trough 8 is located below the feeding trough 7, the manure and water conveyor belt 10 is located below each breeding cage 5, and the roller 11 is located at the tail of the manure and water conveyor belt 10;

[0118] The system further includes a feeding vehicle 2. The feeding vehicle 2 is located directly in front of the feeding trough 7 of the multi-layer breeding cages 5. A core processing module 9 is mounted in the middle of the feeding vehicle 2. The core processing module 9 is connected to multiple high-definition cameras 1 and multiple spray heads 3. The spray heads 3 are connected to the feeding vehicle 2 through telescopic rods 4.

[0119] The telescopic rod 4 is driven by an electric push rod or a pneumatic cylinder. The spray head 3 internally integrates a water quality filtering unit, including a replaceable filter element and a water pressure stabilizer.

[0120] Both sides of the manure and water conveyor belt 10 are rolled up to form a sealed enclosure structure, which can effectively prevent the overflow of waterfowl manure and sprayed water during transportation, ensuring a clean environment. The manure and water conveyor belt 10 is transported by a motor drive. A manure scraping plate is installed at the end of the manure and water conveyor belt 10, which can efficiently remove the manure on the conveyor belt surface, and an adjustable roller 11, which can press out the excess water in the wet manure, so that the sprayed water and the water seeping out of the waterfowl manure can be effectively introduced into the drainage ditch and enter the subsequent water resource treatment system. During the manure collection process, a special "wet and dry separation unit" is set up, which can separate the collected wet manure from the excess water and reduce the moisture content of the manure. The wet manure can be further dehydrated through a wet and dry separation device, or guided to a fermentation tank for fermentation treatment to be converted into organic fertilizer.

[0121] Those of ordinary skill in the art will realize that the embodiments described herein are to assist the reader in understanding the principles of the present invention, and it should be understood that the scope of protection of the present invention is not limited to such specific statements and embodiments. Those of ordinary skill in the art can make various other specific deformations and combinations that do not depart from the essence of the present invention based on these technical revelations disclosed in the present invention, and these deformations and combinations are still within the scope of protection of the invention.

Claims

1. An intelligent method for detecting and cleaning dirt on caged waterfowl, characterized in that, Including the following steps: S1: Move the material vehicle with a core processing module, a high-definition camera and a nozzle, and obtain the QR code label and waterfowl data in the breeding cage through the high-definition camera; S2: Remove the breeding cage net in the image based on the defense algorithm through the core processing module, and restore the occluded part; In the S2, the defense algorithm performs the following operations: S21: Perform semantic segmentation on the original image, and obtain a semantic map through the Unet network. The formula is: Among them, is the obtained semantic graph, is the Unet network, is the original image; S22: Perform bitwise operation on the original image according to the semantic map. Through mask operation, remove the part marked as the breeding cage net area in the semantic map from the original image to obtain the image after removing the breeding cage net. The formula is: Among them, is the image after removing the aquaculture cage net, and are pixel points; S23: Based on the image after removing the breeding cage net, use the Pix2pixHD network to restore the image after removing the breeding cage net to a real image, and restore the occluded part. The formula is: Among them, is the restored image, is the Pix2pixHD network; S3: Based on the restored image, use the improved MDA-DepLabV3+ model to detect the QR code label and the degree of waterfowl feather pollution, and obtain the proportion of the stain area by segmenting the waterfowl area and the stain area; The following sub-steps are included in the S3: S31: Based on the restored image, use the improved MDA-DepLabV3+ model to identify the QR code label, and use the Pyzbar library to decode it to extract the breeding cage number; S32: Use the improved MDA-DepLabV3+ model to segment the waterfowl area and the stain area, and obtain the proportion of the stain area. The formula is: Among them, is the proportion of the stain area, is the area of the stain region, is the total area of the waterfowl region; The improved MDA-DepLabV3+ model uses depthwise separable convolution DSConv to replace the traditional convolution Conv, uses the MobileNetV2 network to replace the original backbone network, and introduces a dynamic atrous spatial pyramid pooling ASPP module; S4: The core processing module compares the stain area proportion with the set threshold to obtain a cleaning instruction, and starts the telescopic rod to move the nozzle to the top of the breeding cage according to the cleaning instruction; S5: Use the nozzle to execute the corresponding cleaning instruction, and use the flow sensor and temperature sensor equipped at the nozzle to monitor the relevant data in real time, and feed the relevant data back to the core processing module; S6: Judge whether the material vehicle has moved to the end of the breeding cage. If so, enter step S7. Otherwise, return to step S1 to continue moving the material vehicle; S7: Start the manure water conveyor belt to remove waterfowl feces, and use the manure water separation device to dehydrate the collected wet manure and then transport it to the fermentation tank for fermentation treatment to be converted into organic fertilizer; S8: Use the intelligent detection of waterfowl feather stains and the adaptive feather cleaning management platform to uniformly manage and display the breeding farm.

2. The intelligent sewage detection and feather cleaning method for caged waterfowl according to claim 1, characterized in that, In the S1, the core processing module adopts a parallel scheduling mechanism of multi-threading and thread pool as the control strategy; The specific parallel scheduling mechanism of the multi-threading and thread pool is: The core processing module allocates a control thread for each high-definition camera and the corresponding nozzle control unit, including a camera thread and a nozzle control thread. The camera thread and the nozzle control thread are uniformly managed by the thread pool built in the core processing module.

3. The intelligent sewage detection and feather cleaning method for caged waterfowl according to claim 1, characterized in that The intelligent detection of waterfowl feather stains and the adaptive feather cleaning management platform in the S8 includes: Waterfowl Feather Stain Cleaning Report: The waterfowl feather stain cleaning report is used to record the cleaning operation information of waterfowl feathers each time, including waterfowl code, cleaning time, pollution degree, breeding cage number, cleaning water flow rate, spraying time, spraying water temperature, water consumption, energy consumption, cleaning effect evaluation, and relevant pictures generated during the cleaning process; Group Waterfowl Cleaning Performance Report: The group waterfowl cleaning performance report is used to display the cleaning effect of the waterfowl group, including cleaning time distribution, average water flow rate, cleaning time, and group water resource usage, and provides analysis of cleaning efficiency and effect trends to help optimize the group cleaning plan; User Feedback Report: The user feedback report is used to upload feedback information through a small program, including information on incomplete feather cleaning and equipment failure information. The platform automatically generates a feedback report to help administrators handle problems and improve system service quality; Equipment Alarm Report: The equipment alarm report is used to monitor the equipment status in real time and automatically generate an equipment failure alarm report to remind administrators to repair in time, including equipment type, fault description, and occurrence time; Operation Management Module: The operation management module is used to provide operation management functions, including threshold management, equipment management, farm building management, and waterfowl file management; Report and Visualization Display Platform: The report and visualization display platform has the function of generating various reports, supports detailed statistics and analysis of cleaning data, equipment status, and warning information, and provides multiple visualization display methods.

4. A system for the intelligent sewage detection and feather cleaning method of caged waterfowl according to any one of claims 1-3, characterized in that, The system includes multiple breeding cages (5) with the same structure, and the multiple breeding cages (5) are connected in parallel. A drinking water pipe (12), a QR code label (6), a feeding trough (7), an egg collection trough (8), a manure and wastewater conveyor belt (10), and a roller (11) are arranged on the breeding cage (5). The drinking water pipe (12) is located inside the breeding cage (5), the feeding trough (7) is located outside the breeding cage (5), the QR code label (6) is pasted on the feeding trough (7), the egg collection trough (8) is located below the feeding trough (7), the manure and wastewater conveyor belt (10) is located below each breeding cage (5), and the roller (11) is located at the tail of the manure and wastewater conveyor belt (10); The system further includes a feeding vehicle (2). The feeding vehicle (2) is located directly in front of the feeding troughs (7) of the multi-layer breeding cages (5). A core processing module (9) is mounted in the middle of the feeding vehicle (2). The core processing module (9) is connected to multiple high-definition cameras (1) and multiple nozzles (3). The nozzles (3) are connected to the feeding vehicle (2) through telescopic rods (4).

5. The system according to claim 4, wherein The telescopic rod (4) is driven by an electric push rod or a pneumatic cylinder. The nozzle (3) is internally integrated with a water quality filtering unit, including a replaceable filter element and a water pressure stabilizer.

6. The system according to claim 4, wherein Both sides of the manure and wastewater conveyor belt (10) are rolled up to form a sealed enclosure structure. The manure and wastewater conveyor belt (10) is transported by a motor. A manure scraping plate and an adjustable roller (11) are installed at the end of the manure and wastewater conveyor belt (10), which can squeeze out the excess water in the wet manure.

Images