A cage-bred meat duck inspection device
By combining mobile robot devices with deep learning and environmental perception technologies, a comprehensive health assessment and environmental monitoring of caged ducks has been achieved, overcoming the shortcomings of manual inspections and improving inspection efficiency and accuracy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NANJING AGRI MECHANIZATION INST MIN OF AGRI
- Filing Date
- 2026-06-18
- Publication Date
- 2026-07-31
AI Technical Summary
In large-scale caged duck production, manual inspections are difficult to achieve high-frequency, full-coverage inspections, have low recognition rates, cannot objectively quantify the health status of the duck flock, and lack comprehensive environmental data acquisition and multimodal data fusion analysis, resulting in delayed abnormal early warnings.
The mobile robot device integrates a lower camera, an upper camera, a 360° probe, an environmental perception module, and a lidar. Combined with deep learning algorithms and an inertial measurement unit, it can achieve comprehensive health assessment and environmental monitoring of duck flocks, generate environmental comfort distribution maps, navigate autonomously, and generate graded alarm signals.
It enables comprehensive inspection of multi-layer caged duck flocks, improves the anomaly identification rate and early warning accuracy, reduces the underreporting rate of casualties, and provides accurate location of environmental anomalies and health status assessment.
Smart Images
Figure CN122496618A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of intelligent livestock breeding inspection devices, specifically an inspection device for caged meat ducks. Background Technology
[0002] In large-scale caged duck production, the health status of the duck flock is mainly monitored by the farmers who regularly patrol the sheds. They observe the ducks' mental state, walking posture, feeding and drinking habits, and feces in each cage layer with the naked eye. Based on their experience, they judge whether there are any sick or injured individuals and whether the temperature, humidity, and ammonia concentration in the shed are suitable. This manual inspection method has the following prominent problems: the narrow and long aisles of caged duck houses require keepers to repeatedly bend over or climb to check multiple layers of cages, making it difficult to achieve high-frequency, full-coverage inspections and often missing the critical window for early disease outbreaks; relying on personal experience, the recognition rate of subtle abnormal postures such as lameness and early paralysis is low, and it is impossible to objectively quantify the activity index and distribution density of the duck flock, resulting in delayed early warning of abnormalities; handheld detectors carried by workers can only obtain environmental data from a few discrete points in the aisles, and cannot form an environmental comfort distribution map corresponding to the cage location, making it difficult to accurately locate environmental stressors; visual observation, sound judgment, and environmental detection are disconnected, lacking multimodal data fusion analysis, making it impossible to make a comprehensive and three-dimensional joint diagnosis of the health status of meat ducks. Therefore, an improved technology is urgently needed to solve these problems existing in the current technology. Summary of the Invention
[0003] The purpose of this invention is to provide an inspection device for caged meat ducks to solve the problems mentioned in the background art.
[0004] To achieve the above objectives, the present invention provides the following technical solution: a caged duck inspection device, comprising a mobile robot. The mobile robot has a steering wheel and two drive wheels at its bottom. Lower cameras are provided on both sides of the lower part of the mobile robot. Lifting rods are also provided on both sides of the mobile robot. Upper cameras are provided at the top of each lifting rod facing outward. A 360° probe is provided in the middle of the upper surface of the mobile robot. The mobile robot integrates a duck flock health assessment unit based on deep learning. The health assessment unit receives video streams from the lower cameras, upper cameras, and 360° probes, respectively, and identifies lame, paralyzed, overturned, and dead ducks in real time through a posture estimation algorithm. It also counts the duck flock activity index and distribution density. When the number of abnormal individuals or the distribution density exceeds a set threshold, a graded alarm signal is generated.
[0005] Preferably, the present invention provides a caged duck inspection device, wherein the mobile robot is also embedded with environmental sensing modules on both sides. The environmental sensing modules include at least an ammonia sensor, a hydrogen sulfide sensor, a temperature and humidity sensor, and a wind speed sensor. Each sensor synchronously collects environmental data at fixed time intervals along the inspection path and binds and stores the image data with the corresponding lower and upper cameras to generate an environmental comfort distribution map of different areas of the cage.
[0006] Preferably, the present invention provides an inspection device for caged meat ducks, wherein the front end and both sides of the lower part of the mobile robot are equipped with lidar, and the mobile robot is equipped with an inertial measurement unit and a synchronous positioning and mapping module; the lidar, the inertial measurement unit and the encoder of the drive wheel are fused to generate a two-dimensional grid map of the cage walkway, and the mobile robot is controlled to autonomously navigate along a preset inspection route and stop at a fixed point in front of the designated cage position.
[0007] Preferably, the present invention provides a caged duck inspection device, wherein the mobile robot is equipped with a directional microphone array on both sides, the microphone array collects the duck calls and transmits them to an audio anomaly detection module; the audio anomaly detection module performs real-time analysis of the sound pressure level and frequency characteristics of the duck calls, identifies stress screams or abnormal silence patterns, and confirms the health status of the duck flock in conjunction with the image recognition results.
[0008] Preferably, the present invention provides a caged duck inspection device, wherein the upper surface of the mobile robot is further equipped with a wireless charging receiving coil, which is used to align and couple with a wireless charging transmitting coil buried under the floor at the inspection starting point, so as to automatically replenish the power of the mobile robot during the inspection interval.
[0009] Preferably, the present invention provides an inspection device for caged meat ducks, wherein the surface of the mobile robot is also provided with a control panel and several buttons.
[0010] Compared with the prior art, the beneficial effects of the present invention are: (1) The lower camera takes pictures of the lower cages, the lifting rod is used in conjunction with the upper camera to accurately target each layer of cages, and the 360° probe monitors the entire walkway and also takes into account the top cages, so as to achieve a blind spot inspection of multi-layer caged ducks and avoid blind spots.
[0011] (2) The fusion mapping of lidar, inertial measurement unit and encoder enables the robot to autonomously plan its path in the narrow aisle of the cage and stop precisely in front of each cage position, ensuring that the upper and lower cameras can clearly capture the posture and images of the ducks in the corresponding cage.
[0012] (3) The deep learning-based posture estimation algorithm can identify abnormal postures such as limping, paralysis, falling, and death in real time. Combined with activity index and distribution density statistics, it can automatically generate early warning signals when abnormal indicators exceed the limit, realize objective and high-frequency early disease screening, and significantly reduce the underreporting rate of injuries and deaths.
[0013] (4) The environmental perception module is bound to the location to collect ammonia, hydrogen sulfide, temperature, humidity and wind speed, and generate an environmental comfort distribution map along the inspection route. It can quickly locate abnormal environmental areas and provide accurate basis for ventilation adjustment.
[0014] (5) By capturing abnormal acoustic patterns of duck flocks through a directional microphone array and cross-validating them with visual health assessment results, the alarm can be effectively identified as screaming or abnormal silence caused by stress or disease, thereby improving alarm accuracy and reducing false alarms. Attached Figure Description
[0015] Figure 1 This is a front view structural diagram of the present invention; Figure 2 This is a side view of the structure of the present invention; Figure 3 This is a schematic diagram of the working state of the present invention.
[0016] In the diagram: 1-Mobile robot, 2-Steering wheel, 3-Drive wheel, 4-Lower camera, 5-Lifting rod, 6-Upper camera, 7-360° probe, 8-Environmental perception module, 9-LiDAR, 10-Pickup array, 11-Wireless charging receiver coil, 12-Control panel, 13-Button. Detailed Implementation
[0017] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present invention. It should be noted that in the description of this invention, the terms "inner", "outer", "upper", "lower", "both sides", "one end", "the other end", "left", "right", etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are only for the convenience of describing this invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limiting this invention.
[0018] Please see Figure 1-2The present invention provides a technical solution: a caged duck inspection device, including a mobile robot 1. The bottom of the mobile robot 1 is provided with a steering wheel 2 and two independently driven drive wheels 3. The steering wheel 2 is used to achieve smooth guidance, and the drive wheels 3 are used to drive the device to move.
[0019] Lower cameras 4 are installed on the lower sides of the mobile robot 1 to capture the activities of the ducks at the bottom and second layers of cages on both sides. Lifting rods 5 are also installed on both sides, with upper cameras 6 installed at their tops facing outwards. The lifting rods 5 can be raised or lowered. The driving structure of the lifting rods 5 is roughly as follows: the rod body is equipped with a stepper motor and a transmission screw. The stepper motor drives the screw to rotate, which drives the rod body to extend or retract through the nut, thereby smoothly adjusting the height of the upper cameras 6 so that the upper cameras 6 can be aimed at the activity areas of the ducks in the cages from the third to the top layer. A 360° probe 7 is also installed in the middle of the upper surface of the mobile robot 1. The probe 7 consists of multiple circumferentially arranged visible light cameras and at least one infrared thermal imaging camera to generate a panoramic image of the cage aisle and simultaneously detect abnormalities in the ducks' body surface temperature.
[0020] The mobile robot 1 integrates a core deep learning-based duck flock health assessment unit. This unit receives real-time video streams from the lower camera 4, the upper camera 6, and the 360° probe 7. Using a pre-trained convolutional neural network and pose estimation model (such as an algorithm based on skeleton key point detection), it identifies the lameness, paralysis, overturning, and individual mortality status of each duck frame by frame. At the same time, it counts the activity index and distribution density of the duck flock in each cage. When the number of abnormal individuals identified in a cage exceeds the preset number, or when a sudden change in distribution density indicates abnormal behavior such as crowding or huddling, the unit immediately generates a graded alarm signal (such as a yellow warning, an orange alarm, and a red emergency alarm), which is sent to the host computer or the breeder's handheld terminal via a wireless module.
[0021] To further acquire information about the cage environment, environmental sensing modules 8 are embedded on both side panels of the mobile robot 1. These modules integrate at least an ammonia sensor, a hydrogen sulfide sensor, a temperature and humidity sensor, and a wind speed sensor. During the inspection, each sensor collects data synchronously at fixed time intervals and adds a current timestamp and a location tag provided by the synchronous positioning and mapping module. The collected environmental data is bound to the image data captured by the lower camera 4 and the upper camera 6 at the corresponding locations and stored in the internal memory. After the inspection is completed, the data is transmitted to the host computer software in real time to generate an "environmental comfort distribution map" along the entire walkway. The map uses different colors to mark areas with excessive ammonia concentration, wind speed dead zones, etc., providing an intuitive basis for environmental control adjustment.
[0022] For autonomous navigation, the mobile robot 1 is equipped with LiDAR 9 at the front end and both sides of its lower part. The body integrates an inertial measurement unit (IMU) and a SLAM synchronous localization and mapping module. The SLAM module integrates the point cloud data of LiDAR 9, the attitude data of IMU and the mileage data of drive wheel encoder to build a two-dimensional grid map of the cage walkway in real time. Based on the map, the optimal inspection path is planned. The control unit drives the drive wheel 3 and the steering wheel 2 to make the robot move along the predetermined route and stop at each cage position with a stopping accuracy within ±2cm to ensure that the camera field of view accurately covers the target cage.
[0023] The mobile robot 1 is also equipped with directional microphone arrays 10 on both sides. The arrays are oriented towards the cages on both sides to pick up the calls of the ducks. The microphone arrays 10 are connected to the audio anomaly detection module. This module extracts features such as Mel frequency cepstral coefficients from the collected sound signals and uses a trained classifier to detect stress screaming patterns and abnormal silence patterns in real time. When the visual posture estimation finds a few anomalies but is uncertain, if the sound module detects stress screaming at the same time, the system will increase the alarm confidence. If abnormal silence is detected in a certain cage, it may indicate that the ducks are sick, such as having a fever or being lethargic. That is, the health status of the ducks is confirmed in conjunction with the image recognition results, which greatly reduces the false alarm rate.
[0024] To address the energy replenishment issue during continuous inspections, a wireless charging receiving coil 11 is integrated on the upper surface of the mobile robot 1. A wireless charging transmitting coil is pre-embedded under the floor at the inspection starting point or the end of the walkway. When the robot returns to the same position after completing a round of inspections, the receiving coil 11 and the transmitting coil automatically align and couple, charging the onboard battery pack via electromagnetic induction, thus achieving automatic energy replenishment without intervention during inspection intervals.
[0025] In addition, a control panel 12 and several buttons 13 are provided on the surface of the casing of the mobile robot 1. The control panel 12 can be a touch screen to display the inspection status, alarm information, remaining power and current level. The buttons 13 include an emergency stop button, an inspection start / pause button, and a manual control button for the lifting rod, which facilitates on-site operation and emergency intervention by the zookeeper.
[0026] The device operates as follows: Before inspection, the inspection route, camera layer switching strategy, and alarm threshold are set via control panel 12. Mobile robot 1 autonomously moves along the aisle using a grid map generated by the mapping module, stopping at each cage position. During this stop, lower camera 4 continuously captures images of the lower cages, while lifting rod 5 drives upper camera 6 to sequentially inspect each cage position in the middle and upper layers. Figure 3As shown, the 360° probe 7 simultaneously collects panoramic data of the walkway and the ducks' body temperature, while the environmental sensing module 8 simultaneously records the air quality, temperature, and humidity at the current location. The microphone array 10 continuously monitors the ducks' sounds. All data is aggregated to the health assessment unit and the audio anomaly detection module, which outputs real-time duck posture recognition results, activity index, abnormal individuals, and environmental comfort indicators. Once an abnormality is detected, a graded alarm is immediately issued, and the abnormal cage number and corresponding screenshot are recorded for timely intervention by the keepers. After completing the entire inspection, the robot automatically returns to the charging point for wireless charging, awaiting the next round of inspections or going into hibernation as planned.
[0027] Any aspects of this invention not described in detail are well-known to those skilled in the art.
[0028] Finally, it should be noted that the above specific embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications and equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications and substitutions should be covered within the scope of the claims of the present invention.
Claims
1. A device for inspecting caged meat ducks, characterized in that: The system includes a mobile robot (1), which has a steering wheel (2) and two drive wheels (3) at its bottom. The lower sides of the mobile robot (1) are equipped with lower cameras (4) respectively. The sides of the mobile robot (1) are equipped with lifting rods (5) respectively. The top of the lifting rods (5) is equipped with upper cameras (6) facing outward. The middle position of the upper surface of the mobile robot (1) is equipped with a 360° probe (7). The mobile robot (1) integrates a duck flock health assessment unit based on deep learning. The health assessment unit receives video streams from the lower camera (4), the upper camera (6) and the 360° probe (7) respectively. It identifies individuals of ducks that are lame, paralyzed, overturned and dead in real time through a posture estimation algorithm, and counts the duck flock activity index and distribution density. When the number of abnormal individuals or the distribution density exceeds a set threshold, a graded alarm signal is generated.
2. The inspection device for caged meat ducks according to claim 1, characterized in that: The mobile robot (1) is also equipped with an environmental perception module (8) on both sides. The environmental perception module (8) includes at least an ammonia sensor, a hydrogen sulfide sensor, a temperature and humidity sensor and a wind speed sensor. Each sensor collects environmental data synchronously at fixed time intervals on the inspection path and binds and stores the image data of the lower camera (4) and the upper camera (6) at the corresponding positions to generate an environmental comfort distribution map of different areas of the cage.
3. The inspection device for caged meat ducks according to claim 1, characterized in that: The mobile robot (1) is equipped with a laser radar (9) at the front end and both sides of the lower part. The mobile robot (1) is equipped with an inertial measurement unit and a synchronous positioning and mapping module. The encoder data of the laser radar (9), the inertial measurement unit and the drive wheel (3) are fused to generate a two-dimensional grid map of the cage walkway, and control the mobile robot (1) to navigate autonomously along the preset inspection route and stop at the designated cage position.
4. The inspection device for caged meat ducks according to claim 1, characterized in that: The mobile robot (1) is also equipped with a directional microphone array (10) on both sides. The microphone array (10) collects the duck calls and transmits them to the audio anomaly detection module. The audio anomaly detection module analyzes the sound pressure level and frequency characteristics of the duck calls in real time, identifies stress screams or abnormal silence patterns, and confirms the health status of the ducks in conjunction with the image recognition results.
5. The inspection device for caged meat ducks according to claim 1, characterized in that: The mobile robot (1) is also equipped with a wireless charging receiving coil (11) on its upper surface. The wireless charging receiving coil (11) is used to align and couple with the wireless charging transmitting coil buried under the floor at the inspection starting point, so as to automatically replenish the power of the mobile robot (1) during the inspection interval.
6. The inspection device for caged meat ducks according to claim 1, characterized in that: The mobile robot (1) is also equipped with a control panel (12) and several buttons (13) on its surface.