Laboratory material storing and taking goods shelf with AI recognition function and using method of laboratory material storing and taking goods shelf

Abstract

The invention discloses a laboratory material storing and taking goods shelf with an AI recognition function and a using method thereof, and relates to the technical field of laboratory material management. The device comprises a frame body, a storage space adjusting assembly and an intelligent recognition assembly, a fixing frame is fixedly connected to the right side of an inner cavity of the frame body, transparent sealing doors are movably connected to the two sides of the front side of the frame body, a temperature control mechanism is fixedly connected to the top of the frame body, and the storage space adjusting assembly comprises a control shell. Through the storage space adjusting assembly, a stepping motor is used for driving a screw rod to rotate, so that a threaded sleeve drives a supporting plate to vertically ascend and descend, an adjustable interlayer spacing structure is formed in cooperation with a positioning plate, and the optimal storage height can be automatically matched according to size information of reagent bottles or instrument equipment; the problem of space waste caused by the fixed layer height of a traditional goods shelf is solved, and when the sensor detects oversized goods and materials, the stepping motor can be in linkage with the multiple sets of supporting plates to cooperatively lift, and the single-layer effective volume is expanded.

Description

Technical Field

[0001] The present invention belongs to the technical field of laboratory material management, and particularly relates to a laboratory material access shelf with AI recognition and its usage method. Background Technique

[0002] A laboratory material access shelf is a dedicated storage device designed for laboratory scenarios, integrating functional modules such as classified storage, quick positioning, and convenient access. It can accommodate various materials such as experimental consumables, reagents, and small instruments. Through scientific planning of storage space and reasonable layout, the orderly placement and efficient management of materials are achieved.

[0003] Laboratory shelves are an indispensable part of scientific research and medical work. Traditional laboratory shelves are usually of a fixed design and cannot flexibly adjust the storage space according to actual needs. Moreover, when laboratory personnel access materials, they need to manually search and carry them. Especially in the case of a large variety and quantity of materials, the operation is cumbersome and prone to errors, which easily leads to laboratory personnel taking or placing the wrong materials and is not conducive to use.

[0004] Therefore, we provide a laboratory material access shelf with AI recognition and its usage method to solve the above problems. Summary of the Invention

[0005] The purpose of the present invention is to provide a laboratory material access shelf with AI recognition and its usage method. Through the cooperation of an adjustment component and an intelligent recognition component, the problems in the prior art that the laboratory material access shelf cannot automatically adjust the storage space according to the specifications of stored materials and cannot automatically take and count materials, which easily leads to laboratory personnel taking or placing the wrong materials, are solved.

[0006] To solve the above technical problems, the present invention is realized through the following technical solutions.

[0007] The present invention relates to a laboratory material access shelf with AI recognition, which includes a rack body, a storage space adjustment component, and an intelligent recognition component. A fixed rack is fixedly connected to the right side of the inner cavity of the rack body. Transparent sealing doors are movably connected to both sides of the front side of the rack body. A temperature control mechanism is fixedly connected to the top of the rack body. The storage space adjustment component includes a control shell, the rear side of which is fixedly connected to the inner wall of the rack body. A stepping motor is fixedly connected to the rear side of the inner cavity of the control shell. Screws are fixedly connected to the top and bottom of the output end of the stepping motor. A threaded sleeve is threadedly connected to the surface of the screw. The front side of the threaded sleeve penetrates through the control shell and is fixedly connected to a support plate. A positioning plate is fixedly connected to the front side of the control shell. The intelligent recognition component includes a controller, the rear side of which is fixedly connected to the transparent sealing door. Photographing mechanisms are fixedly connected to both sides of the rack body. The controller includes a central processing unit, which is bidirectionally electrically connected to a visual data acquisition and preprocessing module, a hand movement tracking and intention recognition module, a material recognition and status detection module, and a real-time feedback and error prevention execution module.

[0008] The present invention is further configured such that the temperature control mechanism includes an exhaust hood, the bottom of which is communicated with the rack body. Exhaust fans are communicated with both sides of the rear side of the exhaust hood. A heating pipe is fixedly connected between both sides of the inner cavity of the exhaust hood. A perforated plate is fixedly connected to the bottom of the exhaust hood. The exhaust fan can convey external air into the exhaust hood. The heating pipe is used for heating the air. The perforated plate can prevent external impurities from entering the rack body.

[0009] The present invention is further configured such that humidity sensors are fixedly connected to both sides of the inner cavity of the fixed rack, and a human proximity sensor is fixedly connected to the front side of the top of the rack body. The humidity sensors can monitor the temperature and humidity inside the rack body in real time. When the temperature and humidity are abnormal, the temperature control mechanism can be used to adjust the temperature and humidity inside the rack body. The human proximity sensor can detect whether an experimenter is approaching. After a person approaches, the photographing mechanism can be controlled to start through the controller.

[0010] The present invention is further configured such that the photographing mechanism includes two linear motors, the opposite sides of which are fixedly connected to the rack body. Mounting frames are fixedly connected to the output ends of the opposite sides of the two linear motors. A camera is fixedly connected to the bottom of the mounting frame. The linear motor can control the mounting frame and the camera to move up and down, so that the camera can photograph the types of materials taken by the experimenter.

[0011] The present invention is further configured such that sliding rods are fixedly connected to the rear sides of both sides of the rack body. A sliding sleeve is slidably connected to the surface of the sliding rod. The front side of the sliding sleeve is fixedly connected to the mounting frame. The sliding rod and the sliding sleeve can limit the mounting frame, so that it can stably control the movement of the camera and prevent it from shaking during the movement process.

[0012] The present invention is further configured such that limiting grooves are provided on both sides of the rear side of the support plate, a limiting strip is arranged in the inner cavity of the limiting groove, the rear side of the limiting strip is fixedly connected to the inner wall of the frame body, and the limiting grooves and the limiting strip can limit the support plate so that it can move up and down stably and prevent it from tilting during the adjustment process.

[0013] The present invention is further configured such that the central processor is bidirectionally electrically connected to an adaptive learning and system optimization module, an environment perception and anomaly monitoring module, and an interactive guidance module. The adaptive learning and system optimization module constructs a dynamic model update mechanism to continuously improve the new material recognition ability and environmental adaptability. The environment perception and anomaly monitoring module realizes all-weather intelligent supervision and leakage warning of the chemical storage environment through thermal imaging and dynamic analysis technology. The interactive guidance module combines augmented reality projection and voice interaction technologies to provide screenless operation guidance and material positioning services.

[0014] The present invention is further configured such that the adaptive learning and system optimization module includes an incremental feature library update unit and a light intensity adaptive switching unit, the environment perception and anomaly monitoring module includes a temperature and humidity anomaly detection unit and a liquid leakage identification unit, and the interactive guidance module includes an AR projection positioning unit and a voice command parsing unit. The incremental feature library update unit and the light intensity adaptive switching unit continuously adapt to the new material recognition requirements through incremental feature library updates, and dynamically switch model parameters in combination with the light intensity mapping table to ensure the operation stability in complex environments. The temperature and humidity anomaly detection unit and the liquid leakage identification unit realize intelligent monitoring of the temperature and humidity of chemical storage based on thermal imaging technology, and quickly locate the diffusion boundary of liquid leakage by combining the inter-frame difference method and the region growing algorithm. The AR projection positioning unit and the voice command parsing unit use AR spot projection technology to achieve millimeter-level operation guidance positioning, and integrate offline voice command parsing to support screenless material retrieval and process guidance.

[0015] The present invention is further configured such that the visual data acquisition and preprocessing module includes a wide-angle distortion correction unit and a dynamic ROI segmentation unit, and the hand motion tracking and intention recognition module includes a multi-joint motion analysis unit and a temporal action classification unit. The wide-angle distortion correction unit and the dynamic ROI segmentation unit eliminate the distortion of the wide-angle lens through image distortion correction technology, and accurately focus on the operation area by combining the dynamic region segmentation algorithm, effectively shielding background interference such as personnel movement, and providing a standardized input for subsequent analysis. The multi-joint motion analysis unit and the temporal action classification unit analyze the three-dimensional motion trajectory of the operator based on bone key point modeling, and identify the complete behavior chain of "grab - move - release" through temporal action classification, and synchronously detect the risk of abnormal jitter or deviation from the path.

[0016] The present invention is further configured such that the material identification and status detection module includes a multispectral feature extraction unit and a three-dimensional space verification unit, and the real-time feedback and error prevention execution module includes a multimodal alarm linkage unit and a mechanical locking control unit. The multispectral feature extraction unit and the three-dimensional space verification unit fuse visible light and near-infrared spectra to enhance the identification of reflective surface features, and combine three-dimensional point cloud reconstruction technology to monitor the spatial posture of materials in real time, preventing potential safety hazards caused by stacking imbalance or incorrect bottle orientation. The multimodal alarm linkage unit and the mechanical locking control unit integrate audible and visual hierarchical alarms and electromagnetic locking devices to immediately block incorrect operations, support physical isolation of high-risk product cabinets and biometric recording, and form a forced error correction mechanism.

[0017] A method for using a laboratory material storage and retrieval shelf with AI recognition includes the following steps: S1. When an experimenter approaches the shelf body, the human proximity sensor detects the human signal, triggering the imaging mechanism to start working. The linear motor drives the mounting frame to move along the sliding rod, so that the camera is at the best shooting angle. The cameras on both sides of the shelf body synchronously collect images of the operation area. After eliminating image distortion through the wide-angle distortion correction unit, the images are transmitted to the central processor; S2. When the experimenter stores materials, the storage space adjustment component automatically adjusts the layer height according to the size of the materials. When the materials are placed on the pallet, the camera takes pictures and detects the height of the materials. The stepping motor drives the screw to rotate, so that the screw drives the threaded sleeve to lift and lower to a suitable position. Multiple pallets can keep synchronous lifting and lowering through the limit slots and limit bars, realizing intelligent allocation of the storage space; S3. When retrieving materials, the hand movement tracking and intention recognition module captures the operation actions through the camera and compares them with the characteristic data stored in the material identification and status detection module. When a misretrieval behavior is detected, the mechanical locking control unit immediately locks the transparent sealing door, and at the same time, the AR projection positioning unit projects a light spot at the correct material position for prompting. During the whole working process, the central processor collects data through various sensors and cameras, coordinates the operation of the storage space adjustment component and the intelligent recognition component, and realizes the intelligent management and safe storage and retrieval of laboratory materials; S4. During the material storage process, the temperature and humidity sensor continuously monitors the internal environment data of the shelf body. When an abnormal temperature is detected, the temperature control mechanism is automatically started. The exhaust fan inhales external air through the exhaust hood, and the heating pipe heats the air flow according to the set temperature. The adjusted air is evenly distributed to the inside of the shelf through the perforated plate. If a liquid leakage occurs, the liquid leakage identification unit quickly locates the leakage position by analyzing the change of the images taken by the camera.

[0018] The present invention has the following beneficial effects.

[0019] 1. The present invention uses a storage space adjustment component and a stepper motor to drive the screw to rotate, so that the threaded sleeve drives the pallet to move vertically up and down, and cooperates with the positioning plate to form an adjustable layer spacing structure. It can automatically match the optimal storage height according to the size information of the reagent bottle or instrument equipment, eliminating the space waste problem caused by the fixed layer height of the traditional shelf. When the sensor detects out-of-specification materials, the stepper motor can link multiple groups of pallets to lift collaboratively to expand the effective volume of a single layer.

[0020] 2. The present invention uses an intelligent recognition component and a camera to collect the video stream of the operation area. After the visual data acquisition and preprocessing module eliminates the wide-angle distortion, the material recognition and status detection module integrates multi-spectral imaging to identify the material label. When the hand motion tracking and intention recognition module detects that the experimenter takes the material by mistake, the mechanical locking control unit immediately locks the transparent sealed door and synchronously triggers the AR projection positioning unit to project the correct material position light spot, thereby reducing the wrong taking rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] In order to more clearly illustrate the technical solution of the embodiment of the present invention, the drawings required for describing the embodiment are briefly introduced below.

[0022] Figure 1 A three-dimensional diagram of a laboratory material storage and retrieval shelf with AI recognition and its use method;

[0023] Figure 2 A side view of a laboratory material storage and retrieval shelf with AI recognition and a method of using the shelf;

[0024] Figure 3 A schematic diagram of a storage space adjustment component in a laboratory material storage and retrieval shelf with AI recognition and a method for using the shelf;

[0025] Figure 4 A cross-sectional view of a laboratory material storage and retrieval shelf with AI recognition and a method for using the shelf;

[0026] Figure 5 A cross-sectional view of a control shell in a laboratory material storage and retrieval shelf with AI recognition and a method for using the same;

[0027] Figure 6 An exploded diagram of a temperature control mechanism in a laboratory material storage and retrieval shelf with AI recognition and a method for using the shelf;

[0028] Figure 7 A schematic diagram of the controller system in a laboratory material storage and retrieval shelf with AI recognition and a method for using the shelf;

[0029] Figure 8Schematic diagram of the system principle of the visual data acquisition and preprocessing module, hand movement tracking and intention recognition module, material recognition and status detection module, and real-time feedback and error prevention execution module in a laboratory material access shelf with AI recognition and its usage method;

[0030] Figure 9 Schematic diagram of the system principle of the adaptive learning and system optimization module, environment perception and anomaly monitoring module, and interactive guidance module in a laboratory material access shelf with AI recognition and its usage method.

[0031] In the attached drawings: 1. Frame body; 2. Fixed frame; 3. Transparent sealed door; 4. Temperature control mechanism; 5. Storage space adjustment component; 51. Control shell; 52. Stepper motor; 53. Screw rod; 54. Threaded sleeve; 55. Support plate; 56. Positioning plate; 6. Intelligent recognition component; 61. Controller; 62. Shooting mechanism; 611. Central processing unit; 612. Visual data acquisition and preprocessing module; 613. Hand movement tracking and intention recognition module; 614. Material recognition and status detection module; 615. Real-time feedback and error prevention execution module; 616. Adaptive learning and system optimization module; 617. Environment perception and anomaly monitoring module; 618. Interactive guidance module; 6161. Incremental feature library update unit; 6162. Light adaptation switching unit; 6171. Temperature and humidity anomaly detection unit; 6172. Liquid leakage recognition unit; 6181. AR projection positioning unit; 6182. Voice command parsing unit; 6121. Wide-angle distortion correction unit; 6122. Dynamic ROI segmentation unit; 6131. Multi-joint movement analysis unit; 6132. Temporal action classification unit; 6141. Multi-spectral feature extraction unit; 6142. Three-dimensional space verification unit; 6151. Multi-modal alarm linkage unit; 6152. Mechanical locking control unit; 41. Exhaust hood; 42. Exhaust fan; 43. Heating pipe; 44. Mesh plate; 7. Human proximity sensor; 621. Linear motor; 622. Mounting frame; 623. Camera. Detailed implementation manners

[0032] Next, the technical solutions in the embodiments of the present invention will be described with reference to the accompanying drawings in the embodiments of the present invention. The described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments.

[0033] Embodiment 1

[0034] Please refer to Figures 1-9, the present invention is a laboratory material access shelf with AI recognition, including a frame body 1, a storage space adjustment component 5 and an intelligent recognition component 6. A fixed frame 2 is fixedly connected to the right side of the inner cavity of the frame body 1. Transparent sealing doors 3 are movably connected to both sides of the front side of the frame body 1. A temperature control mechanism 4 is fixedly connected to the top of the frame body 1. The storage space adjustment component 5 includes a control housing 51, the rear side of the control housing 51 is fixedly connected to the inner wall of the frame body 1, a stepping motor 52 is fixedly connected to the rear side of the inner cavity of the control housing 51, screw rods 53 are fixedly connected to the top and bottom of the output end of the stepping motor 52, a threaded sleeve 54 is threadedly connected to the surface of the screw rod 53, the front side of the threaded sleeve 54 penetrates through the control housing 51 and is fixedly connected to a support plate 55, and a positioning plate 56 is fixedly connected to the front side of the control housing 51. The intelligent recognition component 6 includes a controller 61, the rear side of the controller 61 is fixedly connected to the transparent sealing door 3, and shooting mechanisms 62 are fixedly connected to both sides of the frame body 1. The controller 61 includes a central processing unit 611, and the central processing unit 611 is bidirectionally electrically connected to a visual data acquisition and preprocessing module 612, a hand movement tracking and intention recognition module 613, a material recognition and status detection module 614, and a real-time feedback and error prevention execution module 615.

[0035] Specifically: The fixed frame 2 can facilitate the placement of materials. The transparent sealing door 3 can facilitate the experimenter to observe the materials placed inside the frame body 1. The stepping motor 52 can cooperate with the screw rod 53 to control the use height of the threaded sleeve 54 and the support plate 55. The support plate 55 can cooperate with the positioning plate 56 to control the storage spacing, so that it can freely adjust the storage space according to the size of the stored materials. The controller 61 can automatically record the access and storage of materials in and out of the warehouse, and cooperate with the shooting mechanism 62 to take pictures and record the pick-up and delivery personnel. The visual data acquisition and preprocessing module 612 uses image correction and dynamic region segmentation technology to eliminate environmental interference and focus on the core operation area, providing a high-quality input source for subsequent analysis. The hand movement tracking and intention recognition module 613 is based on bone key point tracking and behavior chain analysis, accurately captures operation actions and predicts behavior intentions, and effectively identifies abnormal operation modes. The material recognition and status detection module 614 integrates multi-spectral imaging and three-dimensional reconstruction technology to break through the problem of identifying reflective objects, and real-time monitors the spatial posture and stability of materials. The real-time feedback and error prevention execution module 615 integrates multi-modal warning and physical blocking devices to achieve immediate intervention for operation errors and forced isolation of high-risk areas.

[0036] Embodiment 2

[0037] Please refer to Figures 1-6, on the basis of Embodiment 1, the temperature control mechanism 4 includes an exhaust hood 41. The bottom of the exhaust hood 41 is connected to the frame 1. Exhaust fans 42 are connected to both sides of the rear side of the exhaust hood 41. A heating pipe 43 is fixedly connected between both sides of the inner cavity of the exhaust hood 41. A perforated plate 44 is fixedly connected to the bottom of the exhaust hood 41. Humidity sensors are fixedly connected to both sides of the inner cavity of the fixed frame 2. A human proximity sensor 7 is fixedly connected to the front side of the top of the frame 1.

[0038] Specifically: The exhaust fans 42 can convey external air into the exhaust hood 41. The heating pipe 43 is used to heat the air. The perforated plate 44 can prevent external impurities from entering the inside of the frame 1. The humidity sensors can monitor the temperature and humidity inside the frame 1 in real time. When the temperature and humidity are abnormal, the temperature control mechanism 4 can be used to adjust the temperature and humidity inside the frame 1. The human proximity sensor 7 can detect whether an experimenter is approaching. After a person approaches, the shooting mechanism 62 can be controlled to start by the controller 61.

[0039] Embodiment 3

[0040] Please refer to Figure 2 and 5 , on the basis of Embodiment 1, the shooting mechanism 62 includes two linear motors 621. The opposite sides of the two linear motors 621 are fixedly connected to the frame 1. The output ends of the opposite sides of the two linear motors 621 are fixedly connected with mounting frames 622. A camera 623 is fixedly connected to the bottom of the mounting frame 622. Slide rods are fixedly connected to the rear sides of both sides of the frame 1. The surface of the slide rod is slidably connected with a sliding sleeve. The front side of the sliding sleeve is fixedly connected to the mounting frame 622. Limit grooves are opened on both sides of the rear side of the support plate 55. A limiting strip is arranged in the inner cavity of the limit groove. The rear side of the limiting strip is fixedly connected to the inner wall of the frame 1.

[0041] Specifically: The linear motors 621 can control the mounting frames 622 and the camera 623 to move up and down, so that the camera 623 can photograph the types of materials taken by the experimenter. The slide rods and the sliding sleeves can limit the mounting frames 622, so that it can stably control the movement of the camera 623 and prevent it from shaking during the movement. The limit grooves and the limiting strips can limit the support plate 55, so that it can move up and down stably and prevent it from tilting during the adjustment process.

[0042] Embodiment 4

[0043] Please refer to Figures 7-9, on the basis of Embodiment 1, the central processing unit 611 is bidirectionally electrically connected to an adaptive learning and system optimization module 616, an environmental perception and anomaly monitoring module 617, and an interactive guidance module 618. The adaptive learning and system optimization module 616 includes an incremental feature library update unit 6161 and a lighting adaptive switching unit 6162. The environmental perception and anomaly monitoring module 617 includes a temperature and humidity anomaly detection unit 6171 and a liquid leakage identification unit 6172. The interactive guidance module 618 includes an AR projection positioning unit 6181 and a voice command parsing unit 6182. The visual data acquisition and preprocessing module 612 includes a wide-angle distortion correction unit 6121 and a dynamic ROI segmentation unit 6122. The hand movement tracking and intention recognition module 613 includes a multi-joint motion analysis unit 6131 and a temporal action classification unit 6132. The material identification and status detection module 614 includes a multi-spectral feature extraction unit 6141 and a three-dimensional space verification unit 6142. The real-time feedback and error prevention execution module 615 includes a multi-modal alarm linkage unit 6151 and a mechanical locking control unit 6152.

[0044] Specifically, the adaptive learning and system optimization module 616 constructs a dynamic model update mechanism to continuously improve the new material recognition ability and environmental adaptability. The environmental perception and anomaly monitoring module 617 realizes all-weather intelligent supervision and leakage warning of the chemical storage environment through thermal imaging and dynamic analysis technologies. The interactive guidance module 618 combines augmented reality projection and voice interaction technologies to provide screenless operation guidance and material positioning services. The incremental feature library update unit 6161 and the light intensity adaptive switching unit 6162 continuously adapt to the new material recognition requirements through incremental feature library updates, and dynamically switch model parameters in combination with the light intensity mapping table to ensure the operation stability in complex environments. The temperature and humidity anomaly detection unit 6171 and the liquid leakage identification unit 6172 realize intelligent monitoring of the temperature and humidity of chemical storage based on thermal imaging technology, and quickly locate the liquid leakage diffusion boundary in combination with the inter-frame difference method and the region growing algorithm. The AR projection positioning unit 6181 and the voice command parsing unit 6182 use AR spot projection technology to achieve millimeter-level operation guidance positioning, and integrate offline voice command parsing to support screenless material retrieval and process guidance. The wide-angle distortion correction unit 6121 and the dynamic ROI segmentation unit 6122 eliminate the distortion of the wide-angle lens through image distortion correction technology, and accurately focus on the operation area in combination with the dynamic region segmentation algorithm, effectively shielding background interferences such as personnel movement, and providing a standardized input for subsequent analysis. The multi-joint motion analysis unit 6131 and the temporal action classification unit 6132 analyze the three-dimensional motion trajectory of the operator based on the modeling of skeletal key points, identify the complete behavior chain of "grab-move-release" through temporal action classification, and synchronously detect the risks of abnormal jitter or deviation from the path. The multi-spectral feature extraction unit 6141 and the three-dimensional space verification unit 6142 fuse visible light and near-infrared spectra to enhance the recognition of reflective surface features, and monitor the spatial posture of materials in real time in combination with the three-dimensional point cloud reconstruction technology to prevent safety hazards caused by stacking imbalance or incorrect bottle orientation. The multi-modal alarm linkage unit 6151 and the mechanical locking control unit 6152 integrate audible and visual graded alarms and electromagnetic locking devices to immediately block incorrect operations, support physical isolation of high-risk product cabinets and biometric recording, and form a forced error correction mechanism.

[0045] A method for using a laboratory material access shelf with AI recognition includes the following steps: S1. When an experimenter approaches the rack 1, the human proximity sensor 7 detects the human signal and triggers the shooting mechanism 62 to start working. The linear motor 621 drives the mounting rack 622 to move along the slide rod, so that the camera 623 is at the best shooting angle. The cameras 623 on both sides of the rack 1 synchronously collect images of the operation area, and after eliminating the image distortion through the wide-angle distortion correction unit 6121, the picture is transmitted to the central processing unit 611.

[0046] S2. When the experimental personnel store materials, the storage space adjustment component 5 automatically adjusts the layer height according to the size of the materials. When the materials are placed on the pallet 55, the camera 623 takes pictures and detects the height of the materials. The stepper motor 52 drives the screw 53 to rotate, so that the screw 53 drives the threaded sleeve 54 to lift to a suitable position. Multiple pallets 55 can be lifted and lowered synchronously through the limit slots and limit bars, realizing the intelligent allocation of the storage space;

[0047] S3. When retrieving materials, the hand movement tracking and intention recognition module 613 captures the operation actions through the camera 623 and compares them with the characteristic data stored in the material recognition and status detection module 614. When a misretrieval behavior is detected, the mechanical locking control unit 6152 immediately locks the transparent sealing door 3. At the same time, the AR projection positioning unit 6181 projects a light spot prompt at the correct material position. During the whole working process, the central processor 611 collects data through various sensors and the camera 623, coordinates the operation of the storage space adjustment component 5 and the intelligent recognition component 6, and realizes the intelligent management and safe access of laboratory materials;

[0048] S4. During the material storage process, the temperature and humidity sensor continuously monitors the internal environment data of the rack 1. When an abnormal temperature is detected, the temperature control mechanism 4 is automatically started. The exhaust fan 42 sucks the external air through the exhaust hood 41, and the heating pipe 43 heats the air flow according to the set temperature. The adjusted air is evenly distributed to the inside of the shelf through the perforated plate 44. If a liquid leakage occurs, the liquid leakage recognition unit 6172 quickly locates the leakage position by analyzing the change of the picture taken by the camera 623.

[0049] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor limit the invention to the specific embodiments described. The present specification selects and specifically describes these embodiments to better explain the principle and practical application of the present invention, so that those skilled in the art in the technical field can understand and utilize the present invention well.

Claims

1. A laboratory material storage and retrieval shelf with AI recognition, comprising a shelf body (1), a storage space adjustment component (5) and an intelligent recognition component (6), characterized in that: A fixed frame (2) is fixedly connected to the right side of the inner cavity of the frame (1), transparent sealing doors (3) are movably connected to both sides of the front side of the frame (1), and a temperature control mechanism (4) is fixedly connected to the top of the frame (1); The storage space adjustment component (5) comprises a control shell (51), the rear side of the control shell (51) is fixedly connected to the inner wall of the frame body (1), the rear side of the inner cavity of the control shell (51) is fixedly connected to a stepper motor (52), the top and bottom of the output end of the stepper motor (52) are fixedly connected to a screw rod (53), the surface of the screw rod (53) is threadedly connected to a threaded sleeve (54), the front side of the threaded sleeve (54) passes through the control shell (51) and is fixedly connected to a support plate (55), and the front side of the control shell (51) is fixedly connected to a positioning plate (56); The intelligent identification component (6) comprises a controller (61), the rear side of the controller (61) is fixedly connected to the transparent sealing door (3), both sides of the frame (1) are fixedly connected with a shooting mechanism (62), the controller (61) comprises a central processing unit (611), and the central processing unit (611) is bidirectionally electrically connected to a visual data acquisition and preprocessing module (612), a hand movement tracking and intention recognition module (613), a material recognition and status detection module (614) and a real-time feedback and error-proofing execution module (615).

2. According to claim 1, a laboratory material storage and retrieval shelf with AI recognition is characterized by: The temperature control mechanism (4) comprises an exhaust hood (41), the bottom of the exhaust hood (41) is connected to the frame (1), both sides of the rear side of the exhaust hood (41) are connected to exhaust fans (42), a heating pipe (43) is fixedly connected between the two sides of the inner cavity of the exhaust hood (41), and a mesh plate (44) is fixedly connected to the bottom of the exhaust hood (41).

3. According to claim 1, a laboratory material storage and retrieval shelf with AI recognition is characterized by: Temperature and humidity sensors are fixedly connected to both sides of the inner cavity of the fixing frame (2), and a human body proximity sensor (7) is fixedly connected to the front side of the top of the frame body (1).

4. According to claim 1, a laboratory material storage and retrieval shelf with AI recognition is characterized by: The shooting mechanism (62) comprises two linear motors (621), the opposite sides of the two linear motors (621) are fixedly connected to the frame (1), the output ends of the opposite sides of the two linear motors (621) are fixedly connected to a mounting frame (622), and the bottom of the mounting frame (622) is fixedly connected to a camera (623).

5. According to claim 1, a laboratory material storage and retrieval shelf with AI recognition, characterized in that: The rear sides of both sides of the frame body (1) are fixedly connected with sliding rods, the surfaces of the sliding rods are slidably connected with sliding sleeves, and the front sides of the sliding sleeves are fixedly connected with the mounting frame (622).

6. The laboratory material storage and retrieval shelf with AI recognition according to claim 1, characterized in that: Limiting grooves are provided on both sides of the rear side of the support plate (55), and the inner cavity of the limiting groove is provided with a limiting strip, and the rear side of the limiting strip is fixedly connected to the inner wall of the frame body (1).

7. The laboratory material storage and retrieval shelf with AI recognition according to claim 1, characterized in that: The central processor (611) is bidirectionally electrically connected to an adaptive learning and system optimization module (616), an environment perception and abnormality monitoring module (617), and an interactive guidance module (618).

8. The laboratory material storage and retrieval shelf with AI recognition according to claim 7, characterized in that: The adaptive learning and system optimization module (616) includes an incremental feature library updating unit (6161) and an adaptive illumination switching unit (6162); the environmental perception and abnormality monitoring module (617) includes a temperature and humidity abnormality detection unit (6171) and a liquid leakage identification unit (6172); and the interactive guidance module (618) includes an AR projection positioning unit (6181) and a voice command parsing unit (6182).

9. The laboratory material storage and retrieval shelf with AI recognition according to claim 1, characterized in that: The visual data acquisition and preprocessing module (612) includes a wide-angle distortion correction unit (6121) and a dynamic ROI segmentation unit (6122); the hand motion tracking and intention recognition module (613) includes a multi-joint motion analysis unit (6131) and a sequential motion classification unit (6132); the material identification and status detection module (614) includes a multi-spectral feature extraction unit (6141) and a stereoscopic space verification unit (6142); and the real-time feedback and error-proofing execution module (615) includes a multi-modal alarm linkage unit (6151) and a mechanical locking control unit (6152).

10. A method for using a laboratory material storage and retrieval shelf with AI recognition according to any one of claims 1 to 9, characterized in that: The steps include: S1. When an experimenter approaches the frame (1), the human body proximity sensor (7) detects a human body signal, triggering the shooting mechanism (62) to start working, and the linear motor (621) drives the mounting frame (622) to move along the slide bar, so that the camera (623) is at the best shooting angle. The cameras (623) on both sides of the frame (1) synchronously collect images of the operation area, and after eliminating image deformation through the wide-angle distortion correction unit (6121), the picture is transmitted to the central processor (611); S2. When the experimenter stores materials, the storage space adjustment component (5) automatically adjusts the layer height according to the size of the materials. When the materials are placed on the pallet (55), the camera (623) takes pictures and detects the height of the materials. The stepper motor (52) drives the screw rod (53) to rotate, so that the screw rod (53) drives the threaded sleeve (54) to rise and fall to a suitable position. Multiple pallets (55) can be kept rising and falling synchronously through the limit grooves and the limit bars, thereby realizing intelligent allocation of storage space. S3. When taking materials, the hand motion tracking and intention recognition module (613) captures the operation action through the camera (623) and compares it with the characteristic data stored in the material recognition and status detection module (614). When the wrong taking behavior is detected, the mechanical locking control unit (6152) immediately locks the transparent sealing door (3), and the AR projection positioning unit (6181) projects a light spot prompt at the correct material position. During the whole working process, the central processor (611) collects data through various sensors and cameras (623), coordinates the storage space adjustment component (5) and the intelligent recognition component (6) to work together, and realizes the intelligent management and safe access of laboratory materials. S4. During the material storage process, the temperature and humidity sensor continuously monitors the internal environmental data of the rack (1). When an abnormal temperature is detected, the temperature control mechanism (4) automatically starts, the exhaust fan (42) draws in external air through the exhaust hood (41), the heating pipe (43) heats the air flow according to the set temperature, and the regulated air is evenly distributed to the inside of the shelf through the mesh plate (44). If liquid leakage occurs, the liquid leakage identification unit (6172) quickly locates the leakage position by analyzing the changes in the picture taken by the camera (623).

Images