Visual intelligent identification device and method for intelligent warehouse goods classification

Abstract

The invention relates to the technical field of storage devices, in particular to a visual intelligent recognition device and method for intelligent storage goods classification, the visual intelligent recognition device comprises a left cross beam, a right cross beam and a conveying gap arranged between the two cross beams, and the tops of the two cross beams are fixedly provided with a mounting frame; a plurality of image acquisition units and a plurality of illumination adjusting units are detachably mounted on the mounting frame, each image acquisition unit comprises a fixing plate and a high-definition camera mounted on the front side surface of the fixing plate, the fixing plate is detachably connected with the mounting frame, and each illumination adjusting unit comprises a limiting plate detachably connected with the mounting frame. A gooseneck pipe is detachably installed on the front side face of the limiting plate, a lamp holder is fixedly installed at the front end of the gooseneck pipe, and an LED lamp is fixedly installed on the front side face of the lamp holder. And a data processing unit for data processing is also arranged on one cross beam. Position adjustment operation is facilitated, operation is simple, and use is convenient.

Description

Technical Field

[0001] The present invention relates to the technical field of warehousing devices, and in particular to a visual intelligent recognition device and method for intelligent warehousing cargo classification. Background Art

[0002] In the modern logistics industry, intelligent warehousing systems have become a key technology for improving logistics efficiency and reducing operating costs due to their high efficiency and precision. Cargo classification is one of the core links of intelligent warehousing systems. Its efficiency and accuracy directly affect the operational effectiveness of the entire warehousing system. At present, cargo classification in intelligent warehousing mainly adopts traditional barcode scanning, RFID recognition and other methods; barcode scanning relies on close-range scanning by manual or robotic arms. When the barcode is damaged or blocked, the recognition success rate is greatly reduced, and it is difficult to obtain detailed information such as the appearance and size of the goods; although RFID recognition can achieve non-contact recognition, the label cost is high, and there are problems such as signal interference and difficulty in reusing labels. At the same time, it is also unable to effectively identify the appearance characteristics of the goods, and it is difficult to meet the diverse cargo classification needs in complex warehousing environments.

[0003] In recent years, cargo classification methods based on visual recognition technology have gradually emerged. Cargo images are collected by cameras, and cargo classification is achieved using image processing and pattern recognition algorithms. This type of visual intelligent recognition device can collect cargo images from multiple angles by setting up multiple high-definition cameras of different types, and adjust the lighting conditions in real time with the lighting adjustment unit to obtain comprehensive and clear cargo image information. Edge computing devices are then used as data processing units with built-in efficient deep learning image recognition algorithms to process image data locally in real time; data interaction and collaborative control are carried out with other equipment in the intelligent warehousing system through communication units, realizing seamless connection between cargo classification and warehousing operations such as transportation, sorting, and storage, optimizing the operating process of the intelligent warehousing system, improving the automation level and operating efficiency of the entire warehousing system, and reducing manual intervention and operating costs.

[0004] The utility model patent with authorization announcement number CN213504274U discloses an intelligent bulk picking device, including a bidirectional telescopic fork and a picking robot. The bidirectional telescopic fork includes a telescopic drive, a fork tine seat, an intermediate sliding fork, and a top sliding fork. The telescopic drive is fixedly connected to the fork tine seat, the intermediate sliding fork is slidably connected to the fork tine seat, and the top sliding fork is slidably connected to the intermediate sliding fork. The telescopic drive can drive the intermediate sliding fork and the top sliding fork to telescopically move and transport turnover boxes. The picking robot includes a multi-axis robotic arm and a visual scanner and picking actuator connected to the multi-axis robotic arm. The multi-axis robotic arm can drive the visual scanner and picking actuator to move above the turnover box. The visual scanner can scan the goods in the turnover box and identify the location. The picking actuator grabs the corresponding goods based on the identified location. This utility model intelligent bulk picking device can quickly pick specific goods and has high picking efficiency.

[0005] Patent application number CN114261678A discloses a three-dimensional warehouse with intelligent item sorting capabilities. The warehouse comprises a sorting device and multiple sets of shelves with movable storage compartments arranged around the sorting device. The sorting device includes a sorting pipe, a material receiving portion, a visual recognition component, a discharge channel, a lifting drive, a rotation drive, and a vibration component. The axis of the sorting pipe is vertically arranged. The material receiving portion is movably arranged within the sorting pipe, with its top surface tilted. The visual recognition component is used to visually identify the goods. The discharge channel is arranged around the axis of the sorting pipe. The lifting drive and the rotation drive are used to drive the material receiving portion to rise and fall and rotate, respectively. The vibration component is used to cause the material receiving portion to intermittently vibrate during the process of rising and falling along the axis of the sorting pipe. This application has a simple structure and is easy to operate, effectively improving sorting efficiency while reducing costs.

[0006] While the aforementioned technical solutions offer corresponding advantages, most current visual intelligent recognition devices with cameras and light sources have fixed mounting positions during use, preventing easy adjustment of these locations. Furthermore, if screws are used for installation, corresponding through-holes must be reserved. Subsequent changes to the mounting position make installation and removal difficult and inconvenient for the user. Therefore, we propose a visual intelligent recognition device and method for intelligent warehousing cargo classification. Summary of the Invention

[0007] The purpose of the present invention is to provide a visual intelligent recognition device and method for intelligent warehouse cargo classification to solve the defects mentioned in the above background technology.

[0008] To achieve the above object, the present invention provides the following technical solutions: One of the objectives of the present invention is to provide a visual intelligent recognition device for intelligent warehouse cargo classification, comprising two mutually symmetrical left and right beams and a conveying gap arranged between the two beams, a mounting frame is fixedly installed on the top of the two beams, and a plurality of image acquisition units and a plurality of light adjustment units are detachably installed on the mounting frame, the image acquisition unit comprises a fixing plate and a high-definition camera installed on the front side of the fixing plate, the fixing plate and the mounting frame are detachably connected, the light adjustment unit comprises a limiting plate detachably connected to the mounting frame, a gooseneck tube is detachably installed on the front side of the limiting plate, a lamp holder is fixedly installed on the front end of the gooseneck tube, and an LED lamp is fixedly installed on the front side of the lamp holder; a data processing unit for data processing is also provided on one of the beams.

[0009] Preferably, a light sensor is fixedly mounted on the inner side surface of the beam, and the light sensor is used to detect the ambient light intensity.

[0010] Preferably, a conveyor belt is provided at the conveying gap, and the conveyor belt is rotated by corresponding rollers.

[0011] Preferably, a plurality of through holes are provided on each side plate of the mounting frame, and two mutually symmetrical first studs are fixedly mounted on the rear side surface of the fixing plate. The first studs pass through the through holes, and the other ends of the first studs are threadedly connected to first screw sleeves. Preferably, the fixing plate and the first screw sleeve are respectively against the inner and outer side surfaces of the mounting frame, and a first knob is fixedly mounted on the first screw sleeve; The above two settings facilitate the installation and removal and position adjustment operations of the high-definition camera.

[0012] Preferably, two second studs symmetrical to each other are fixedly mounted on the back of the limiting plate, the second studs pass through the through holes, and a second screw sleeve is threadedly connected to the second studs; Preferably, the limiting plate and the second screw sleeve are respectively against the inner and outer side surfaces of the mounting frame, and a second knob is fixedly mounted on the second screw sleeve; The above two settings facilitate the installation and removal of the LED lamp and the position adjustment operations.

[0013] Preferably, a wiring hole is provided at the center of the limiting plate, a hollow threaded joint is fixedly installed at the center of the front side surface of the limiting plate, a hard pipe is fixedly installed at the rear end of the gooseneck tube, a third knob is fixedly installed on the hard pipe, a hollow stud is fixedly installed at the rear end of the hard pipe, and the hollow stud is threadedly connected to the hollow threaded joint; This setting facilitates the installation, removal and replacement of LED lights.

[0014] Preferably, the data processing unit includes a data processing host, a communication module is installed on the data processing host, a vertically arranged support rod is fixedly installed at the bottom of the shell of the data processing host, a fixed base is fixedly installed at the bottom end of the support rod, and the fixed base is fixedly installed on the corresponding beam.

[0015] A second object of the present invention is to provide a visual intelligent recognition method for intelligent warehousing cargo classification, including the above-mentioned visual intelligent recognition device for intelligent warehousing cargo classification, specifically comprising the following steps: S1: The cargo enters the recognition area of ​​the visual intelligent recognition device through the conveyor belt arranged in the conveyor gap. At this time, the light sensor installed on the inner side of the beam detects the ambient light intensity in real time and transmits the detection data to the data processing host in the data processing unit; S2: Based on the received light intensity data, the data processing host sends instructions to the light adjustment unit through the communication module to control the brightness and color temperature of the LED light. The operator adjusts the position of the limit plate by rotating the second knob, adjusts the position of the gooseneck by turning the third knob, and bends the gooseneck to adjust the bending angle, thereby precisely adjusting the lighting conditions on the cargo surface. At the same time, the data processing host controls the high-definition camera of the image acquisition unit. The operator adjusts the installation angle and position of the high-definition camera using the first knob on the fixed plate to capture cargo images from different angles. S3: The image acquisition unit transmits the captured cargo images to the data processing unit. The data processing host uses a built-in deep learning image recognition algorithm to preprocess, extract features, and classify the images. First, the images are preprocessed by noise reduction and enhancement to improve image quality. Then, the algorithm is used to extract feature information such as the cargo's shape, color, texture, and logo. Finally, the extracted feature information is matched with pre-set classification rules to determine the cargo's category. S4: The data processing host transmits the recognition results to the sorting robot equipment in the intelligent warehousing system through the communication module. The sorting robot moves the goods to the designated storage area to complete the goods classification operation.

[0016] Compared with the prior art, the present invention has the following beneficial effects: 1. The present invention realizes convenient installation and removal and position adjustment of the high-definition camera in the image acquisition unit by providing multiple through holes on the mounting frame, cooperating with the first stud, the first screw sleeve and the first knob of the fixing plate. The operator can flexibly adjust the position of the high-definition camera on the mounting frame according to actual needs without reserving mounting holes. This solves the problems of fixed installation position and inconvenient adjustment of traditional devices, achieves the effect of quickly adapting to the shooting needs of different goods and improving the flexibility of image acquisition.

[0017] 2. The present invention realizes the free adjustment of the installation position of the LED lamp in the light adjustment unit and the rapid assembly and disassembly through the cooperation of the second stud, second screw sleeve and second knob of the limit plate and the through hole. This structural design makes it easy to change the installation position and angle of the LED lamp when facing different lighting conditions and cargo types, ensuring uniform lighting on the surface of the cargo, thereby improving the quality of image acquisition and ensuring recognition accuracy.

[0018] 3. The present invention realizes convenient assembly and disassembly and angle fine-tuning of LED lamps through the coordination of the wiring holes and hollow threaded joints on the limit plate, the hard tube, the hollow stud and the third knob. This not only facilitates the replacement and maintenance of the lamp, but also the gooseneck tube can be bent and shaped at will, which can adjust the LED lamp to the ideal position and angle, accurately fill in the light, and meet the diverse lighting needs in complex storage environments, thereby improving the applicability of the visual recognition device. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the overall structure of the present invention; Figure 2 It is a schematic diagram of the explosion structure of the present invention; Figure 3 Schematic diagram of the exploded structure of the image acquisition unit of the present invention; Figure 4 This is one of the exploded structural diagrams of the light adjustment unit of the present invention; Figure 5 This is the second exploded structural diagram of the light adjustment unit of the present invention; Figure 6 Schematic diagram of the structure of the data processing unit of the present invention; The meaning of each number in the figure is: 1. Beam; 10. Conveyor gap; 11. Light sensor; 12. Conveyor belt; 13. Mounting bracket; 131. Through hole; 2. Image acquisition unit; 20. Fixing plate; 21. High-definition camera; 22. First stud; 23. First screw sleeve; 24. First knob; 3. Lighting adjustment unit; 30. Limiting plate; 301. Wiring hole; 31. Second stud; 32. Second screw sleeve; 33. Second knob; 34. Hollow threaded connector; 35. Hard tube; 351. Third knob; 352. Hollow stud; 36. Gooseneck tube; 37. Lamp holder; 371. LED lamp; 4. Data processing unit; 40. Data processing host; 41. Communication module; 42. Support rod; 43. Fixed base. DETAILED DESCRIPTION

[0020] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0021] In the description of the present invention, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", "clockwise", "counterclockwise" and the like to indicate orientations or positional relationships based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as limiting the present invention.

[0022] See also Figures 1-6 The present invention provides a technical solution: a visual intelligent recognition device for intelligent warehousing cargo classification, comprising two mutually symmetrical left and right beams 1 and a conveying gap 10 arranged between the two beams 1, a mounting frame 13 is fixedly installed on the top of the two beams 1, and a plurality of image acquisition units 2 and a plurality of light adjustment units 3 are detachably mounted on the mounting frame 13, the image acquisition unit 2 comprises a fixing plate 20 and a high-definition camera 21 mounted on the front side of the fixing plate 20, the shooting angles of the plurality of high-definition cameras 21 cover the entire conveying area of ​​the cargo conveying line, and are used to collect image information of the cargo from different angles; the fixing plate 20 and the mounting frame 13 are detachably connected, and a plurality of through holes 131 are provided on each side plate of the mounting frame 13, Two symmetrical first studs 22 are fixedly installed on the rear side surface of the fixing plate 20. The first stud 22 passes through the through hole 131. The other end of the first stud 22 is threadedly connected to the first screw sleeve 23. The fixing plate 20 and the first screw sleeve 23 are respectively against the inner and outer side surfaces of the mounting bracket 13. A first knob 24 is fixedly installed on the first screw sleeve 23, which realizes convenient loading and unloading and position adjustment of the high-definition camera 21 in the image acquisition unit 2. The operator can flexibly adjust the position of the high-definition camera 21 on the mounting bracket 13 according to actual needs without reserving mounting holes, which solves the problems of fixed installation position and inconvenient adjustment of traditional devices, and achieves the effect of quickly adapting to the shooting needs of different goods and improving the flexibility of image acquisition.

[0023] In this embodiment, the light adjustment unit 3 includes a stopper plate 30 detachably connected to the mounting frame 13. A gooseneck tube 36 is detachably mounted on the front side of the stopper plate 30. A lamp holder 37 is fixedly mounted on the front end of the gooseneck tube 36. An LED lamp 371 is fixedly mounted on the front side of the lamp holder 37. Two symmetrical second studs 31 are fixedly mounted on the back side of the stopper plate 30. The second studs 31 extend through the through-hole 131 and are threadedly connected to the second studs 31. The stopper plate 30 and the second screw sleeve 32 respectively abut against the inner and outer sides of the mounting frame 13. A second knob 33 is fixedly mounted on the second screw sleeve 32, enabling free adjustment of the installation position of the LED lamp 371 in the light adjustment unit 3 and quick installation and removal. This structural design allows the installation position and angle of the LED lamp 371 to be easily adjusted to accommodate different lighting conditions and cargo types, ensuring uniform illumination of the cargo surface, thereby improving image acquisition quality and ensuring recognition accuracy.

[0024] In this embodiment, a light sensor 11 is fixedly installed on the inner side of the beam 1. The light sensor 11 is used to detect the ambient light intensity in real time and transmit the detection data to the data processing unit 4. The data processing unit 4 controls the brightness, color temperature and other parameters of the adjustable LED light source 371 according to the light intensity data to adjust the lighting conditions on the surface of the goods.

[0025] like Figure 1 and Figure 2 As shown, a conveyor belt 12 is provided at the conveying gap 10. The conveyor belt 12 is rotated by corresponding rollers. The conveyor belt 12 is a conveyor line for intelligent storage goods. The conveyor belt 12 rotates to transport goods.

[0026] In addition, a wiring hole 301 is provided at the center position of the limiting plate 30, a hollow threaded joint 34 is fixedly installed at the center position of the front side of the limiting plate 30, a hard tube 35 is fixedly installed at the rear end of the gooseneck tube 36, a third knob 351 is fixedly installed on the hard tube 35, and a hollow stud 352 is fixedly installed at the rear end of the hard tube 35. The hollow stud 352 is threadedly connected to the hollow threaded joint 34, which realizes the convenient loading and unloading and angle fine-tuning of the LED lamp 371, which not only facilitates the replacement and maintenance of the lamp, but also the gooseneck tube 36 can be bent and shaped at will, and the LED lamp 371 can be adjusted to the ideal position and angle, accurately fill in the light, so as to meet the diverse lighting needs in complex storage environments and improve the applicability of the visual recognition device.

[0027] It is worth noting that a data processing unit 4 for data processing is also provided on one of the beams 1. The data processing unit 4 includes a data processing host 40. A communication module 41 is installed on the data processing host 40. A vertically arranged support rod 42 is fixedly installed at the bottom of the shell of the data processing host 40. A fixed base 43 is fixedly installed at the bottom end of the support rod 42. The fixed base 43 is fixedly installed on the corresponding beam 1 by a plurality of fastening screws, which is convenient for fixed installation and disassembly operations.

[0028] It is worth noting that the data processing unit 4 is electrically connected to the image acquisition unit 2, the light adjustment unit 3 and the communication module 41. The data processing unit 4 has a built-in deep learning image recognition algorithm for processing and analyzing the collected cargo images, extracting the characteristic information of the cargo, and classifying and identifying the cargo according to preset classification rules; the communication module 41 is used to interact with other equipment in the intelligent warehousing system, such as conveyor line controllers, sorting robots, etc., and transmit the recognition results to related equipment to achieve coordinated control of cargo classification and warehousing operations.

[0029] In this embodiment, the high-definition camera 21 includes a wide-angle camera and a macro camera. The wide-angle camera is used to capture images of the overall appearance of the goods, and the macro camera is used to capture images of detailed features of the goods, such as product logos and textures, to obtain more comprehensive image information of the goods. The LED lamp 371 with an adjustable light source is a dimmable lamp that achieves brightness adjustment through PWM dimming technology and color temperature adjustment by adjusting the mixing ratio of LED lamp beads of different colors, which can quickly and accurately adjust the lighting conditions. The data processing unit 4 uses edge computing equipment to locally process and analyze the collected image data in real time, reducing data transmission delays and improving recognition speed and real-time performance.

[0030] Finally, it should be noted that the components involved in the present invention, such as the high-definition camera 21, LED light 371, data processing host 40 and communication module 41, are all universal standard parts or components known to those skilled in the art. Their structures and principles are known to those skilled in the art through technical manuals or through conventional experimental methods. In the idle space of this device, all the above-mentioned electrical components, which refer to power elements, electrical components, and adapted controllers and power supplies, are connected through wires. The specific connection means should refer to the working principle of the present invention. The electrical connection between each electrical component is completed in a sequential working order, and the detailed connection means are all well-known technologies in the art.

[0031] In addition, this embodiment also provides a visual intelligent recognition method for intelligent warehousing cargo classification, which specifically includes the following steps: S1: Goods enter the recognition area of ​​the visual intelligent recognition device through the conveyor belt 12 provided at the conveying gap 10. At this time, the light sensor 11 installed on the inner side of the beam 1 detects the ambient light intensity in real time and transmits the detection data to the data processing host 40 in the data processing unit 4; S2: The data processing host 40 sends instructions to the light adjustment unit 3 via the communication module 41 based on the received light intensity data to control the brightness and color temperature of the LED light 371. The operator removes the limit plate 30 by rotating the second knob 33, then inserts the second stud 31 through the corresponding through hole 131 to adjust the position of the limit plate 30. After adjustment, the second knob 33 is tightened to adjust the position of the gooseneck 36. The gooseneck 36 is bent to adjust the bending angle, thereby precisely adjusting the lighting conditions on the cargo surface. At the same time, the data processing host 40 controls the high-definition camera 21 of the image acquisition unit 2. The operator adjusts the installation angle and position of the high-definition camera 21 using the first knob 24 on the fixing plate 20 to enable it to capture images of the cargo from different angles. S3: The image acquisition unit 2 transmits the captured cargo image to the data processing unit 4. The data processing host 40 uses a built-in deep learning image recognition algorithm to preprocess, extract features, and classify the image. First, the image is preprocessed by noise reduction and enhancement to improve image quality. Then, the algorithm is used to extract feature information such as the cargo's shape, color, texture, and logo. Finally, the extracted feature information is matched with preset classification rules to determine the cargo's category. S4: The data processing host 40 transmits the recognition result to the sorting robot equipment in the intelligent warehousing system through the communication module 41. The sorting robot moves the goods to the designated storage area to complete the goods classification operation.

[0032] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions are merely preferred examples of the present invention and are not intended to limit the present invention. Various changes and improvements may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and improvements fall within the scope of the present invention. The scope of protection claimed in the present invention is defined by the appended claims and their equivalents.

Claims

1. A visual intelligent recognition device for intelligent warehousing cargo classification, comprising two mutually symmetrical left and right beams (1) and a conveying gap (10) arranged between the two beams (1), characterized in that: A mounting frame (13) is fixedly mounted on the top of the two crossbeams (1), and a plurality of image acquisition units (2) and a plurality of light adjustment units (3) are detachably mounted on the mounting frame (13), wherein the image acquisition unit (2) comprises a fixing plate (20) and a high-definition camera (21) mounted on the front side of the fixing plate (20), the fixing plate (20) is detachably connected to the mounting frame (13), and the light adjustment unit (3) comprises a limiting plate (30) detachably connected to the mounting frame (13), a gooseneck tube (36) is detachably mounted on the front side of the limiting plate (30), a lamp holder (37) is fixedly mounted on the front end of the gooseneck tube (36), and an LED lamp (371) is fixedly mounted on the front side of the lamp holder (37); a data processing unit (4) for data processing is also provided on one of the crossbeams (1).

2. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 1 is characterized by: A light sensor (11) is fixedly mounted on the inner side surface of the crossbeam (1), and the light sensor (11) is used to detect ambient light intensity.

3. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 1 is characterized by: A conveyor belt (12) is provided at the conveying gap (10), and the conveyor belt (12) is driven to rotate by a corresponding roller shaft.

4. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 1 is characterized by: A plurality of through holes (131) are provided on each side plate of the mounting frame (13), and two mutually symmetrical first studs (22) are fixedly mounted on the rear side surface of the fixing plate (20), the first studs (22) passing through the through holes (131), and the other end of the first stud (22) is threadedly connected to a first screw sleeve (23).

5. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 4 is characterized in that: The fixing plate (20) and the first screw sleeve (23) respectively abut against the inner and outer side surfaces of the mounting frame (13), and a first knob (24) is fixedly mounted on the first screw sleeve (23).

6. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 5, characterized in that: Two upper and lower symmetrical second studs (31) are fixedly mounted on the back of the limiting plate (30), the second studs (31) pass through the through holes (131), and the second screw sleeves (32) are threadedly connected to the second studs (31).

7. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 6, characterized in that: The limiting plate (30) and the second screw sleeve (32) respectively abut against the inner and outer side surfaces of the mounting frame (13), and a second knob (33) is fixedly mounted on the second screw sleeve (32).

8. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 1 is characterized by: A wiring hole (301) is provided at the center of the limiting plate (30), a hollow threaded joint (34) is fixedly installed at the center of the front side of the limiting plate (30), a hard tube (35) is fixedly installed at the rear end of the gooseneck tube (36), a third knob (351) is fixedly installed on the hard tube (35), a hollow stud (352) is fixedly installed at the rear end of the hard tube (35), and the hollow stud (352) is threadedly connected to the hollow threaded joint (34).

9. The visual intelligent recognition device for intelligent warehousing cargo classification according to claim 1, characterized in that: The data processing unit (4) comprises a data processing host (40), a communication module (41) is installed on the data processing host (40), a vertically arranged support rod (42) is fixedly installed at the bottom of the housing of the data processing host (40), a fixed base (43) is fixedly installed at the bottom end of the support rod (42), and the fixed base (43) is fixedly installed on the corresponding crossbeam (1).

10. A method for intelligent visual recognition of goods for intelligent warehousing classification, comprising the device for intelligent visual recognition of goods for intelligent warehousing classification according to any one of claims 1 to 9, characterized in that: The specific steps include: S1: The goods enter the recognition area of ​​the visual intelligent recognition device through the conveyor belt (12) arranged at the conveying gap (10). At this time, the light sensor (11) installed on the inner side of the beam (1) detects the ambient light intensity in real time and transmits the detection data to the data processing host (40) in the data processing unit (4); S2: The data processing host (40) sends a command to the light adjustment unit (3) through the communication module (41) based on the received light intensity data to control the brightness and color temperature of the LED light (371); the operator adjusts the position of the limit plate (30) by rotating the second knob (33), adjusts the position of the gooseneck (36) by rotating the third knob (351), and bends the gooseneck (36) to adjust the bending angle, thereby accurately adjusting the light conditions on the surface of the goods; at the same time, the data processing host (40) controls the high-definition camera (21) of the image acquisition unit (2), and the operator adjusts the installation angle and position of the high-definition camera (21) through the first knob (24) on the fixing plate (20) so that it can capture images of the goods from different angles; S3: The image acquisition unit (2) transmits the collected cargo image to the data processing unit (4), and the data processing host (40) uses the built-in deep learning image recognition algorithm to preprocess, extract features, and classify and identify the image; first, the image is preprocessed by noise reduction, enhancement, and other operations to improve the image quality; then, the algorithm is used to extract the cargo's shape, color, texture, logo, and other feature information; finally, the extracted feature information is matched with the preset classification rules to determine the category of the cargo; S4: The data processing host (40) transmits the recognition result to the sorting robot device in the intelligent warehousing system through the communication module (41). The sorting robot moves the goods to the designated storage area to complete the goods sorting operation.

Images