An AI vision-based network filter automatic foot detection device

Abstract

The utility model relates to AI vision detection technical field, and specifically is a kind of network filter automatic winding foot detection device based on AI vision, including support column and AI vision identification host computer, the support column is fixedly installed with first support plate and second support plate, the first support plate is provided with through slot, the first support plate below one side is fixedly installed with several coaxial nested annular LED light source, at least one item in the luminous radius, luminous color and illumination angle of different annular LED light source is different, the second support plate is fixedly installed with industrial camera, the lens of the industrial camera is directly opposite the through slot and is towards the irradiation area of annular LED light source, the AI vision identification host computer is provided with several wires, and the AI vision identification host computer is electrically connected with industrial camera and each annular LED light source respectively by different wires, solve the problem that detection adaptability is poor, misjudgment rate is high under traditional single light source environment.

Description

Technical Field

[0001] This utility model relates to the field of AI vision inspection technology, and more specifically, to an automatic foot detection device for network filters based on AI vision. Background Technology

[0002] Network filter lead winding inspection is a quality control step in the network filter production process that uses a vision inspection device to wind the enameled wire ends of the magnetic ring around the inside of the lead. This inspection process focuses on monitoring key indicators such as lead shape, spacing, winding quality (e.g., coil forming, whether there are broken wires), and the welding strength between the lead and the wire. For example, intelligent algorithms are used to process visual images to identify whether the lead is skewed or whether the spacing meets the standards, ensuring the quality of the network filter leads, improving product performance and stability, and ensuring that the filter can accurately filter out specific frequency signals in the power line, laying a solid foundation for the stable operation of network communication equipment.

[0003] For example, Chinese patent publication CN202120120750.8 discloses an AI-based visual inspection device, including a frame with a motor housing at the bottom. Side baffles are located at both ends of the inner radial side of the frame, with a base plate between the two side baffles. A detection chamber is located at the top of the base plate, and a top plate is located at the top of the detection chamber. A set of lifting and adjusting components is located on each side of the base plate. An industrial camera is located at the center of the bottom of the top plate. An AI visual recognition host electrically connected to the industrial camera is located on the side of the detection chamber away from the industrial camera. A horizontal plate is located inside the detection chamber, with a ring light source device at the bottom. The overall structure is highly flexible, provides good protection, and is not prone to dust accumulation, exhibiting good dustproof performance. This allows the device to be used immediately upon startup, offering more flexible operation and ensuring normal use.

[0004] However, in the above-mentioned technical solutions, the automatic inspection equipment based on machine vision on the market usually adopts a single type of light source illumination system, which is difficult to adapt to the detection needs of different defect types. Since the network filter has various defect forms such as warping, deformation, and short circuit, the characteristics of various defects are significantly different under different lighting conditions. A fixed light source cannot provide the best imaging effect for all defects at the same time. Especially when inspecting highly reflective metal pins, a single light source is prone to strong reflection or shadows, which can cover up the true defect characteristics, resulting in a high rate of misjudgment and missed detection, and cannot meet the requirements of high-quality production. Utility Model Content

[0005] The main purpose of this invention is to provide an automatic detection device for network filters based on AI vision. This device can effectively solve the problem that current automatic detection equipment based on machine vision on the market typically uses a single type of light source illumination system, which is difficult to adapt to the detection needs of different defect types. Since network filters have various defect forms such as warping, deformation, and short circuits, the characteristics of various defects differ significantly under different lighting conditions. A fixed light source cannot provide the best imaging effect for all defects at the same time. Especially when detecting highly reflective metal pins, a single light source is prone to strong reflection or shadows, which can obscure the true defect characteristics, leading to high misjudgment and missed detection rates, and failing to meet the requirements of high-quality production.

[0006] To achieve the above objectives, the technical solution adopted by this utility model is as follows: an automatic foot detection device for network filters based on AI vision, comprising a support column and an AI vision recognition host. A first support plate and a second support plate are fixedly installed on the support column. A through groove is provided on the first support plate. Several coaxially nested ring LED light sources are fixedly installed on one side below the first support plate. At least one of the luminous radius, luminous color, and illumination angle of the different ring LED light sources is different. An industrial camera is fixedly installed on the second support plate. The lens of the industrial camera faces the through groove and is directed towards the illumination area of ​​the ring LED light sources. Several wires are provided on the AI ​​vision recognition host. The AI ​​vision recognition host is electrically connected to the industrial camera and each ring LED light source through different wires. The AI ​​vision recognition host is configured to independently control the on / off state and brightness of each ring LED light source.

[0007] Preferably, the shape and size of the through slot are adapted to the illumination area of ​​the annular LED light source to define the detection area.

[0008] Preferably, the first support plate includes a support plate base, and a liquid cooling channel is provided inside the support plate base. The two ends of the liquid cooling channel are respectively provided with a liquid inlet and a liquid outlet.

[0009] Preferably, a diffuse scattering glass cover is fixedly installed below the support plate base.

[0010] Preferably, the liquid inlet, liquid outlet, and liquid cooling channel are connected.

[0011] Preferably, the diffuse scattering glass cover covers the outside of the annular LED light source.

[0012] Compared with the prior art, the present invention has the following beneficial effects:

[0013] (1) In this invention, the network filter element to be tested is placed in the testing station under the ring LED light source. The AI ​​vision recognition host precisely controls the opening and closing state and brightness of each ring LED light source through independent wires according to the preset testing program, so that light sources with different radii, colors and angles can be turned on individually or in combination to form the optimal lighting scheme for specific defect features. The industrial camera collects the image of the illuminated element through the through slot and transmits the image data to the AI ​​vision recognition host in real time. The image processing algorithm in the host analyzes and processes the image, and judges whether there are defects such as warping, deformation and short circuits around the foot through feature extraction and pattern recognition, which solves the problem of poor detection adaptability and high misjudgment rate in the traditional single light source environment. Attached Figure Description

[0014] Figure 1 This is a schematic diagram of the structure of an automatic foot-circling detection device for network filters based on AI vision according to this utility model;

[0015] Figure 2 This is a schematic diagram of the internal structure of an automatic foot-circling detection device for network filters based on AI vision according to this utility model;

[0016] Figure 3 This is a schematic diagram of the structure of the first support plate in an AI vision-based network filter automatic foot detection device of this utility model.

[0017] In the diagram: 1. Support column; 2. AI visual recognition host; 3. First support plate; 301. Support plate base; 302. Liquid cooling channel; 303. Liquid inlet; 304. Liquid outlet; 305. Diffuse scattering glass cover; 4. Second support plate; 5. Through groove; 6. Wire; 7. Ring LED light source; 8. Industrial camera. Detailed Implementation

[0018] The technical solutions of this utility model will be clearly and completely described below with reference to the embodiments of this utility model. Obviously, the described embodiments are only some embodiments of this utility model, and not all embodiments. Based on the embodiments of this utility model, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this utility model.

[0019] like Figure 1 and Figure 2As shown, an automatic foot detection device for network filters based on AI vision includes a support column 1 and an AI vision recognition host 2. A first support plate 3 and a second support plate 4 are fixedly installed on the support column 1. The first support plate 3 is provided with a through groove 5. Several coaxially nested ring LED light sources 7 are fixedly installed on one side below the first support plate 3. At least one of the light emission radius, light emission color and illumination angle of the different ring LED light sources 7 is different. An industrial camera 8 is fixedly installed on the second support plate 4. The lens of the industrial camera 8 faces the through groove 5 and is directed towards the illumination area of ​​the ring LED light sources 7. Several wires 6 are provided on the AI ​​vision recognition host 2. The AI ​​vision recognition host 2 is electrically connected to the industrial camera 8 and each ring LED light source 7 through different wires 6. The AI ​​vision recognition host 2 is configured to independently control the opening and closing and brightness of each ring LED light source 7.

[0020] like Figure 3 As shown, in another embodiment of the present invention, the first support plate 3 includes a support plate base 301, and a liquid cooling channel 302 is provided in the support plate base 301. The two ends of the liquid cooling channel 302 are respectively provided with a liquid inlet 303 and a liquid outlet 304.

[0021] A diffuse glass cover 305 is fixedly installed below the support plate base 301;

[0022] Coolant is injected into the liquid cooling channel 302 through the inlet 303. When the coolant enters the liquid cooling channel 302 through the inlet 303 and circulates in it, it can effectively absorb and carry away the heat generated by the ring LED light source 7 during operation. Then the coolant is discharged from the outlet 304, realizing continuous heat dissipation control of the light source system. The diffuse scattering glass cover 305 can soften and uniformly scatter the light emitted by the LED, eliminate direct light spots, and provide a uniformly distributed illumination field, thereby providing an ideal optical environment for the industrial camera 8 to acquire images.

[0023] The working principle of this AI vision-based network filter automatic foot detection device:

[0024] In use, the network filter element to be tested is placed in the testing station below the ring LED light source 7. The AI ​​vision recognition host 2 precisely controls the on / off state and brightness of each ring LED light source 7 through independent wires 6 according to the preset testing program. This allows light sources of different radii, colors and angles to be turned on individually or in combination to form the optimal lighting scheme for specific defect features. The industrial camera 8 collects images of the illuminated element through the through slot 5 and transmits the image data to the AI ​​vision recognition host 2 in real time. The image processing algorithm in the host analyzes and processes the image, and judges whether there are defects such as warping, deformation, and short circuits around the foot through feature extraction and pattern recognition. This solves the problems of poor adaptability and high misjudgment rate in traditional single light source environments.

[0025] Obviously, the above embodiments of this utility model are merely examples for clearly illustrating the present invention, and are not intended to limit the implementation of the present invention. For those skilled in the art, other variations or modifications can be made based on the above description. It is impossible to exhaustively list all the implementation methods here. All obvious variations or modifications derived from the technical solutions of the present invention are still within the protection scope of the present invention.

Claims

1. An automatic foot-circling detection device for network filters based on AI vision, comprising a support column (1) and an AI vision recognition host (2), characterized in that: A first support plate (3) and a second support plate (4) are fixedly installed on the support column (1). A through groove (5) is provided on the first support plate (3). Several coaxially nested ring LED light sources (7) are fixedly installed on one side below the first support plate (3). At least one of the light emission radius, light emission color and illumination angle of the different ring LED light sources (7) is different from each other. An industrial camera (8) is fixedly installed on the second support plate (4). The lens of the industrial camera (8) faces the through groove (5) and is directed toward the illumination area of ​​the ring LED light source (7). Several wires (6) are provided on the AI ​​visual recognition host (2). The AI ​​visual recognition host (2) is electrically connected to the industrial camera (8) and each ring LED light source (7) through different wires (6). The AI ​​visual recognition host (2) is configured to independently control the opening and closing and brightness of each ring LED light source (7).

2. The automatic foot-circling detection device for network filters based on AI vision according to claim 1, characterized in that: The shape and size of the through groove (5) are adapted to the illumination area of ​​the annular LED light source (7) to define the detection area.

3. The automatic foot-circling detection device for network filters based on AI vision according to claim 2, characterized in that: The first support plate (3) includes a support plate base (301), and a liquid cooling channel (302) is provided inside the support plate base (301). The two ends of the liquid cooling channel (302) are respectively provided with a liquid inlet (303) and a liquid outlet (304).

4. The automatic foot-circling detection device for network filters based on AI vision according to claim 3, characterized in that: A diffuse glass cover (305) is fixedly installed below the support plate base (301).

5. The automatic foot-circling detection device for network filters based on AI vision according to claim 4, characterized in that: The liquid inlet (303), liquid outlet (304), and liquid cooling channel (302) are connected.

6. The automatic foot-circling detection device for network filters based on AI vision according to claim 5, characterized in that: The diffuse glass cover (305) covers the outside of the ring LED light source (7).

Images