Precision part surface defect recognition device based on deep learning and CCD imaging

Abstract

The utility model provides a precision part surface defect identification device based on deep learning and CCD imaging, which comprises a machine base, a conveying mechanism, a CCD imaging detection mechanism, a defect elimination mechanism and a limiting assembly are arranged in the machine base, the conveying mechanism comprises two conveying belt modules which are arranged in parallel, the two conveying belt modules are fixedly arranged in the machine base, and the CCD imaging detection mechanism and the defect elimination mechanism are arranged in the machine base. Shock pads are further arranged between the conveyor belt modules and the machine base, the CCD imaging detection mechanism comprises an adjusting support, two high-resolution CCD cameras, an annular shadowless light source and a strip-shaped auxiliary light source, the high-resolution CCD cameras and the conveyor belt modules are arranged in a one-to-one correspondence mode, the limiting assembly is located at the front ends of the conveyor belt modules, and the defect removing mechanism is located at the tail ends of the conveyor belt modules. The defect removing mechanism comprises a turnover type sorting plate, a rotating motor and a waste collecting box. The utility model has the beneficial effects that the combined structure of the conveying mechanism, the CCD imaging detection mechanism, the defect eliminating mechanism and the limiting component is adopted, the conveying is stable, the eliminating accuracy is high, and the detection efficiency of the whole machine is improved.

Description

Technical Field

[0001] This utility model mainly relates to the field of visual inspection equipment technology, specifically to a precision parts surface defect identification device based on deep learning and CCD imaging. Background Technology

[0002] In the field of precision parts manufacturing, traditional inspection technologies are no longer sufficient to meet the demands of high-quality and high-efficiency production. Early methods based on manual visual inspection or simple optical inspection suffer from low inspection efficiency, susceptibility to subjective factors, and difficulty in identifying minute defects. Although some equipment using single CCD imaging technology has achieved automated inspection, it lacks intelligent algorithm support, resulting in low accuracy in identifying surface defects of parts under complex background interference. Furthermore, it cannot adapt to the inspection requirements of different types of parts, leading to frequent missed and false detections.

[0003] During the actual implementation process, the inventors discovered the following defects:

[0004] With the breakthroughs in deep learning technology in the field of image recognition, combining it with CCD imaging technology for precision parts inspection has become a new trend. However, existing related equipment has many shortcomings in structural design, such as poor stability of parts transportation, unreasonable light source layout, and weak coordination between detection and rejection mechanisms, which limits the overall detection efficiency and accuracy.

[0005] It should be noted that the above content falls within the scope of the inventor's technical knowledge. Due to the vast and complex nature of the technical content in this field, the above content of this application does not necessarily constitute prior art. Utility Model Content

[0006] 1. The technical problem to be solved by the utility model:

[0007] The present invention provides a precision part surface defect identification device based on deep learning and CCD imaging to solve the technical problems existing in the background art.

[0008] 2. Technical Solution:

[0009] To achieve the above objectives, the technical solution provided by this utility model is as follows: a precision part surface defect identification device based on deep learning and CCD imaging, comprising a base, wherein the base is provided with a conveying mechanism, a CCD imaging detection mechanism, a defect rejection mechanism, and a limiting component. The conveying mechanism includes two parallel conveyor belt modules, which are fixedly installed within the base. A shock-absorbing pad is also provided between the conveyor belt modules and the base. The CCD imaging detection mechanism includes an adjustment bracket, two high-resolution CCD cameras, a ring-shaped shadowless light source, and a strip-shaped auxiliary light source. The high-resolution CCD cameras are configured in a one-to-one correspondence with the conveyor belt modules. The limiting component is located at the front end of the conveyor belt modules. The defect rejection mechanism is located at the end of the conveyor belt modules and includes a flip-type sorting plate, a rotary motor, and a waste collection box.

[0010] Furthermore, both the high-resolution CCD camera and the ring-shaped shadowless light source are suspended on the adjustment bracket via a clamping frame.

[0011] Furthermore, the annular shadowless light source is fixed to the front end of the high-resolution CCD camera via the clamping bracket.

[0012] Furthermore, the two strip-shaped auxiliary light sources are fixed to both sides of the base by positioning posts.

[0013] Furthermore, the flip-type sorting plate is provided with rotating rods that pass through the base on both sides, and the drive end of the rotary motor is connected to the rotating rods.

[0014] Furthermore, the waste collection box is located below the flip-type sorting plate and has a receiving port near the upper end of the flip-type sorting plate.

[0015] Furthermore, the limiting component includes width-adjusting baffles fixed to both sides of the conveyor belt module, and the width-adjusting baffles are fixed to the positioning posts on the side of the machine base by adjusting bolts.

[0016] Furthermore, a defect identification and judgment module is provided on the side of the base near the defect rejection mechanism. The defect identification and judgment module consists of an image acquisition card, an industrial computer, and a display. The high-resolution CCD camera is connected to the defect identification and judgment module via a data cable.

[0017] 3. Beneficial effects:

[0018] Compared with the prior art, the technical solution provided by this utility model has the following advantages:

[0019] This device uses a dual conveyor belt module with shock-absorbing pads to improve conveying stability and inspection efficiency; the adjustable bracket and multi-light source structure ensure clear imaging and can identify minute defects; the flip-type sorting plate combined with a stepper motor has a high rejection accuracy; the limit component provides precise positioning and is compatible with parts of different specifications; the defect identification and judgment module processes efficiently, greatly improving the overall inspection efficiency.

[0020] It should be noted that the structures not described in this utility model are the same as or can be implemented using existing technology, and will not be elaborated here, as they do not involve the design points and improvement directions of this utility model. Attached Figure Description

[0021] Figure 1 This is a schematic diagram of the overall structure of this utility model;

[0022] Figure 2 This is a schematic diagram of the overall structure of this utility model from another angle;

[0023] Figure 3 This is a partial structural schematic diagram of the present invention.

[0024] Figure label:

[0025] 1. Base; 2. Conveying mechanism; 201. Conveyor belt module; 3. CCD imaging detection mechanism; 301. Adjusting bracket; 302. High-resolution CCD camera; 303. Ring-shaped shadowless light source; 304. Strip auxiliary light source; 305. Clamping frame; 4. Defect rejection mechanism; 401. Tilting sorting plate; 402. Rotary motor; 403. Waste collection box; 5. Limiting component; 501. Width adjustment baffle; 502. Adjusting bolt; 6. Defect identification and judgment module. Detailed Implementation

[0026] To facilitate understanding of this utility model, a more comprehensive description of the utility model will be given below with reference to the accompanying drawings, which show several embodiments of the utility model. However, the utility model can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided so that the disclosure of the utility model will be more thorough and complete.

[0027] In the description of this utility model, it should be understood that the terms "center", "longitudinal", "lateral", "length", "width", "thickness", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "page", "bottom", "inner", "outer", "clockwise", "counterclockwise", etc., indicating the orientation or positional relationship are based on the orientation or positional relationship shown in the accompanying drawings, and are only for the convenience of describing this utility model and simplifying the description, and are not intended to indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of this utility model.

[0028] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this utility model, "a plurality of" means two or more, unless otherwise explicitly specified.

[0029] In this utility model, unless otherwise explicitly specified and limited, the terms "installed," "connected," "linked," "fixed," "provided with," and "located in" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection of two components. Those skilled in the art can understand the specific meaning of the above terms in this utility model according to the specific circumstances. Example

[0030] See attached document Figure 1-3A precision part surface defect identification device based on deep learning and CCD imaging includes a base 1. The base 1 houses a conveying mechanism 2, a CCD imaging detection mechanism 3, a defect rejection mechanism 4, and a limiting component 5. The conveying mechanism 2 includes two parallel conveyor belt modules 201, which are fixedly mounted within the base 1. A shock-absorbing pad is provided between the conveyor belt modules 201 and the base 1. A 5mm thick nitrile rubber shock-absorbing pad is sandwiched between the conveyor belt support and the base 1. This shock-absorbing pad is adhered to the bottom surface of the conveyor belt support with adhesive backing, effectively absorbing the vibration energy during conveyor belt operation and controlling the conveyor vibration amplitude within ±0.03mm. Inside, to ensure the stability of parts transport, the CCD imaging inspection mechanism 3 includes an adjustment bracket 301 and two high-resolution CCD cameras 302, a ring shadowless light source 303 and a strip auxiliary light source 304. The high-resolution CCD cameras 302 are set one-to-one with the conveyor belt module 201. The limiting component 5 is located at the front end of the conveyor belt module 201, and the defect rejection mechanism 4 is located at the end of the conveyor belt module 201. The defect rejection mechanism 4 includes a flip-type sorting plate 401, a rotary motor 402 and a waste collection box 403. The dual conveyor belt module design can realize the synchronous operation of two independent inspection channels, which improves efficiency compared to the single channel and is suitable for large-scale production line inspection scenarios.

[0031] Both the high-resolution CCD camera 302 and the ring-shaped shadowless light source 303 are suspended on the adjustment bracket 301 via the clamping frame 305. The ring-shaped shadowless light source 303 is fixed to the front end of the high-resolution CCD camera 302 via the clamping frame 305. The two high-resolution CCD cameras 302 are suspended on the horizontal beam of the adjustment bracket 301 via the L-shaped clamping frame 305. The clamping frame 305 adopts a quick-release buckle structure and can slide laterally along the guide rail of the beam to achieve precise alignment between the camera and the conveyor belt module 201 below. Each CCD camera 302 has a ring-shaped shadowless light source 303 fixedly installed at its front end, ensuring that the center of the light source is coaxial with the camera's optical axis, providing 360° uniform illumination and eliminating reflections and shadows on the surface of the parts; two strip-shaped auxiliary light sources 304 are fixed to both sides of the base 1 by positioning posts. The left and right side walls of the base 1 are fixedly installed with the strip-shaped auxiliary light sources 304 by positioning posts. The positioning posts are equipped with height adjustment holes, which can realize the pitch angle adjustment of the light source from 0-45°, forming multi-angle supplementary lighting with the ring light source, improving the detection accuracy of edge defects of curved parts.

[0032] The tilting sorting plate 401 has rotating rods on both sides that penetrate the base 1. The drive end of the rotary motor 402 is connected to the rotating rods. The waste collection box 403 is located below the tilting sorting plate 401 and has a receiving port near the upper end of the tilting sorting plate 401. The tilting sorting plate 401 is horizontally arranged and its height is the same as the conveyor belt. The rotating rods on both sides penetrate the side wall of the base 1 and are connected to the base through deep groove ball bearings. Dustproof sealing rings are installed on the outside of the bearings. The rotary motor 402 is connected to the right rotating rod via a coupling. The moving rod connection uses a stepper motor with a positioning accuracy of 0.1°, which can precisely control the flipping angle of the sorting plate. When a defective part is detected, the motor drives the sorting plate to flip 60°, so that the part slides down the slope to the waste collection box 403 below. The width of the material receiving port at the top of the collection box is the same as the width of the conveyor belt, and the edge is provided with a guide slope to ensure that 100% of the parts fall into the box, thus improving the rejection accuracy. In another embodiment, pulleys can be installed at the bottom of the waste collection box 403 to facilitate horizontal removal and cleaning.

[0033] The limiting component 5 includes a width adjustment baffle 501 fixed on both sides of the conveyor belt module 201. The width adjustment baffle 501 is fixed to the positioning column on the side of the base 1 by adjusting bolts 502. The width adjustment baffle 501 adjusts the spacing by adjusting bolts 502 according to the product parts being detected, so as to ensure that the product parts are within the recognition range of the high-precision CCD camera during conveying.

[0034] A defect identification and judgment module 6 is also provided on the side of the base 1 near the defect rejection mechanism 4. The defect identification and judgment module 6 consists of an image acquisition card, an industrial computer, and a monitor. A high-resolution CCD camera 302 is connected to the defect identification and judgment module 6 via a data cable and to the image acquisition card via a gigabit network cable. The image data transmission rate reaches 1Gbps, ensuring real-time transmission without delay. The industrial computer is an embedded industrial control computer, which is connected to the monitor via an HDMI cable, allowing operators to view real-time inspection images and defect annotation results. This module integrates customized vision algorithms to quickly complete defect identification of a single image. Combined with dual conveyor belt channels, it improves the overall inspection efficiency of the machine.

[0035] The above-described embodiments are merely illustrative of certain implementations of this utility model, and their descriptions are relatively specific and detailed. However, they should not be construed as limiting the scope of this utility model patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this utility model, and these modifications and improvements all fall within the protection scope of this utility model. Therefore, the protection scope of this utility model patent should be determined by the appended claims.

Claims

1. A precision part surface defect identification device based on deep learning and CCD imaging, comprising a base (1), characterized in that: The base (1) is equipped with a conveying mechanism (2), a CCD imaging detection mechanism (3), a defect removal mechanism (4), and a limiting component (5). The conveying mechanism (2) includes two parallel conveyor belt modules (201), which are fixedly installed inside the base (1). A shock-absorbing pad is also provided between the conveyor belt modules (201) and the base (1). The CCD imaging detection mechanism (3) includes an adjusting bracket (301) and two high-resolution CCDs. The high-resolution CCD camera (302), the ring shadowless light source (303), and the strip auxiliary light source (304) are arranged one-to-one with the conveyor belt module (201). The limiting component (5) is located at the front end of the conveyor belt module (201), and the defect rejection mechanism (4) is located at the end of the conveyor belt module (201). The defect rejection mechanism (4) includes a flip-type sorting plate (401), a rotary motor (402), and a waste collection box (403).

2. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 1, characterized in that: The high-resolution CCD camera (302) and the ring-shaped shadowless light source (303) are both suspended on the adjustment bracket (301) by a clamp (305).

3. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 2, characterized in that: The ring-shaped shadowless light source (303) is fixed to the front end of the high-resolution CCD camera (302) by the clamp (305).

4. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 1, characterized in that: The two strip-shaped auxiliary light sources (304) are fixed to both sides of the base (1) by positioning posts.

5. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 1, characterized in that: The flip-type sorting plate (401) has rotating rods on both sides that pass through the base (1), and the drive end of the rotary motor (402) is connected to the rotating rods.

6. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 1, characterized in that: The waste collection box (403) is located below the flip-type sorting plate (401) and has a receiving port near the upper end of the flip-type sorting plate (401).

7. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 1, characterized in that: The limiting component (5) includes a width adjustment baffle (501) fixed on both sides of the conveyor belt module (201), and the width adjustment baffle (501) is fixed on the positioning column on the side of the base (1) by adjusting bolts (502).

8. The precision part surface defect identification device based on deep learning and CCD imaging according to claim 1, characterized in that: The base (1) is also provided with a defect identification and judgment module (6) on the side near the defect removal mechanism (4). The defect identification and judgment module (6) consists of an image acquisition card, an industrial computer and a display. The high-resolution CCD camera (302) is connected to the defect identification and judgment module (6) via a data cable.

Images