Light source illumination mode and control circuit design of AOI inspection system

Abstract In view of the different defect characteristics under different lighting modes, a two-level lighting system and control circuit of the light source are designed according to the characteristics that the detection methods of different defects are closely related to the lighting methods adopted. This system can adopt corresponding lighting methods according to the different objects to be detected. Different defects can detect the optimal image quality of the object under the corresponding lighting methods. The control circuit can adjust the brightness of the light source in different environments and can also adjust the area where a defect part needs to be highlighted separately, meeting the requirements of the AOI inspection system for image acquisition and inspection of various components.

The automatic optical inspection system for PCB patch installation defects is an optomechanical and electrical high-tech product that integrates computer vision, image processing, CNC precise positioning and automatic control technology. Its hardware mainly includes image acquisition card, light source, optical lens, CCD and computer. Due to the limitation of the CCD field of view in the defect detection system, the system must move the PCB board or camera of the patch to collect images multiple times during detection. Therefore, it is necessary to design a motion control module to achieve precise control of the mobile platform and PCB board transmission. Defects will have different characteristics under different lighting methods. The detection methods of different defects are closely related to the light source lighting methods used. Therefore, it is necessary to design a light source system that can use the corresponding lighting method according to the detection object to optimize the image quality of the detection object collected. This paper designs a control circuit and a secondary light source system to detect the best quality image of the object under various corresponding lighting modes. The designed control circuit can automatically adjust the light source controller through the light sensor to control the brightness of the light source so that the brightness of the light source remains consistent. Secondly, there are 16 brightness control areas in total, which can be adjusted together in any combination or independently. The requirements of the AOI detection system for image acquisition and detection of each component are met.

1 Detection system principle and light source design

1.1 Principle of the detection system

The detection system detects defects in the PCB mounting process, and improves the mounting process or process control after feedback to reduce or eliminate defects. Its basic principle is shown in Figure 1. The computer program controls the camera to automatically scan the mounting PCB in different areas, collect images and process the graphics. The test images are analyzed by software such as feature extraction, information expression, classification and recognition to check the condition of the PCB.

The image acquisition module is a system that collects images of SMD PCB boards and then inputs them into the computer system for storage for display, transmission and other processing. Image preprocessing is the first step in image processing. Its main purpose is to eliminate noise, improve image quality, enhance the detectability of useful information, and create conditions for subsequent processing. Common preprocessing includes denoising, grayscale transformation and sharpening. The detection of PCB defects is mainly to detect defects with different characteristics, including missing parts, offset, skew, flipping, pin lift, side stand, wrong parts and bridging. The extraction of the above defect features is the basis for identifying defects. Defect classification and recognition is mainly based on image matching, and appropriate classifiers are constructed to identify and classify defects.

1.2 Light source design principles

Defects will show different characteristics under different lighting methods, and the detection methods of different defects are closely related to the lighting methods used. The design of a secondary light source lighting system can adopt corresponding lighting methods according to different inspection objects. Collecting the image of the inspection object under this corresponding lighting method can obtain the optimal quality image of the inspection object.

1.2.1 Light intensity transfer function

For AOI systems, an incoherent light source illumination system is usually used, and the imaging system can be equivalent to the optical model shown in Figure 2. The equivalent light intensity is

Where i is the incident angle; I0 is the outgoing light intensity; ρd(x, y) is the reflectivity distribution of the surface; d0 is the distance from the light source to the object; CCD is the image receiving device, and the lateral and longitudinal dimensions and lateral and longitudinal spacing of the CCD photosensitive element are a, b, c, and d respectively. The actual specific distribution of the charge is:

Among them, t is the charge integration time; k is the proportional coefficient; d1 is the distance from the optical system to the object's pupil; S0 is the area of ​​the human pupil; h(x, y) is the expansion function; M is the system magnification factor.

1.2.2 Light source design of optical lighting system

From formula (2), it can be obtained that the amount of electricity obtained by CCD integration is inversely proportional to the square of the distance from the object to the light source, and is proportional to the cosine value of the incident angle and the light intensity. For a given CCD, exposure should be performed within a certain range. If the exposure is too small, some points of the object will be submerged by noise due to insufficient illumination; on the contrary, if the illumination is too large, the CCD pixels will be saturated or close to saturation due to excessive exposure, which will cause a lot of distortion in the picture and make the measurement error larger. Therefore, the maximum illumination of the photosensitive surface is usually adjusted to not higher than the maximum saturation illumination, so that the dynamic range of components can be fully utilized. Therefore, a two-level lighting system is studied and designed in this paper. The brightness distribution of the welding part of the component under several lighting modes is shown in Figure 3. The brightness of the solder joint is higher under horizontal lighting, while the brightness of the upper end of the component is higher under vertical lighting. Based on this feature, two images with different visual characteristics are designed, namely horizontal lighting and vertical lighting, which use different incident angles of horizontal lighting and vertical lighting.

The brightness of the light source when using the secondary lighting system is shown in Figure 4. When the light is vertically projected onto the components parallel to the circuit board, most of the light is reflected to the CCD, while the light that is projected onto the electrode is reflected to other directions and dissipated. When the vertically projected light reaches the welding part, most of the light will not be reflected to the camera side, but will be reflected to other directions and dissipated.

2 Design and application scope of various lighting modes

The detection of different defects is closely related to the lighting method used. Many defects will have characteristics at different lighting angles. Based on these different characteristics, it is possible to comprehensively judge whether it is a defect. Therefore, the research on defect detection needs to be carried out together with the lighting method. Therefore, when detecting different defects in different parts of components, it is necessary to adopt different lighting methods according to the different characteristics obtained under lighting. In this paper, different lighting modes are designed according to the defect detection of different parts, as shown in Table 1.

2.1 Horizontal lighting

The low brightness and high brightness ranges formed under horizontal illumination are as follows: (1) Low brightness range. IC component wire ends, wire flanks, pads, component electrodes, circuit boards, etc. (2) High brightness range. Bends of wires with large inclination angles, solder joints, etc.

According to the different illumination brightness ranges, horizontal illumination is mainly used to detect cold solder joints, leaking solder joints in ICs or their wires, or to check bridges in SOICs and QFPs. Figure 5 is a typical effect obtained under horizontal illumination.

2.2 Vertical lighting

In vertical lighting mode, two main brightness ranges are formed: low and high: (1) Low brightness range. The bends of wires with large inclination angles, solder joints, etc. (2) High brightness range. The wire ends, wire wings, pads, electrodes of components, and circuit boards of IC components, etc.

Therefore, according to the different illumination brightness ranges, vertical illumination is mainly used to inspect the lead ends of components and ordinary chips. A typical effect image under vertical illumination is shown in Figure 6.

2.3 Vertical-Horizontal Lighting

Due to the diffuse reflection of the component body, text and silk-screen logo, the detection effect under vertical lighting is not ideal, so the vertical-horizontal lighting method can be used to improve the detection effect. The typical effect of subtracting the horizontal image from the vertical image is shown in Figure 7.

2.4 Horizontal-vertical lighting

Horizontal lighting is usually used to detect bridges of SOIC and QFP, or leaks, cold soldering, and wrong soldering of IC wires and chips. When horizontal lighting is not ideal for high-brightness parts such as component bodies, text, and white silk-screen logos, "horizontal-vertical" lighting can be used. The typical effect of subtracting the vertical image from the horizontal is shown in Figure 8.