Surface defects in industrial products can negatively impact their aesthetics, comfort, and performance. Therefore, manufacturers need to inspect for surface defects to detect and control them in a timely manner.
Machine vision inspection methods can largely overcome the drawbacks of manual inspection methods, such as low sampling rate, low accuracy, poor real-time performance, low efficiency, and high labor intensity, and are being increasingly widely researched and applied in modern industry.
This paper focuses on machine vision-based surface defect detection. Based on a comprehensive review of relevant literature and development achievements, it summarizes the applications of machine vision in the field of surface defect detection. The working principles and basic structures of typical machine vision surface defect detection systems are analyzed, and the current research status, existing vision software, and hardware platforms for surface defect visual detection are described.
This paper reviews the theoretical and algorithmic research related to image preprocessing, image segmentation, image feature extraction and selection, and image recognition involved in machine vision inspection. It summarizes the basic ideas, characteristics, and limitations of each major method and looks forward to possible future development directions.
In machine vision surface defect detection systems, image processing and analysis algorithms are crucial components, each with its own advantages, disadvantages, and applicable range. Improving the accuracy, real-time performance, and robustness of these algorithms has been a continuous research focus.
In conclusion, machine vision is a simulation of human vision. Machine vision surface inspection involves many disciplines and theories. Further research is needed to further develop inspection towards automation and intelligence.
Surface defect detection
Machine vision technology is a non-contact, non-destructive automated inspection technology. It is an effective means to achieve equipment automation, intelligence, and precision control, and has outstanding advantages such as safety and reliability, wide spectral response range, ability to work for extended periods in harsh environments, and high production efficiency. A machine vision inspection system acquires surface images of a product through an appropriate light source and image sensor (CCD camera), extracts feature information from the image using corresponding image processing algorithms, and then performs operations such as locating, identifying, and classifying surface defects, as well as statistical analysis, storage, and retrieval based on the feature information.
Basic Components of a Machine Vision Surface Defect Detection System
It mainly includes an image acquisition module, an image processing module, an image analysis module, a data management module, and a human-computer interface module.
The image acquisition module consists of an industrial camera, optical lens, light source, and clamping device. Its function is to acquire images of the product surface. Under the illumination of the light source, the product surface is imaged onto the camera sensor through the optical lens. The light signal is first converted into an electrical signal, and then into a digital signal that can be processed by a computer. Currently, industrial cameras are mainly based on CCD or CMOS chips. CCD is currently the most commonly used image sensor in machine vision.
The light source in machine vision directly affects image quality. Its function is to overcome ambient light interference, ensure image stability, and obtain images with the highest possible contrast. Currently, commonly used light sources include halogen lamps, fluorescent lamps, and light-emitting diodes (LEDs). LED light sources have gained widespread application due to their advantages such as small size, low power consumption, fast response speed, good monochromatic light emission, high reliability, uniform and stable light, and easy integration.
Illumination systems composed of light sources can be classified according to their illumination methods into bright-field illumination and dark-field illumination, and structured light illumination and strobe illumination. Bright-field and dark-field illumination primarily describe the positional relationship between the camera and the light source. Bright-field illumination refers to the camera directly receiving the reflected light from the target; generally, the camera and light source are located on opposite sides, which facilitates installation.
Dark-field illumination refers to the camera indirectly receiving scattered light from a light source on the target. Typically, the camera and light source are positioned on the same side. Its advantage is the ability to obtain high-contrast images. Structured light illumination projects a grating or line light source onto the object being measured, and demodulates the object's 3D information based on the distortions produced. Stroboscopic illumination illuminates an object with high-frequency light pulses; the camera must be synchronized with the light source during shooting.
The image processing module mainly involves image denoising, image enhancement and restoration, defect detection, and target segmentation. Because the environment, CCD image photoelectric conversion, transmission circuits, and electronic components all contribute to image noise, this noise degrades image quality and negatively impacts image processing and analysis. Therefore, image preprocessing is necessary to reduce noise.
Image enhancement aims to purposefully emphasize the overall or local characteristics of a given image for a specific application, making an originally blurry image clearer or highlighting certain features of interest, amplifying the differences between features of different objects in the image, and suppressing features of little interest, thereby improving image quality, enriching information, and enhancing image interpretation and recognition. Image restoration, on the other hand, is a computer-based process for reconstructing or restoring images that have suffered quality degradation.
Image restoration often employs the same methods as image enhancement, but the results of image enhancement require further verification in the next stage; while image restoration attempts to utilize prior knowledge of the degradation process to restore the original appearance of a degraded image, such as eliminating additive noise and restoring motion blur. Image segmentation aims to separate target regions from an image for further processing.
Machine vision surface defect detection applications
Its applications are very wide, mainly including steel metallurgy, non-ferrous metal processing, high-precision copper strip, aluminum strip, aluminum foil, stainless steel manufacturing, electronic materials, non-woven fabrics, textiles, glass, paper, film and other fields.
Why use a surface defect detection system?
Ensure product quality, improve production processes, and reduce labor costs.
The main components of a line scanning surface defect detection system are:
The visual acquisition unit mainly includes a line scan camera, lens, light source, and image acquisition card.
The system support components include: camera bracket, light source bracket, and control panel bracket.
The electrical components (communication/control section) include encoders, motion control cards or PLCs, and may also include motors.
Other: Various wires and cables, CL wires, power cords, various SMPS, lighting controllers, etc.
Article source: Robotics Lecture Hall, ACT Vision System Design
