1. Introduction to Machine Vision Inspection Technology
Visual inspection uses machines to replace human eyes for measurement and judgment. It involves using machine vision products (image acquisition devices, either CMOS or CCD) to convert the captured target into an image signal, which is then transmitted to a dedicated image processing system. Based on pixel distribution and information such as brightness and color, this signal is converted into a digital signal. The image system performs various calculations on these signals to extract the target's features, and then controls the on-site equipment based on the judgment results. It is a valuable mechanism for production, assembly, or packaging. It has immeasurable value in detecting defects and preventing defective products from being delivered to consumers.
Machine vision inspection is characterized by improved production flexibility and automation. In hazardous work environments unsuitable for manual labor or where human vision is insufficient, machine vision is often used to replace manual vision. Furthermore, in large-scale industrial production, manual inspection of product quality is inefficient and lacks precision; machine vision inspection methods can significantly improve production efficiency and automation. Moreover, machine vision facilitates information integration, making it a fundamental technology for computer-integrated manufacturing.
2. Basic Components and Principles of Machine Vision Systems
A typical industrial machine vision application system includes digital image processing technology, mechanical engineering technology, control technology, light source illumination technology, optical imaging technology, sensor technology, analog and digital video technology, computer hardware and software technology, human-machine interface technology, etc.
Image components
A camera captures an electronic image of the object being detected and then sends it to a processor for analysis. The electronic image is converted into digital representations, the smallest parts of the image, called pixels. The number of pixels displayed in an image is called its resolution. The higher the image resolution, the more pixels it contains; therefore, the more pixels in the image, the more accurate the detection result.
Camera
A vision inspection system's camera has three variables that need to be adjusted to optimize the captured image: aperture, contrast, and shutter speed.
Lighting components
Proper lighting is crucial for creating the contrast needed for effective inspection. When evaluating the correct system setup for a product, designers spend considerable time determining the optimal lighting required for inspection. The type, geometry, color, and intensity of the lighting solution should provide the strongest possible contrast.
Software tools
Visual inspection systems use software to process images. The software employs algorithmic tools to aid in image analysis. Visual inspection solutions utilize combinations of these tools to perform the required inspections. Commonly used tools include search tools, boundary tools, feature analysis tools, process tools, and visual printing tools.
3. Advantages of machine vision inspection
In the inspection industry, machine vision has significant advantages over human vision.
High precision: Human vision has 64 gray levels and is weak at distinguishing tiny targets; machine vision can significantly improve the gray level and can observe targets at the micrometer level.
High speed: Humans cannot see fast-moving targets clearly, while machine shutter speeds can reach the microsecond level;
High stability: Machine vision solves a very serious problem for humans – instability. Manual visual inspection is a tedious and arduous task; regardless of reward or punishment systems, a high rate of missed detections will occur. However, machine vision inspection equipment does not suffer from fatigue or emotional fluctuations. It will diligently execute whatever is written into the algorithm every time, greatly improving the controllability of results in quality control.
Information integration and retention: The amount of information obtained by machine vision is comprehensive and traceable, and the relevant information can be easily integrated and retained.
4. Applications of machine vision inspection
Applications of visual inspection in the printing industry
Online/offline vision systems can be used to detect quality problems in the printing process, such as die cutting, ink piling, ink splatter, missing/shallow prints, misregistration, and color deviation. At the same time, online equipment can feed back the detection results of color deviation and ink quantity to the PLC to control the ink supply of the printing equipment and adjust the ink supply online to improve printing quality and efficiency.
Application of visual inspection in PCB board inspection
A vision system is used to inspect bare PCB boards, detecting errors in the position and spacing of conductors and components, errors in the size of circuits and components, errors in component shape, continuity of circuits, and contamination on the board.
Application of visual inspection in parts inspection
Machine vision inspection can easily handle quality control in the production of metal parts, such as coins, automotive components, and connectors. Through image processing, it can detect defects on the surface of metal parts, such as scratches, defects, discoloration, and adhesive residue, and guide the mechanical transmission system to reject defective products, significantly improving production efficiency. Furthermore, statistical analysis of defect types can guide the adjustment of production parameters, improving product quality.
Applications of visual inspection in automotive safety
The working principle of this type of digital system is to use visual sensors to detect and measure in real time the geometric and movement characteristics of a person's eyelids and eyeballs, the gaze angle and its dynamic changes, and changes in head position and direction. This establishes a model of the relationship between the driver's eye and head characteristics and fatigue state, and studies multi-parameter comprehensive description methods for fatigue state. Simultaneously, it researches methods for rapid fusion of multi-source information to improve the reliability and accuracy of fatigue detection, thereby developing a stable and reliable driver fatigue monitoring system. It employs many detection methods, such as rapid face detection methods, fatigue level detection methods, and fatigue driving problem detection, etc.
Automatic surface damage control system for metal plates
The automatic flaw detection system for metal plates utilizes machine vision technology to automatically inspect metal surface defects. It performs high-speed and accurate detection during the production process, and its non-contact measurement method avoids the possibility of creating new scratches. By combining the self-scanning characteristics of a linear CCD with the X-axis movement of the inspected steel plate, three-dimensional image information of the metal plate surface is obtained.
Automotive body inspection system
The 100% online inspection of the body contour dimensions of the 800 series vehicles manufactured by ROVER in the UK is a typical example of machine vision systems used in industrial inspection. The system consists of 62 measurement units, each including a laser and a CCD camera, used to inspect 288 measurement points on the vehicle body shell. The car body is placed under the measurement frame, and its precise position is calibrated by software.
Intelligent Traffic Management System
By placing cameras on major traffic arteries, when a vehicle violates traffic rules (such as running a red light), the camera captures the vehicle's license plate and transmits the image to the central management system. The system then uses image processing technology to analyze the captured image, extract the license plate number, and store it in a database for management personnel to retrieve.
Bottled beer production line inspection system
It can detect whether beer meets the standard volume and whether the beer label is intact.
5. The Development and Future of Machine Vision
Image acquisition technology is developing rapidly.
With increasingly mature firmware for CCD and CMOS sensors, the size of image-sensitive devices continues to shrink, the number of pixels and data rates are constantly increasing, and the speed of improvement in resolution and frame rate can be described as rapid. Product series are also becoming more and more abundant, with continuous optimization of parameters such as gain, shutter speed and signal-to-noise ratio. By using core test indicators (MTF, distortion, signal-to-noise ratio, light source brightness, uniformity, color temperature, comprehensive evaluation of system imaging capabilities, etc.) to comprehensively select light sources, lenses and cameras, many previously difficult imaging problems have been continuously overcome.
Image processing and pattern recognition are developing rapidly.
In image processing, with the extraction of high-precision edge information from images, many low-contrast defects that were originally difficult to detect directly due to being mixed in with background noise have begun to be distinguished.
In pattern recognition, it can be viewed as a labeling process, classifying the patterns to be recognized into their respective categories based on a certain amount of measurement or observation. Decision theory and structural methods are frequently used in image recognition. Decision theory is based on decision functions, which are used to classify and recognize pattern vectors, relying on time-dependent descriptions (such as statistical textures). Structural methods decompose objects into patterns or pattern primitives, with different object structures having different primitive strings (or strings). By using given pattern primitives to find the encoding boundaries of unknown objects, strings are obtained, and their class is determined based on these strings. In feature generation, many new algorithms are constantly emerging, including wavelet, wavelet packet, and fractal-based features, as well as unique binary component analysis; further developments include support vector machines, deformable template matching, and the design of linear and nonlinear classifiers.
Breakthroughs brought by deep learning
Traditional machine learning relies primarily on human analysis and logic for feature extraction. Deep learning, however, simulates brain function through multilayer perceptrons, constructing deep neural networks (such as convolutional neural networks) to learn simple features, build complex features, learn mappings, and output results. During training, all layers are continuously optimized. Specific applications include automatic ROI segmentation; punctuation localization (allowing flexible detection of unknown defects through simulated vision); re-detection of indescribable or quantifiable defects such as orange peel defects from heavily noisy images; and distinguishing between real and fake defects in glass cover detection. As more and more deep learning-based machine vision software enters the market (including Vidi from Switzerland, SUALAB from South Korea, and the Hong Kong Applied Science and Technology Research Institute), the empowering effect of deep learning on machine vision will become increasingly evident.
The development of 3D vision
3D vision is still in its early stages. Many applications are using 3D surface reconstruction, including navigation, industrial inspection, reverse engineering, surveying, object recognition, measurement and classification, etc. However, accuracy issues limit the application of 3D vision in many scenarios. Currently, the most widely used application in engineering is the volume measurement of standard parts in logistics, and it is believed that this area has huge potential in the future.
Disclaimer: This article is a reprint. If it involves copyright issues, please contact us promptly for deletion (QQ: 2737591964). We apologize for any inconvenience.
