I. Definition of Machine Vision
Machine vision is a device that automatically receives and processes images of real objects using optical equipment and non-contact sensors to obtain the necessary information or control the movement of robots.
Machine vision is the use of machines to replace human eyes for measurement and judgment. Essentially, machine vision is the application of image analysis technology in factory automation. It uses optical systems, industrial digital cameras, and image processing tools to simulate human visual functions and make corresponding decisions, thereby directing specific devices to achieve these decisions.
In modern automated production processes, machine vision has gradually begun to replace human vision, especially in areas such as condition monitoring, finished product inspection, and quality control. With the advent of Industry 4.0, this trend is irreversible.
II. Why use machine vision instead of human vision?
There are many reasons, the main ones being as follows:
1. From the perspective of production efficiency, operators are prone to fatigue after working for a long time, resulting in low quality and low accuracy of manual vision. Machine vision can greatly improve production efficiency and automation.
2. From a cost control perspective, training a qualified operator requires significant human and material resources from company management. Furthermore, simple training is far from sufficient; substantial time is needed afterward to improve the operator's practical skills. In contrast, a machine vision system, if properly designed, debugged, and operated, can be used continuously for extended periods while ensuring production efficiency.
3. In certain special industrial environments, such as welding and gunpowder manufacturing, manual vision may pose a threat to the personal safety of operators, while machine vision effectively avoids these risks to a certain extent.
III. What fields does machine vision cover?
Machine vision systems consist of different functional modules, so designing a successful machine vision system requires a high level of skill from engineers.
Typically, machine vision encompasses the following specialized fields:
1. Electrical Engineering: Hardware and software design for machine vision systems.
2. Engineering Mathematics: The foundation of image processing technology.
3. Physics: The foundation of lighting system design.
4. Mechanical Engineering: Application of Machine Vision Systems. A good machine vision system can provide better technical support for manufacturing, thereby improving product quality and production efficiency.
IV. Components of a Machine Vision System
A complete machine vision system typically consists of an optical system (light source, lens, industrial camera), an image acquisition unit, an image processing unit, actuators, and a human-machine interface. These functional modules are complementary and indispensable.
1. Lighting (light source)
Illumination is a crucial factor affecting the input of machine vision systems. The design of the light source system is extremely important, as it is directly related to the input data, namely the image quality and the effectiveness of the application.
Engineers need to first determine the effective lighting conditions and select appropriate lighting equipment based on user needs and product characteristics to ensure that the images generated under these lighting conditions can highlight the target information features required by the user.
Light sources are generally divided into visible light sources and invisible light sources. Visible light sources commonly used in industry include LEDs, halogen lamps, and fluorescent lamps. Invisible light sources mainly include near-infrared light, ultraviolet light, and X-rays.
LED light sources are widely used in machine vision systems for educational purposes, featuring high efficiency, long lifespan, moisture resistance, shock resistance, energy saving, and environmental friendliness. This is why engineers choose them when designing lighting systems.
Invisible light sources are mainly used to meet certain specific needs, such as the inspection of pipe welding processes, because the penetration of invisible light can only reach the inspection point.
2. Lens
The lens is an important component of a machine vision system, and its function is optical imaging.
The main parameters of a lens are focal length, depth of field, resolution, working distance, and field of view.
Depth of field refers to the range of distances between the subject and the area in front of and behind the focal point when the lens can capture an image.
Field of view (FOP) represents the range that a camera can observe, usually expressed in angles. Generally, the larger the FOP, the wider the observation range.
Working distance refers to the distance between the lens and the subject being photographed. The longer the working distance, the higher the cost.
When designing a machine vision system, please select a lens whose parameters meet the user's needs.
3. Industrial cameras
Industrial cameras are indispensable in machine vision systems. They act like human eyes, capturing images. Industrial cameras can be categorized based on their image sensors: CCD cameras and CMOS cameras.
CCD cameras are more expensive, but their image quality, transparency, and color richness are far superior to CMOS cameras. CCD cameras can be categorized into two types based on the CCD image sensor they use: linear array and area array.
Linear CCD cameras use a "line" format and process image information only on a line-by-line basis, offering high resolution and high speed. They are primarily used in machine vision systems for industrial, medical, and scientific research fields.
Area array cameras can acquire information about the entire image at once and are relatively inexpensive.
4. Image Acquisition Unit
A key element in the image acquisition unit is the image acquisition card, which serves as the interface between the image acquisition unit and the image processing unit. It is used to digitize acquired images and input and store them in the computer.
5. Image Processing Unit
It contains a large number of image processing algorithms. After acquiring images, these algorithms are used to process digital images, perform analysis and calculations, and then output the results.
6. Actuators and Human-Machine Interface
After completing image acquisition and processing, the image processing results need to be output, and operations matching the results need to be performed, such as rejecting defective products, setting alarms, etc., and production information should be displayed through the human-machine interface.
V. Principles of Machine Vision Systems
The optical system converts the photographed object into an image signal, which is then transmitted to the image acquisition card. Based on information such as pixel distribution, brightness, and color, the signal is converted into a digital signal. The image processing unit effectively calculates these digital signals and obtains the feature values of the target, so as to guide the device to perform corresponding actions based on the discrimination results.
