Machine vision is the use of cameras, lighting, and image-processing software to inspect, measure, identify, or guide automatically in industrial settings. It turns images into decisions or control signals, such as pass/fail or a part position.
A machine vision system has five parts: lighting designed to make the feature visible, a camera or smart camera, optics, a processor, and software. The output is a result or signal sent to a PLC or robot: pass/fail, a measurement, a code read, or coordinates.
Common tasks include presence and absence checks, dimensional gauging, barcode and 2D code reading, optical character recognition, defect detection, and robot guidance. Traditional tools are rule-based: edge detection, blob analysis, pattern matching. Deep learning is now often added for variable defects that are hard to describe with rules.
Machine vision is related to computer vision but not the same. Computer vision is a research field; machine vision is the engineering discipline of applying imaging to real production problems. In practice, lighting and fixturing often decide success more than the algorithm. A good camera cannot fix glare, motion blur, or inconsistent part position.
Questions about Machine vision
Is machine vision the same as computer vision?
No. The two are related, but computer vision is a research field, while machine vision is the engineering discipline of applying imaging to real production problems.
What are the components of a machine vision system?
Five parts: lighting designed to make the feature visible, a camera or smart camera, optics, a processor and software. The system sends its result to a PLC or robot as a pass/fail, a measurement, a code read or coordinates.
Does machine vision need deep learning?
No. Traditional tools are rule-based: edge detection, blob analysis and pattern matching. Deep learning is often added for variable defects that are hard to describe with rules. In practice, lighting and fixturing often decide success more than the algorithm.
Machine vision in real plants
- Endress+Hauser and MVTec Software: Endress+Hauser Flow in Reinach uses MVTec MERLIC, integrated into its MES, to verify manual flowmeter assembly steps by camera; in use since March 2025.
- La Próxima Estación and KEYENCE: La Próxima Estación runs two KEYENCE IV4 AI vision sensors on its line in Irapuato to detect foreign matter and off-color vegetables, linked to a rejection system it built itself.
- MAB Möbel and MVTec Software: At MAB Möbel in Muotathal, a HOMAG robot cell uses a 3D laser scanner and MVTec HALCON to pick wooden workpieces from a chaotic stack and load a CNC machining center.
- Gupta Permold and FANUC: Gupta Permold runs a robotic beam-sanding cell with a FANUC robot, 3D vision and AI guidance, built by Premier Automation with PushCorp, across all three shifts.
- Gebr. Baldauf and MVTec Software: The dairy Gebr. Baldauf uses a mobile cheese-care robot from Eberle whose 4K camera and MVTec HALCON deep learning inspect ripening cheese wheels for mold and defects.
Vendors linked to Machine vision
MVTec Software (6) · FANUC (4) · Mujin (4) · SICK (3) · Yaskawa Electric (3) · ABB (2) · Basler (1) · Cognex (1) · KEYENCE (1) · OMRON (1) · Universal Robots (1) · Teledyne DALSA
Related terms
Computer vision inspection
Computer vision inspection is the automated checking of parts or products for defects using cameras and image analysis, increasingly with deep learning models.
Bin picking
Bin picking is a robotic task in which a robot, guided by a vision system, locates and grasps individual parts from a container where they lie in random positions and orientations.
Edge AI
Edge AI is running machine learning models on or near the device that produces the data, such as a camera, gateway, or industrial PC, instead of in a remote cloud.
Synthetic data
Synthetic data is artificially generated data that mimics some statistical properties of real data.
Sources