What Is a Vision Sensor?
A vision sensor is a device that can perform various inspections by processing images captured by a camera using a computer.
In recent years, the manufacturing industry has been increasing the speed of production lines and diversifying the types of products manufactured.
In conjunction with this trend, the required level of quality has become higher and higher.
In order to meet these requirements, visual inspections of all products are unavoidable, but there are various problems and issues, such as soaring labor costs, personnel shortages, and the time required to train inspectors.
Therefore, vision sensors are increasingly being adopted as an alternative to conventional visual inspections.
Uses of Vision Sensors
Vision sensors are used in a wide variety of industries in multiple ways.
They can be broadly categorized into the following five applications:
- Inspection, to detect the presence or absence of products, to count the number of items, to detect different types of products, etc.
- Positioning, to provide feedback of product position information to industrial robots, etc.
- Dimensional inspection, to ensure that product dimensions are in accordance with specifications.
- Visual inspection, to find various visual defects such as scratches and rust.
- Recognition inspection, such as recognizing characters on a board.
In the case of visual inspections by inspectors, it is possible to miss defects due to errors in quantity counts, wrong type of product, or ambiguous judgment criteria.
On the other hand, in the case of vision sensors, the judgment criteria can be controlled by numerical values, and by setting them appropriately, stable inspections can be performed.
Principle of Vision Sensors
Image inspection with vision sensors consists of hardware and software that operates the hardware.
Hardware includes cameras, lenses, and lighting.
The first step in image inspection is to ensure that the content to be inspected can be captured.
Proper hardware selection is essential for accurate image inspection.
Software includes algorithms and programs.
An image inspection algorithm can be defined as a means of analyzing an image and inspecting it for correctness.
In order for the computer to detect such as scratches, rust, or foreign objects in the image as NG, the image inspection algorithm must be properly constructed.
By organizing images captured by a camera using an image inspection algorithm, various defects and positional information can be calculated and output for use in the operation of the equipment.
Based on the algorithm, a program is assembled using a computer language, such as C, to associate inputs and outputs so that the inspection can proceed.
When the hardware and software function properly, it can be utilized as a vision sensor.
Difference Between a Vision Sensor and a Camera
Both are similar to the extent that they are equipment for capturing and converting images and video to digital data by combining an imaging device such as CMOS or CCD and an optical lens.
In a camera (digital camera), the images and video obtained in this way are stored for later review or for use in creating documents, etc. In a vision sensor, the images and video are stored for later review or for use in creating documents, etc.
Vision sensors, on the other hand, use computers to analyze images and video in real time. They can be used to inspect the appearance and quantity of products, measure dimensions and angles, and replace sensors, which until now have been done visually by humans. Depending on the model and settings, it is also possible to keep the images and videos used for inspection and use them for traceability.
How to Use Vision Sensors
First, it is necessary to prepare the camera of the vision sensor so that it can correctly capture images and video. No matter how sophisticated the vision sensor is, if it cannot capture images properly in the first place, it will not be able to demonstrate its true performance.
Many considerations must be taken into account, including the selection of the vision sensor and lens to be used, the fixing and positioning of the vision sensor itself and the object to be inspected, and the brightness of the surroundings (including changes in sunlight with time).
Depending on the product, vision sensors can perform many processes, such as finding a specific shape registered as a master from the captured image or video, examining the color of a specific area, finding a straight line or circle, or measuring dimensions. Depending on what you want to inspect, set the appropriate processing to be performed.
In some cases, “pre-processing” such as shooting extremely bright (or dark), inverting colors, or removing noise can make subsequent inspections easier or improve accuracy.
Inspection results can be output via contact output or via Ethernet or other communications. Select a compatible host (PC, PLC, etc.).