- SPECIAL REPORTS
Advanced vision systems drive quality and process automation for manufacturers. In order to achieve its mandate, each vision system is composed of many components, including the camera, digitizer, optics, lighting, processor, inspection software tools, inspection application software, communications, display, calibration requirements and more. Staying mindful of the following areas will help you design these components to work robustly together in an industrial environment.
- Required inspection cycle time– This can influence the type of processor used as well as the number of cameras per processor.
- Inspection tolerances– Tighter inspection tolerances require higher resolution cameras and higher quality optics to achieve the required image resolution.
- Features to be inspected– The types of features to be inspected will influence the choice of lighting and the types of inspection software tools required.
- Inspection environment - Factors such as whether the part will be in motion during the inspection, whether the environment is dusty/radioactive/high temperature, etc., will influence hardware choices (e.g., special shielding, dust protection)
- Mechanical considerations- Taking note of such factors as no-go zones for camera and lights will ensure that the vision system is not damaged by other equipment moving in the work cell.
- Range of tolerances - Analyzing sample part types that are representative of the range of tolerances that will be seen during the normal inspection process helps to ensure that the vision system will be robust during actual production.
In order to select a camera for your vision system you need to know the required image resolution. There are three parameters that determine image resolution:
- Inspection tolerance requirement
- Size of the part to be inspected
- Mechanical repeatability of the system that is presenting the parts to the camera
Your image resolution requirements will drive your cameras sensor size, or pixel resolution. You should choose the camera sensor type based on the type of inspection the system will perform. Some examples of camera sensor choices include monochrome, color, IR, and area scan vs. line scan.
The optical design determines the required lenses while taking into account the following:
- Camera standoff requirements
- Desired size of the field of view
- Pixel size to be resolved
- Camera sensor size
- Telecentricity requirements, if any
Note that one or more filters often need to be added to the optical path either to enhance the lighting system or suppress outside sources of light.
If the inspection system includes a gauging component, you will need to determine a calibration method. This could involve creation of a calibration target, calibration of image scales and, in some cases, calibration across multiple cameras and between cameras and motion platforms (robots, servo axes, etc.). Defining a calibration method enables you to determine image scales so that results can be produced in engineering rather than pixel coordinates.
The purpose of the lighting design is to map out how to light the inspection object in such a way that there is enough contrast between the inspection features and the background. If the part will be in motion during the inspection then the lighting system might have to be strobed to stop the motion with minimal image blur. It also may be important to synchronize the lighting to the image capture.
Testing your design
The design of the camera, optics and lighting combination should be tested in laboratory conditions, using the actual parts that the vision system will eventually inspect, to verify achievement of the appropriate field of view, camera resolution and lighting. The configuration should be recalibrated and retested until results are satisfactory.
Documentation and Reporting
Once you have your final vision system design, documenting it in a Vision Design Specification document allows you to share it easily with other design disciplines, such as mechanical and electrical.
You will also need a method for reporting the results of the vision inspection to the Programmable Logic Controller (PLC) that is controlling the main process or the operator. The vision results data will need to be documented, along with any communications handshake between the PLC and the vision system. “The communication method is usually determined based on the type of device that is receiving the vision results,” says Paul Sommers, Senior Vision Designer for ATS. “For example, Allen Bradley PLCs typically communicate using Ethernet/IP, while Siemens PLCs typically use Profinet or Profibus.”
No two vision systems are the same. There is an art and a science to designing a vision system so that all the components work together to deliver the results you’re looking for. Following the recommendations above will help you design the best solution for your needs.
Contact: Steve Wardell