How do you think the new GigE standards will influence the machine vision industry?
Respond or ask your question now!
By Daryl Martin
The vision industry now boasts a considerable technology base and growing product offering of task-specific lighting sources and techniques; therefore the appropriate choice of lighting can directly affect the quality and robustness of a vision system implementation. It is important, then, for an integrator or VAR to consider how the choices of lighting and optics, cameras, and immediate inspection environment (mounting, sample handling and presentation) interrelate (Fig. 1).
Lighting and Optics
Appropriate lighting and optics are important for a successful vision implementation, and indeed most vision engineers would agree that lighting may be the most critical aspect. It is logical, then, to begin this process of component specification and system development by first determining the lighting that produces the most effective and robust contrast for the features of interest, followed by lens, filters, camera, and mounting options. The choice of lens focal length is often a function of the type and/or working distance of the lighting. For example, by their nature, diffuse dome and some circular dark field lights require close proximity to the sample, and many have relatively narrow viewing “port holes,” thus requiring a short focal length lens to prevent vignetting. Conversely, a sharply focused spot or coaxial ring light may require a long focal length lens so that the camera and lens assembly does not interfere with the lighting. Finally, to effectively block ambient white light contribution with a band pass filter, the light and filter must be matched by wavelength, which precludes the use of white light for the inspection – a potential problem if a color camera has been specified.
When selecting lighting with respect to cameras, there are two primary issues to consider: wavelength compatibility and intensity. Vision camera sensors on the market today have different spectral sensitivity response curves, and all exhibit a non-linear response to visible light, as well as near UV and IR. Therefore, it is useful to optimize sensitivity by matching lighting wavelength peaks and maximum camera sensitivity range. For example, when using IR lighting, which is invisible to human vision, a camera with an IR enhanced sensor can increase light sensitivity 2x-4x, which may be critical for inspections on a high-speed line, or when using a high frame rate camera, because of the shorter exposure times necessary for freezing motion.
Immediate Inspection Environment
There is a complex relationship among the camera and light geometry, and proximity vs. sample handling and presentation. Accordingly, it is useful to conduct a thorough analysis of the immediate physical inspection environment, or the 3-D Working Envelope (camera – light – sample geometry; See also Fig. 1), as early as possible in the system development process. Ideally, the mounting, sample handling and presentation mechanicals are designed to accommodate the best lighting geometry and camera position requirements. However, real world circumstances often force a compromise. In example Fig. 2, based on careful analysis of the features of interest, a coaxial bright field ring light is specified for a 2-D peen matrix code, but because of pre-existing parts handling mechanical restrictions, the light can be placed neither directly above the part for the correct geometry, nor in close proximity for sufficient light intensity. Therefore, a different 3-D working envelope is indicated, which includes employing an off-axis dark field linear light and increasing the camera’s working distance, with a commensurate longer focal length lens. Additionally, the vision system software must also be changed to inspect for bright peen “dots” in a dark field, rather than vice-versa.
In summary, it is important to recognize that there are complex interrelationships among the hardware components that comprise a vision system, primarily lighting, lenses and filters, cameras, and mounting options. The selection of one component, particularly lighting, can have cascading effects on the selection of the other primary components. If an integrator, VAR, or end user has the luxury of designing the entire system, including the sample handling and presentation hardware components, then it is useful to begin by determining the optimum lighting and optics that most accurately and consistently accentuate the features of interest, and then build the remainder of the system around those choices.