This is essentially computer hardware, specialized for image storage, processing and analysis. The hardware is typically configured as either a stand-alone system, complete with its own enclosure or power supply, or as a board level product, to be plugged into a VME-bus or a personal computer. The hardware also provides the connections necessary for sending and receiving communication and control signals to and from external devices and other automation machinery.
Application Development Software
Software application development tools allow a developer to customize a vision system to suit a particular application. Two basic elements of application development software need to be configured:
1.
The machine vision software tools needed to solve the vision task itself - for example, to specify what edge detection tools the vision system will use to locate the edges of a part to be measured. These are commonly referred to as the “vision tools”; and
2.
The system’s interface to its factory floor environment - to design an operator interface for instance. The application development software itself is available as either a set of low-level C++ program routines and functions which allow experienced computer programmers to build a complete machine vision application from the ground up, or as a graphical and intuitive development environment, which requires little or no previous programming experience.
Operator Interface
Using the application development software, a developer can create a customized user interface to allow line operator to communicate with the vision system at run time. Once the system is deployed on the factory floor, the line operator will be able to configure and control the vision system (start and stop an inspection process, for example) through this interface. The operator interface might consist of graphic text boxes, pull-down menus, icons and push buttons displayed on a monitor and accessed through a keyboard, trackball, or touch screen - or it may simply consist of a start button or an inspection pass/fail light.
An electronic camera provides input to the vision system in the form of a coded video image, or “picture”. From the camera, this image is acquired by the vision system and held in video memory from where it can then be processed and analyzed to obtain a desired result (pass/fail for example) and displayed on the operator’s monitor screen.
Input / Output
A machine vision system often needs to be able to communicate with other machinery in its environment. In addition to the user interface, the vision system must also be able to accept input from, and provide output to other line automation equipment such as switches, triggers, PLC’s and motion controllers. An example of this kind of external input might be a trigger used to signal the vision system that a part to be inspected is under the camera lens, ready to be inspected. Output might include discrete signal sent to a PLC for process control, or perhaps serial output to help guide the movement of a robot arm or to send results data to a statistical process control package.
Fixtures / Lighting
Although the parts handling and lighting mechanisms are technically not part of the machine vision system itself, they are nevertheless important considerations in any vision application. Conveyors, indexing tables and robots are all examples of part handling devices that can be employed to present a part to the camera. As for lighting, requirements vary form one application to another; sometimes existing ambient light can be used to successfully illuminate a part; at other times, more specialized lighting techniques are needed. A strobe light, for example, is often employed to “freeze” the images of parts moving by rapidly on a conveyor belt, eliminating blurred part images.
