Smart cameras combine a camera, lens, programmable processing, communications and sometimes lighting, in a small enclosure. Smart, small and flexible - they are an ideal solution for many machine vision tasks. This article discusses when to use a smart camera and what attributes to look for when choosing one.

Smart cameras acquire and process images and return results rather than-or in addition to-images. Their small size, ease of integration, and low cost allow them to satisfy a wide range of machine vision applications. They are an excellent choice for tasks such as gauging, assembly verification, robot guidance or visually detecting part defects. They can be designed to operate in harsh environments, for example exposure to water in food in food processing, and their small size and weight allows them to fit into existing manufacturing equipment or ride on a robot arm.

A key consideration in choosing a smart camera is how easy it is to program the camera software. The camera's user interface and processing algorithms should help you quickly develop and solve vision tasks without requiring a lot of specialized knowledge.

When to replace human vision with machine vision?

Replacing human vision with machine vision can reduce costs and improve product quality on tasks that are fast, precise or repetitive. Machine vision systems have three general capabilities:

• Finding part positions and guiding automated assembly
• Identification of parts, perhaps by shape or an optical code (e.g. a bar code)
• Finding defects, such as incorrect part dimensions, bad colors, or surface flaws

Vision tasks that need these abilities and capabilities might be appropriate for automation by machine vision. Generally, tasks where you cannot limit part shape, part presentation or lighting are difficult to automate. For example, currently it is more cost-effective to use people to pick and pack strawberries than to automate this task with machine vision.

Your machine vision vendor or systems integrator can help you decide if your vision task can cost-effectively be done by machine vision and if a smart camera is appropriate. As you develop machine vision applications, you will learn what vision tasks are possible and make financial sense.

Is a smart camera appropriate?

Generally, a smart camera is appropriate when:

• size, cost, and/or power requirements must be small
• The processing rate is moderate, perhaps 20 parts per second for simple gauging and less as the vision task increases in complexity

You and your vision vendor or integrator should test the smart camera to be sure it solves your vision task within the allowed time and tolerances. This usually requires trying various ways to light and position the part, selecting a suitable lens, and establishing communication between the smart camera and the equipment it is embedded in. Easy-to-use software is, again, critical for quickly testing and solving your machine vision task.

An example of an appropriate application for a smart camera is inspecting pharmaceutical pills in blister packs. In this example a conveyer presents cards with pills in the smart camera's field of view. Diffuse lighting allows the smart camera to see through the plastic enclosing the pills so it can identify incorrect pills and inspect pills for damage. The BOA^TM smart camera from Teledyne DALSA communicates with a PLC (programmable logic controller) to reject cards with incorrect or damaged pills.

Despite its small size, the BOA smart camera is a "server" on an Ethernet network, and is programmed by another "client" machine, typically a personal computer, on the same network. Programming is done using iNspect^TM vision software. This software allows you to quickly develop vision applications through its graphical interface. For example, to measure the diameter of a bottle neck, you would simply drag and drop a caliper graphic across the bottle neck.

What to look for in a smart camera

First, a smart camera must have the computational ability to solve your machine vision task within the cycle time of your process.

Second, a camera small size allows it to fit it in tight spaces or to retrofit into existing processes. Small size and weight are also important when the camera is mounted on a moving platform, such as a robot arm.

Third, the camera must be designed to withstand the environment it is used in. This could include temperature range, splash from wash down, dust, vibration, and electrical noise. Cables and connectors must also be rated for your environment.

Fourth, ease of use is critical. Smart camera software should minimize the knowledge of machine vision you need and make programming as quick and easy as possible.

Last, there are features and accessories that make using a smart camera quicker and easier. You should be able to remotely program and monitor the smart camera. The smart camera should use your factory communication protocols. Most smart cameras have a few input/output (IO) lines, limited by the space on the camera for connectors, so you might need an IO "expansion box." You also want a camera that can control lighting, so that the lighting can be synchronized with the smart camera's image acquisition.