Cameras in Complex Vision Systems – How and When to Use Them

Table of Contents

When and How to Use Multiple Smart Cameras in Complex Vision Systems
Start with the Cars
80 Smart Cameras Can’t be Wrong
So How to Choose?
Future Thinking

When and How to Use Multiple Smart Cameras in Complex Vision Systems

When smart cameras were launched about 10 years ago, the machine vision industry acknowledged that they were perfect for simple, straightforward applications such as reading bar codes or presence/absence verification. It was believed that a smart camera was best suited for applications that required just a single operation.

Smart cameras were and still are recognized as stand-alone vision system; however, several machine vision systems are complex and require multiple operations. The question is can multiple smart cameras be employed in more complex applications? Or is a conventional PC-based frame grabber and multiple camera system the only choice? Of course, the classic machine vision answer is, it depends on the type of application. In certain cases, multiple smart camera systems can perform much better than a traditional PC-based, multi-camera vision system.

Start with the Cars

The automotive industry was among the first to implement machine vision in an attempt to enhance the quality of the product and its production process. The vision-guided robots for the manufacturing process were the center of interest. Before the advent of smart cameras, PC-based systems with multiple cameras would have been used for verification of a sub-assembly or inspection of the final product. The present-day’s smart technology enables companies to include vision at all levels of the manufacturing process. An automotive customer might start with just a single smart camera for a particular task, and after seeing the advantages and results may add more, because the platform made it an economical and practical solution.

automotive industry

The automotive industry was one of the first to embrace machine vision.

A good example of how multiple smart cameras can be used together can be seen in the assembly line for an engine block. In this case, multiple manufacturing cells are configured to assemble the product as it moves along a line. Beginning from a tray of parts– bolts, valves, gaskets, cast assemblies, etc – smart cameras in each cell guide robots to pick and place the parts on to the assembly – with each camera configured to pick a distinct part.

Kia Sportage Assembly Line Video

After the completion of partial assembly, additional cameras confirm that the operation was carried out correctly and the assembly shifts to the next cell. Therefore, in the first part of the cell, the smart cameras are configured to execute part identification using high-level pattern matching or 2D Matrix reading functions, whereas the other cameras in the cell are programed to perform post-assembly verification.

80 Smart Cameras Can’t be Wrong

Usually, an engine assembly line has 8 to 10 cells and consists of 4 to 8 cameras for every cell. That’s a system with as many as 80 cameras – which is very expensive and very complex! So why are smart cameras a good option for such applications?

smart cameras on an assembly line

Multiple smart cameras is a smart way to add more inspection stations on an assembly line

As each smart camera is its own entity, it is totally separate from the other cameras. In case, after a system runs for a few months, a customer wants to include a new component or add more vision for inspection, the smart camera can be added by simply configuring it for the necessary task and adding to the network. On the other hand, with a traditional PC-based frame grabber system, a systems integrator would have to, at the very least, change the source code to include an additional camera, or even add a new frame grabber.

Generally, it is much more convenient to program one smart camera for one part of an application than it is to take a single large program – which takes the input from two or three cameras – and develop new code and guarantee that everything is synchronized such that all the cameras run simultaneously. A flexible system that is easy to reconfigure when specifications change is required by the automobile assembly plant.

The smart cameras’ modular nature also makes them ideal for complex applications. Often, adding them to an existing network is easier to do than writing a new code. Moreover, when new functionality needs to be added to the system, it is possible to program the camera offline and connect it. Hence, it can be tested live without disturbing ongoing inspections. Thanks to smart cameras, integrators can easily add new functionality to an existing system.

BOA smart cameras

BOA smart cameras (from Teledyne DALSA) can perform part identification and ensure part traceability throughout the product lifecycle.

Developing a system with multiple smart cameras can also help to reduce development costs. As smart cameras are configured rather than programed, vision applications can be developed rapidly – requiring just half a day for a simple application. Teledyne DALSA’s smart camera environments – iNspect or Sherlock – use graphical interfaces where a user, instead of writing lines of code, ‘builds’ an application. Again for an application that requires multiple cameras, configuring half a dozen smart cameras can save a lot of time for engineers instead of writing C++ code for the acquisition and processing of those six cameras. Overall, the ease of programming means it is relatively easy to develop an application; once the application is ready quickly, it can be deployed quickly and the customer can benefit from the return on the investment.

So How to Choose?

Ultimately, you can determine whether multiple smart cameras are a good investment based on the processing and resolution requirements of a particular application. Even when used in complex systems, Smart cameras are well-suited for reading bar-codes, verification, pattern matching, point-to-point measurements, presence/absence, or counting blobs in an image – in other words, any process that can give an expected measurement, or a pass/fail result. As demonstrated by the automobile assembly line application, it’s not hard to see how a smart camera can be assigned a single task in the overall process.

It is to be noted that due to the nature of embedded components, smart cameras will never match the performance of PC-based systems. A smart camera can perform at 60 or more frames/second, but that will slow down if coupled with a low-level function such as bar code reading with high-level processing, for example, OCR or geometric pattern matching.

Lastly, if the application being used needs round robin or parallel processing, a traditional PC-based multi-camera system, with a vision processor or frame grabber, is the ideal choice. Although a smart camera has a CPU, it is not a PC and, therefore is not intended to run complex, multi-threaded code behind those applications.

Future Thinking

A multiple smart camera “system” is suitable for certain types of applications, and the below table can help estimate the requirements of the application. Opting for the smart camera route needs some planning and good communication with the customer. It is advised to ask the customers how they imagine the application and believe how it could evolve in the future. Do they desire to examine more features or integrate more inspection stations on an assembly line? Will they want to keep a record of all the image data, or want the cameras to communicate results? Their answers will help find the best solution that will satisfy them.

Factors to consider Multiple smart cameras Multiple traditional cameras
Resolution higher than 1600 x 1200   X
Lowest cost possible X  
Need pass/fail results X  
Results determine the next step in process   X
Multiple points of inspection on the assembly line X  
Want/need to incorporate additional functionality as product evolves X  
Want to keep all image data   X

 

Teledyne DALSA.

This information has been sourced, reviewed and adapted from materials provided by Teledyne DALSA.

For more information on this source, please visit Teledyne DALSA.

Tell Us What You Think

Do you have a review, update or anything you would like to add to this article?

Leave your feedback
Submit