To continue our discussion on the ways camera technology can make a better picture, we offer a recent story from a customer regarding the advantages of in-camera image processing. This particular example relates to Flat Field Correction (FFC) functionality.
This example is from a customer who is a sophisticated medical diagnostic system provider. For all previous generations of the system, they developed their own cameras and image processing software. This included advanced flat field correction which was performed in the frame grabber. For the latest upgrade of their system, they evaluated 3 cameras from different suppliers based on the CMOSIS CMV4000 sensor, the Adimec Quartz Q-4A180 included. The engineer completing the testing assumed the image uniformity would be the same because the cameras had the same image sensor. The engineer’s observations were that the 2 cameras combined with their own off-camera image processing performed worse than Adimec’s camera with internal image processing. He did not trust the results and sought our explanation of how this could be. The reality is no matter how good the FFC is on the frame grabber, it will not be as good as what can be done on the camera, especially by an experienced camera manufacturer.
Higher accuracy is achieved by having FFC on the camera for a couple of reasons. One is that a higher bit depth (10 or even 12 bits) is available in the camera for processing before the images are sent over the interface link. This is the benefit that all camera manufacturers could take advantage of. The other reason for higher accuracy is influenced by the experience and knowledge of the camera manufacturer. Calibrations made closer to the source of the artifacts can have greater effectiveness, assuming the designer has a thorough understanding of the potential issues. A good camera provider has years of experience on how to drive the image sensor chip and get an optimized image out. They can also include automatic adjustments through direct loops that correct for temperature changes, lighting conditions, lens defects or other specific system conditions.
In- camera processing has the additional benefit of reduced costs achieved in several ways:
- Fewer software resources to create own image processing or workarounds for inefficient camera performance.
- Reduced load on the PC-frame grabber processing can mean less expensive overall system costs.
- Reduced data rate can allow for cost optimization with the system connectivity infrastructure. For example if images are processed in the camera a 2-tap Camera Link interface could be sufficient to transport the results rather than a 10-tap configuration.
Our goal is for your software resources to be spent on using the image for your measurement rather than fixing the image. We welcome your comments with your thoughts on this.
Click the image below to find out more about other in-camera image processing and how to compare cameras: