Vision Campus Article: Excellent Pre-processing of Acquired Images for Flawless Image Processing
The processing steps used in pre-processing are executed on the FPGA processor by frame grabbers, but also by other vision hardware such as cameras, vision sensors or embedded systems. On all these vision devices, the graphical development environment VisualApplets can be used to implement extensive pre-processing as well as complete image processing solutions on the FPGA, thus also enabling application or software engineers to implement a real-time-capable system.
Frame grabbers allow a much wider range of pre-processing. If the image processing system is set up with standard frame grabbers, pre-processing elements such as high-quality debayering, look-up tables (LUT) and mirroring are usually included, depending on the application. With programmable frame grabbers, the individual design options are much more versatile.
All in all, image pre-processing enables acceleration of computationally intensive algorithms. This permits the system to improve application performance and to process image data with very high resolution and bandwidth efficiently, by increasing the data throughput to enhance the accuracy and robustness of image analysis. The complete bandwidth of a high-speed camera can be safely transferred while reducing load on the host PC’s CPU and the subsequent analysis software.
Curious for more?