Basler's new proprietary feature set: PGI
Several of our newest camera models come with our powerful in-camera image optimization technology already built in: The proprietary PGI feature set enhances your images at the full speed of your camera. PGI is comprised of a unique combination of 5x5 debayering/4x4 debayering, color-anti-aliasing, denoising and improved sharpness.
With PGI, your camera will produce better images than ever, without putting additional load on your processor. The PGI feature set is included in all boost, dart and pulse color cameras as well as ace U, ace L, and ace 2 Pro color cameras.
The following cameras possess the PGI feature set:
Details on PGI
The term debayering generally describes an algorithm that calculates the color image from the image sensor data. However, as the image sensor does not provide color values for each color in each individual pixel, the algorithm must interpolate to determine that information. Cameras typically offer 2x2 debayering, whereby the two closest pixels are used to calculate the actual color of the image. This can lead in some cases to uneven edges between colors and other artifacts. The 5x5 debayering used in the PGI feature set delivers cleaner transitions between colors and eliminates artifacts altogether.
At the same time, the optimized debayering also assists with noise suppression.
*For technical reasons, a 4x4 debayering is used for some sensors / models.
Due to the limits of camera resolution, the use of the debayering algorithms can easily lead to color distortion in the captured images. In practical terms, even colorless structures suddenly appear to have color. The aforementioned resolution limits allow color cameras using debayering to capture only specific colors, which in turn produces false interpolation values. PGI uses an expanded informational range during debayering while also simultaneously correcting the incorrect colors.
Noise is a phenomenon that occurs in all cameras, and arises from a variety of causes (photon shot noise, sensor noise). Color cameras must not only deal with gray noise but also with color noise caused and reinforced by the sequencing of multiple calculation steps and interpolation. PGI accounts for and avoids this type of noise formation from the start through careful linking and parallelization of calculation operations. In addition, active noise filtering can be applied to further reduce the noise level, which in turn further enhances the image.
Color cameras often struggle to depict particularly fine or sharp structures. The results are aliasing effects or reduced image sharpness. This can be traced back to the interpolation algorithms, known as debayering for cameras with Bayer pattern. Because its interpolation algorithm is adapted for the image structure, the PGI feature set delivers significantly improved sharpness, with the option to pursue further improvement via a supplemental sharpness factor. These enhancements are particularly helpful for applications requiring strong sharpness, such as applications using color cameras to detect and process letters and numbers (such as ANPR for traffic applications) or other fine or sharp-edged structures (such as barcodes).