For direct visual examination of inaccessible body cavities, semirigid or flexible endoscopes (fiberscopes) are preferably used. Typical examples in a clinical environment are bronchoscopy, gastroscopy and neurosurgery and, in the technical area, inspection and maintenance of turbines and pipes.

Image Optimization for Fiber Endoscopes

The Challenge

Due to their flexible construction and miniaturization, the fiber-optic image conductors employed exhibit a honeycomb interference pattern (a) and poor brightness and color reproduction. These negative effects lower the image quality and, in conventional systems, are usually reduced through defocusing and adaptation of brightness and contrast.

Our Solution: DeFiberizer

To suppress the honeycomb pattern, all fiber centers in the fiber-optic image are first registered in a calibration step. Direct access to the raw data from the color filter array in the camera enables optimized calculation of individual correction factors for color distribution within the various cells. Together with the information on local adjacency relationships, this forms the basis for subsequent interpolation. With the algorithms described, global shadowing effects and color artifacts can be visibly reduced.

The software automatically adapts to most combinations of commonly used endoscopes and suitable digital camera modules. Operation from a PC platform additionally allows for acceleration via powerful graphics hardware.


  • Winter, C.; Zerfaß, T.; Elter, M.; Rupp, S.; Wittenberg, T.: Physically motivated enhancement of color images for fiber endoscopy. In: Medical Image Computing and Computer-Assisted Intervention (MICCAI), Springer Verlag, Heidelberg, 2007, pp. 360-367.

More Information