Conclusion

In the field of medical imaging, endoscopy is a very important modality for disease diagnosis, patient follow-up and minimally invasive surgery. However, this imaging modality is still under-exploited. One of the obstacles to greater and easier use of this image modality lies in the enlarged field of view of endoscopes. This contribution provides a state-of-the-art review of the various image processing steps required for this enlargement, namely the correction of inherent instrument defects by pre-processing techniques (distortions, honeycomb fiber patterns and inhomogeneous perception of scene illumination due to vignetting), image registration to establish the geometric link between images and the (non-redundant) placement of iconic data in the common panoramic image frame (mosaic), while minimizing textural and color discontinuities in data acquired from different viewpoints....