Agisoft Metashape Professional 2.0.4.17162 free downloads

One way to overcome this obstacle is to use synthetic videos with their rendered depth ground truths. However, the access to real intra-operative bronchoscopic videos is sometimes limited, and it is difficult to obtain the ground-truth data. For further development and evaluation of these methods (especially learning-based methods due to their well-proven outstanding performance for monocular depth estimation), a large amount of bronchoscopic videos and a known ground truth of depth maps corresponding to the bronchoscopic video images are required.

Some examples of such depth estimation approaches include conventional “Structure from Motion” (SfM) method, and learning-based depth estimation methods utilizing different networks architectures. There, several 2D–3D registration approaches have been proposed, among which the approaches trying to recover the geometrical structure of the airway based on depth estimation from monocular bronchoscopic video images prove to be more robust to illumination variations. Motivated by its advantages of low cost, less impact from tissue deformations, and no requirement for additional equipment setup, it has been the focus of research in bronchoscopic navigation in recent years.

Vision-based bronchoscopic navigation system, as an alternative to the electromagnetic navigation system, helps to localize the bronchoscope during the endobronchial inspection and diagnostic procedures by applying video-CT registration techniques.

Agisoft Metashape Professional 2.0.4.17162 free downloads