CVC-ClinicDB is a database of frames extracted from colonoscopy videos. These frames contain several examples of polyps. In addition to the frames, we provide the ground truth for the polyps.  This ground truth consists of a mask corresponding to the region covered by the polyp in the image, as observed in Fig. 1.

Fig1. Examples of content of CVC-ClinicDB database. First row shows original images whereas second row shows corresponding ground truth.

CVC-ClinicDB database consists of two different types of images:
1) Original images: original/frame_number.tiff
2) Polyp mask: ground truth/frame_number.tiff

The correspondence between the number of frame and the video sequence is shown in Table 1:

Table 1: Correspondence between number of frames and sequences in CVC-ClinicDB.

CVC-ClinicDB is the official database to be used in the training stages of MICCAI 2015 Sub-Challenge on Automatic Polyp Detection Challenge in Colonoscopy Videos .

Download link

You can download CVC-ClinicDB database from the following link: CVC-ClinicDB.rar


Images from folder ‘Original’ are property of Hospital Clinic, Barcelona, Spain

Images from folder ‘Ground Truth’ are propery of Computer Vision Center, Barcelona, Spain


The use of this database is completely restricted for research and educational purposes. The use of this database is forbidden for commercial purposes.

If you use this database for your experiments please include the following reference:

Bernal, J., Sánchez, F. J., Fernández-Esparrach, G., Gil, D., Rodríguez, C., & Vilariño, F. (2015). WM-DOVA maps for accurate polyp highlighting in colonoscopy: Validation vs. saliency maps from physicians. Computerized Medical Imaging and Graphics, 43, 99-111 .