Journals Proceedings

Abstract

An automated endoscopic tool for detection of various image features from the endoscopic videos to speed up clinician process. The detection of diseased or problematic area involved in endoscopic analysis requires very enough time 7 to 8 hours to manually review the video to locate problematic area. In an attempt to automate the process, we have developed a tool which accepts the features of the problematic areas from the physician and accordingly features are detected on the image obtained from the endoscopic video. The functional modules of our tool include: Capturing image frames from a live video or already available video, Reducing redundancy involved in the images obtained such as inter-frame redundancy (Temporal) and Feature detection in images through color image segmentation and other techniques.