Endoscopic Video Pre-Processing using Histogram Equalization and Canny Edge Detection Technique and Hough Transform Technique for Polyp Detection

Abstract

Paper An effective image processing approach has been shown to process real time endoscopic videos to support practitioners in taking vital decisions regarding cancer patients. The images are extracted from real time endoscopy videos using the available software and fed into Matlab for image processing, the results of the processed videos are returned back to the host software. Image processing techniques are used to recognize and enhance the visualization of the polyps present in gastro intestinal tract which help the practitioners in decision making. The intention of the system is to assist the physician or medical practitioner for better visualization and identifying abnormal structures like polyps and bleeding regions during the endoscopic procedures. This proposed system is experimented on the recorded gastrointestinal dataset which contains ten sequence videos with 7894 total frames.