Title :
Transform based bleeding detection technique for endoscopic images
Author :
Mathew, Mekha ; Gopi, Varun P.
Author_Institution :
Dept. of Electron. & Commun. Eng, Gov. Eng. Coll. Wayanad, Mananthavady, India
Abstract :
Wireless Capsule Endoscopy (WCE) has become a great device in Gastrointestinal (GI) tract diagnosis, which can examine the entire GI tract, especially the small intestine without invasiveness and sedation. Bleeding in the digestive tract is a symptom of a disease rather than a disease itself. Hence the detection of bleeding is important in diagnosing many diseases. In this paper we proposes a novel method for distinguishing bleeding regions from normal regions based on Contourlet transform and Local Binary Pattern (LBP). Experiments show that this method provides a high accuracy rate of 96.38% in CIE XYZ colour space for k-Nearest Neighbour (k-NN) classifier.
Keywords :
biological organs; biomedical optical imaging; biomedical telemetry; body sensor networks; colour graphics; diseases; endoscopes; feature extraction; haemodynamics; image classification; medical image processing; wavelet transforms; CIE XYZ colour space; GI tract diagnosis; LBP; WCE; bleeding region classification; contourlet transform; digestive tract bleeding detection; disease diagnosis; disease symptom; endoscopic image; gastrointestinal tract diagnosis; k-NN classifier accuracy rate; k-nearest neighbour classifier; local binary pattern; noninvasivene method; normal region classification; sedation; small intestine; transform based bleeding detection; wireless capsule endoscopy; Accuracy; Endoscopes; Feature extraction; Hemorrhaging; Histograms; Image color analysis; Transforms; Contourlet transform; Local binary pattern; Wireless Capsule Endoscopy; k-NN classifier;
Conference_Titel :
Electronics and Communication Systems (ICECS), 2015 2nd International Conference on
Conference_Location :
Coimbatore
Print_ISBN :
978-1-4799-7224-1
DOI :
10.1109/ECS.2015.7124882
