A new Gray-Level Picture Thresholding method for colonoscopic navigation

In this paper, we propose a vision based navigation system to guide an endoscope inside human colon. The target to pursue is a dark spot on colonoscopic images, called “pattern”. A novel methodology for “pattern” extraction and tracking was designed. Surgeons observations leads to the basic idea of this technique. Information about target position is then continues and makes possible prediction of the “pattern” position. A set of endoscopic images is tested to demonstrate the effectiveness of the vision technique. An experiment tool to simulate the endoscope navigation was employed to achieve real time performance. An interpretation of the results and the possible amelioration is presented.