Online extraction of lumen region and boundary from endoscopic images using a quad structure

摘要

A new approach for the automatic extraction of the lumen region and its boundary for the gastrointestinal (GI) endoscopic images is presented. At first, a quasi region of interest representing the darker regions of the image is segmented by using a region splitting scheme named progressive thresholding. Then the lumen region is obtained by using a region growing technique called integrated neighbourhood search (INS). A new quadstructure based technique is introduced to enhance the speed of INS significantly. A back projection algorithm is suggested to optimise the search for the pixels belonging to the lumen region and boundary. A boundary-thinning algorithm is also proposed to remove the redundant pixels from the lumen boundary and to generate a connected single pixel width boundary. The proposed approach does not need a priori knowledge about the image characteristics. The efficiency of this method in terms of speed and accuracy is validated by various GI images and the results of the experiments are presented. The main advantage of the proposed technique is its high-speed response that facilitates real time analysis of the endoscopic images.