Non-Parametric Approach of Video Capsule Endoscope Localization Using Suboptimal Method of Position Bounded CWCL

For proper diagnosis, location of the wireless video capsule endoscope is required to be known by the physicians. In this paper, we propose an algorithm of localizing the capsule using path loss-based calibrated weighted centroid localization (CWCL). The main challenge in path loss-based localization is the highly randomness of measured path loss due to shadow fading and multi-path propagation effects of human body channel. To address the randomness in the measured path loss, we propose two methods of path loss estimation using Gaussian weighted average filter and the multiple input multiple output diversity scheme. Then, we calculate the weight of the sensor receiver position using the estimated path loss. Finally, the position of the capsule is estimated using position-bounded CWCL. We propose a realistic suboptimal method of estimating the calibration coefficient and also compute the optimal value of coefficient to set the benchmark. Additionally, we propose two boundary conditions on the estimated positions to improve the localization accuracy. We simulate our proposed algorithms using MATLAB to validate the accuracy and observe significant improvements without any prior knowledge of channel parameters. The proposed algorithms improve the accuracy up to 5.14-mm root mean square error and outperform the existing literature.