Multi-sensor fusion for real-time object tracking

Accurate orientation and position estimation are critical elements in optimizing real-time object tracking performance when leveraging smartphone sensors such as accelerometers and gyroscopes. The primary challenges encountered in smartphone-based object tracking are attributed to the GPS signal, canyon effect, and orientation errors, accumulation error in sensor. To address these limitations, a novel approach is proposed wherein a smartphone application is developed based on IMU Multi -sensor fusion using Kalman filter and Rotation vector. The proposed approach integrates Kalman filtering to fuse sensor data and leverages the rotation vector for precise orientation estimation. Additionally, geohash filtering is employed to efficiently proficiency in quantifying intricate spatial interdependencies and display track paths on maps within the application. A detailed mathematical analysis and thorough comparison with existing algorithms in the field proves the dexterity of the proposed object tracking scheme. The comprehensive evaluation showcases the algorithm's capability and advancement compared to state-of-the-art approaches.