Research On Indoor Positioning Technology Kinect And Inertial Navigation System Integrated

The robot autonomous navigation and positioning technology is an important prerequisite to achieve indoor robot path planning, obstacle avoidance and reach the target position. Since the GPS signal indoor environment is weak, cannot use GPS technology application in indoor mobile robot navigation. Thus according to indoor robot sensors including vision sensors, inertial navigation system measuring unit, odometer, etc., research of autonomy, stable and rapid positioning navigation system is of great significance for the autonomous navigation robot.This paper mainly studies and analyzes the following three aspects of content:1. This paper introduces the working principle of strapdown inertial navigation and digital iteration algorithm, derives quaternion Picard algorithm and rotation vector method, and describes the inertial navigation error model, and SINS position, velocity and attitude.2. We expounds the principle of the navigation and positioning algorithm based on visual sensor device, through RGB image and depth information collected, to SIFT the adjacent two RGB image matching, and combined with the depth image for all feature points on the three-dimensional coordinates of the device coordinate system. Using the absolute directional algorithm to get rotation matrix and translation vector, through laboratory experiments show, compared to conventional ICP algorithm, more accurate indoor robot navigation and positioning can be obtained.3. The paper analyzes the inertial navigation system and monocular vision error, establishes the SINS-based state equation and observation equation, and gives the Kinect and IMU based on a combination of indoor robot navigation system with Kalman filtering scheme. Through laboratory experiments to prove: Kalman filter integrated navigation system suppresses the error of positioning system, effectively improves navigation accuracy compared to Kinect vision based positioning system, and ensures the robot navigation system navigation positioning accuracy and stability.