Three-dimensional reconstruction method based on optimized IMU tight coupling dense direct RGBD

Abstract

The invention discloses a three-dimensional reconstruction method based on optimized IMU tight coupling dense direct RGBD, and the method comprises the steps: (1) collecting optimized RGB data and optimized depth data from a local model reconstructed at the last moment, and obtaining original RGB data, original depth data and IMU data; (2) constructing a nonlinear cost function according to the inertial navigation error, the dense luminosity error and the ICP error, forming a nonlinear optimization problem by taking minimization of the nonlinear cost function as a target, and solving the nonlinear optimization problem by using a Gauss-Newton method so as to estimate an RGBD sensor optimization pose; (3) carrying out real-time three-dimensional reconstruction by utilizing the original RGB data and the original depth data at the current moment under the optimized pose of the RGBD sensor, and deforming the three-dimensional reconstruction model by using the nodes under the condition of finding the loopback constraint so as to ensure the global consistency. According to the method, the accuracy and efficiency of three-dimensional reconstruction are improved.

Description

technical field [0001] The invention relates to the field of three-dimensional reconstruction, in particular to a three-dimensional reconstruction method based on optimized IMU tightly coupled dense direct RGBD. Background technique [0002] The inertial measurement unit is a device that measures the three-axis angular velocity and specific force of an object. Generally, an IMU includes a three-axis accelerometer and a three-axis gyroscope. The accelerometer detects the specific force signal of the object in the carrier coordinate system, and The gyroscope detects the angular velocity signal of the carrier relative to the inertial coordinate system. [0003] Real-time 3D reconstruction based on the visually dense direct method using depth-detailed RGBD data is a relatively mature application, such as an incremental multi-view 3D reconstruction method published by the application publication number CN110211223A, and an application publication number published by CN111009007A ...

AI Technical Summary

Problems solved by technology

[0004] However, the simple visual dense direct method has a robustness defect in the pose tracking estimation of the visual sensor: if the initial value of the nonlinear optimization of the pose estimation deviates too far from the real value, the estimation result of the nonlinear optimization will become Unreliable, even non-linear optimization results fail to converge

When the visual sensor moves too fast or the visual sensor passes through a section with no ups and downs in the geometric structure and lack of texture, the pure visual dense direct method will cause the above-mentioned situation in which the initial value deviates too far from the true value in the nonlinear optimization

Images