Indoor real-time three-dimensional reconstruction method based on depth camera

The invention discloses an indoor real-time three-dimensional reconstruction method based on a depth camera. The indoor real-time three-dimensional reconstruction method comprises the following steps: step 1, acquiring a depth image and a color image of an indoor scene; 2, complementing holes of the depth map; 3, converting the depth map into a point cloud, and carrying out downsampling on the point cloud; 4, detecting and matching feature points, and calculating initial value pose parameter estimation; step 5, registering the point cloud by using an ICP point-to-plane algorithm, and then realizing pose tracking by using a feature point method in combination with a direct method; step 6, using a bag-of-words method to complete loopback detection, and optimizing a detected closed loop; and 7, constructing a mesh map according to the distance sign function in the TSDF stage, reconstructing the surface of the object by using a moving cube MC algorithm, and completing texture mapping through weighted fusion to ob