Monocular Polarization 3D Reconstruction Method Based on Scattering Information Correction

Abstract

The invention relates to a monocular polarization three-dimensional reconstruction method based on scattering information correction, comprising the following steps: (a) obtaining polarization images of several different polarization angles of an object to be reconstructed; (b) calculating the polarization image to be reconstructed according to the several polarization images Reconstructing the diffuse reflection intensity of the object; (c) calculating the first surface normal of the object to be reconstructed according to the diffuse reflection intensity of several objects to be reconstructed; (d) calculating the to-be-reconstructed object according to the light intensity of the object to be reconstructed Reconstructing the second surface normal of the object; (e) using the second surface normal to correct the first surface normal to obtain a third surface normal of the object to be reconstructed; (f) according to the third Surface normals complete the 3D reconstruction. The monocular polarization three-dimensional reconstruction method based on scattering information correction of the present invention uses the diffuse reflection intensity to calculate the polarization degree of the object to be reconstructed, thereby improving the accuracy of the three-dimensional reconstruction result.

Description

technical field [0001] The invention belongs to the field of computational imaging, and in particular relates to a monocular polarization three-dimensional reconstruction method based on scattering information correction. Background technique [0002] As an important technology in the field of computer vision, 3D reconstruction has broad application prospects because it can restore the surface shape of objects. However, there are many deficiencies in traditional 3D reconstruction: (1) low reconstruction accuracy, such as shading restoration method, binocular vision, etc.; (2) complex equipment, such as stereo vision method, structured light method, etc.; (3) expensive, such as laser Scanning, time-of-flight, etc. Therefore, it is of great research significance to realize a 3D reconstruction system with high precision, simple system and low cost. [0003] Due to the introduction of object polarization information, the polarization 3D reconstruction technology can obtain ric...

AI Technical Summary

Problems solved by technology

However, there are many deficiencies in traditional 3D reconstruction: (1) low reconstruction accuracy, such as shading restoration method, binocular vision, etc.; (2) complex equipment, such as stereo vision method, structured light method, etc.; (3) expensive, such as laser scanning, time-of-flight, etc.

[0022] In the first method, the depth information obtained by kinect is used to correct the surface normal of the object in the reconstruction, which makes the system structure complex and increases the system cost. At the same time, image matching is required, which increases the amount of calculation and calculation error; the second method is in When calculating the zenith angle of the normal zenith angle of the surface of the object to be reconstructed, the polarization degree of the diffuse reflection light of the object is theoretically used. In this method, the polarization degree of the reflected light of the object is used, which makes the calculation of the zenith angle of the normal zenith angle of the reconstructed object surface inaccurate , resulting in distortion of the reconstruction result, and the final reconstruction result of this method only restores the surface shape of the object and the relative depth information of the object surface, and cannot obtain the real depth information of the object

Images