Method, medium and system for reconstructing three-dimensional model through multi-angle image set

Abstract

The invention provides a method, a medium and a system for reconstructing a three-dimensional model through a multi-angle image set, and the method comprises the steps: prompting and obtaining an image set which is continuously shot for a target according to a preset shooting angle and a shooting distance; extracting the positions of the feature point pixels of the target in the images, and calculating the relative coordinates of the feature point pixels in the space to obtain the relative coordinates of all the feature point pixels; and fitting the relative coordinates of all the feature point pixels to obtain a reconstructed three-dimensional model of the target. According to the method, equipment such as special laser, radar and infrared equipment is not needed, a two-dimensional imageof a target is shot through a single two-dimensional image acquisition device by changing angles for multiple times, the space coordinates of the pixels of the feature points are calculated through aspace geometry method, then fitting is carried out, a three-dimensional model of the target is constructed, reliability and convenience are ensured, and the requirement for high precision on a curvedsurface is met.

Description

Technical field [0001] The invention relates to the field of measurement and modeling, and in particular to a method, medium and system for reconstructing a three-dimensional model through a multi-angle image set. Background technique [0002] As the application scenarios based on RGB image surveying and mapping have increased, from local measurement to overall measurement, local measurement of surface details, and evaluation of the optimal distance through several image ratio conversions, there is a high requirement for dimensional accuracy in surveying and mapping human scenes. , The surface curvature changes greatly, and there are many environmental interference factors. [0003] At present, there are three main methods on the market. (1) Scan by a special laser, and reconstruct the curved surface based on the information obtained by laser scanning, and stitch the imaging surface by multiple scanning at the same time or by a single scanning at multiple angles; (2) Through phase ...

AI Technical Summary

Problems solved by technology

These three have their own advantages, but in human body size mapping, the acceptable scanning time within half a minute, and the best shooting distance of 1 meter, the cost of the first special optical device is high, and the accuracy is stable It needs regular manual intervention and adjustment. After fixed installation, the single scanning speed is the fastest, but it is easily interfered by external ambient light and object reflection, and needs to be calibrated irregularly. The second type of 3D information extraction takes a long time, and the accuracy of the surface is difficult. Improvement, the surface information of small objects is seriously lost; the third type relies on proprietary infrared characteristics, within the reliable line-of-sight range of existing infrared sensors, the accuracy of depth information of photographed objects is limited by hardware, and depends on the ambient light and the color of the shooting target The degree is large, and equipment and levels need to be calibrated from time to time

Images