3D model rapid construction method based on image screening

Abstract

The embodiment of the invention provides an image matching method in a three-dimensional model construction process. The method comprises the steps of 1, determining a to-be-matched source image n; step 2, screening images n-1... N-m and n + 1... N + m adjacent to the source image n; 3, performing matching calculation on the screened 2m + 1 images; and step 4, repeating the steps 2-3 on the rest to-be-matched source images, and finally completing matching of all the images. For the first time, aiming at a scheme of surrounding collection of a target object with a limited volume, a mode of screening adjacent pictures is proposed to reduce the calculation difficulty and time of a matching algorithm, and the synthesis speed and precision are considered.

Description

Technical field [0001] The invention relates to the technical field of shape measurement, in particular to the technical field of 3D shape measurement. Background technique [0002] When performing 3D measurement, using 3D measurement data for processing and manufacturing, or using 3D data for display and identification, first of all, a relatively accurate 3D model should be established for the target. At present, commonly used methods include using machine vision to collect pictures of objects from different angles, and match these pictures to form a 3D model. These 3D models can be regarded as the digitization of real things, and using these data can process and manufacture matching objects that match the target. For example, 3D data of human feet can be collected to make more suitable shoes. In addition, these data can also be used for identity verification. For example, the 3D model of the human iris can be used as standard identity data after being synthesized, and the 3D...

AI Technical Summary

Problems solved by technology

However, in practical applications, it is found that unless there is an accurate angle measurement device, the user is not sensitive to the angle, and it is difficult to determine the angle accurately; the size of the target object is difficult to accurately determine, especially in some applications where the target object needs to be replaced frequently. A lot of extra work is required, and professional equipment is required to accurately measure irregular targets

Measurement errors lead to camera position setting errors, which will affect the speed and effect of acquisition and synthesis; the accuracy and speed need to be further improved

Images