Real-time detection method and system for three-dimensional objects

Abstract

The embodiment of the present invention discloses a real-time detection method of a three-dimensional object, which includes the following steps: step 100, constructing a three-dimensional object database; and step 200, detection of the three-dimensional object. The embodiments of the present invention can solve the difficulties in database construction, and can realize fast and real-time detection of three-dimensional objects on mobile devices. The embodiment of the invention also discloses a real-time detection system of a three-dimensional object.

Description

technical field [0001] The embodiment of the present invention relates to a computer vision detection method, in particular to a real-time detection method of a three-dimensional object. The embodiment of the present invention also relates to a real-time detection system of a three-dimensional object. Background technique [0002] In daily life, people are often interested in some three-dimensional objects seen in the scene. To enable people to interact or connect with these three-dimensional objects, they first need to detect them to understand what it is, where it is, and what its posture is like. And it is hoped that this detection process is fast and accurate, and can reach real-time on the mobile devices that are used the most in daily life. [0003] At present, the existing technologies in the industry are relatively mature for the detection and recognition of planar objects in the scene, but the detection and recognition solutions for three-dimensional objects have ...

AI Technical Summary

Problems solved by technology

The disadvantages of this scheme are: it does not involve the construction of the database, and there are restrictions on the input data (only 3D models can be used, and real images cannot be used); in addition, the calculation of shape feature extraction and description is relatively large, and the matching results are not accurate enough

In order to improve the success rate and speed of 3D object detection, this solution needs to cover as many angles as possible in the early shooting, which results in a very large data size in the database, which is not conducive to deployment on mobile devices

In addition, this acquisition method is not conducive to large-scale expansion of 3D objects of interest

[0010] To sum up, the current technical solutions for 3D object detection are not mature, and have many limitations in database construction and detection methods, which cannot meet the needs well.

Images