Human-vehicle interaction test virtual scene detection method and system, storage medium and equipment

By acquiring real-time preview images and depth information background images, positioning three-dimensional virtual images and target images, analyzing differences and using anchoring algorithms to detect posture consistency, the problem of difficulty in distinguishing between real and virtual images in virtual image testing in the prior art is solved, and accurate detection and distinction of virtual scenes is achieved.

CN119992591APending Publication Date: 2025-05-13CHERY AUTOMOBILE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202510064435.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-01-15
Publication Date
2025-05-13

AI Technical Summary

Technical Problem

The lack of technology for identifying virtual images and real people in the prior art makes it difficult to distinguish between real and virtual images in virtual image testing, and there is a lack of effective methods to distinguish between real and virtual images.

Method used

By acquiring real-time preview images and background images with depth information, positioning the three-dimensional virtual images and target images, analyzing the differences between the two, using an anchoring algorithm for analysis, detecting the pose of the target image and determining its consistency with the focus posture, and performing focus shooting.

Benefits of technology

It realizes comprehensive inspection of virtual scenes of human-vehicle interaction testing, which can accurately distinguish between reality and virtual images, and improves the accuracy and comprehensiveness of the test.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119992591A_ABST
    Figure CN119992591A_ABST
Patent Text Reader

Abstract

The invention provides a human-vehicle interaction test virtual scene detection method and system, a storage medium and equipment. The method comprises the following steps: acquiring a real-time preview image and a background image with depth information; positioning a three-dimensional virtual image in the background image with the depth information; positioning a target image in the real-time preview image; analyzing the difference between the three-dimensional virtual image in the background image and the target image in the real-time preview image to obtain an analysis result; and according to an analysis result, carrying out over-detection on the human-vehicle interaction test virtual scene. According to the method, the virtual scene is comprehensively detected from multiple dimensions by combining different types of data (such as images, sounds, sensor data, user behaviors and the like), the advantages of various data sources are fully played, and more accurate and comprehensive virtual scene perception is realized; security risks of an intelligent automobile in different scenes and exposure frequency of a test scene in the real world are considered to construct an importance function to detect scene data.
Need to check novelty before this filing date? Find Prior Art