AR Object Detection via Point Cloud Complement
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality (AR) techniques face challenges in object detection and recognition due to limitations in the operation principle of depth sensors, leading to difficulties in achieving high-precision real-time processing and accurate 3D state understanding of surrounding objects.
Innovation Solution
A method and apparatus for object detection that involves obtaining a first point cloud feature from point cloud data, performing point cloud complement to fill in missing data, and determining objects by extracting appearance and geometric features, calculating relationship weights, and refining features through a series of neural network operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If depth sensors are used for object detection in AR systems, then real-time processing capability is achieved, but measurement precision and accuracy of 3D state understanding deteriorate
Solution Approach 1:
The patent combines multiple point cloud data from different sources and viewpoints to create a more complete and accurate 3D representation of objects. By merging complementary information from various sensors and processing stages, the system achieves both real-time performance and high measurement precision simultaneously
Solution Approach 2:
The patent transitions from 2D image data to 3D point cloud representation, adding spatial dimensionality to enhance object understanding. This dimensional transformation enables more accurate 3D state comprehension while maintaining real-time processing capabilities through efficient point cloud algorithms
2Measurement precision
If point cloud complement is performed to improve data quality, then measurement precision improves, but device complexity and processing time increase
Solution Approach 1:
The patent performs point cloud complement operations in advance during the data processing pipeline, preparing enhanced point cloud data before object detection. This preliminary enhancement improves subsequent detection accuracy without adding complexity to the core detection algorithm
Solution Approach 2:
The patent introduces intermediate processing steps that act as mediators between raw sensor data and final object detection. These intermediate point cloud complement operations refine the data gradually through multiple processing stages, managing complexity by breaking down the enhancement task into manageable steps
Data Source
AI summary
A method with object detection includes obtaining a first point cloud feature based on point cloud data of an image and determining at least one object in the image based on the first point cloud feature.