Image-Guided Subject Distance Extraction for 3D Shape Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing technologies face challenges in accurately extracting the area of a specific subject from distance information, often leading to erroneous results due to range finding errors and background interference, which complicates the estimation of three-dimensional shapes in mixed reality applications.
Innovation Solution
The information processing device employs a combination of image capturing and range finding devices to calculate and extract distance information, using subject extraction units to determine the subject area based on image features, and adjust resolution and projection transforms to enhance accuracy, thereby isolating the subject's distance information from background interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distance information is obtained using range finding devices in mixed reality applications, then three-dimensional shape estimation can be performed, but range finding errors cause erroneous extraction results where background areas are incorrectly identified as subject areas or subject areas are missed
Solution Approach 1:
The patent combines image capturing device data with range finding device data to perform subject extraction. By merging visual information from the image capturing device with distance information from the range finding device, the system can accurately identify subject areas and eliminate erroneous background extractions caused by range finding errors alone
Solution Approach 2:
The image capturing device serves as an intermediary to verify and correct distance information. The extracted subject area from image processing acts as a mediator to filter out erroneous distance measurements from the range finding device, ensuring only valid subject areas are identified
2Device complexity
If only distance information is used to extract subject area, then processing is simplified, but it is difficult to determine the subject area with high accuracy when subject distance is close to background distance
Solution Approach 1:
The patent merges image capturing device processing with range finding device processing to achieve accurate subject extraction. The combination of visual features from image processing and distance information from range finding allows the system to distinguish subjects from background even when their distances are similar, maintaining high accuracy without excessive complexity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An information processing device includes a subject extraction unit for extracting an area of a subject from an image to generate area information of the subject, a distance information calculation unit for calculating distance information corresponding to the image on the basis of information on a distance within the image, and an area extraction unit for using the area information of the subject generated by the subject extraction unit to extract distance information of the area of the subject from the distance information calculated by the distance information calculation unit.