Pattern-Adjustable Projector for ROI-Based Depth Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face recognition technologies in mobile devices face bottlenecks in performance improvement, leading to security issues and inefficiencies, particularly in depth detection and 3D face recognition.
Innovation Solution
A method and apparatus for region-of-interest (ROI)-based depth detection using a pattern-adjustable projector, which captures images, determines ROIs, selects projection regions, projects patterns, and generates depth maps, enhancing performance without introducing side effects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional face recognition algorithms are used in mobile devices, then basic face recognition functionality is achieved, but performance improvement is bottlenecked and security issues arise
Solution Approach 1:
The patent divides the projection field into multiple predetermined projection regions and identifies a region of interest (ROI) containing the target object. The system then selects only the projection region corresponding to the ROI for depth detection, segmenting the detection process to focus computational resources on the most relevant area, thereby improving both accuracy and reliability.
Solution Approach 2:
The patent applies different processing qualities to different regions: full-resolution depth detection is applied only to the ROI region where high precision is needed for security-critical face recognition, while other regions use standard processing. This local quality approach ensures high reliability where needed without unnecessary computational overhead.
2Loss of energy
If the projector projects patterns across the entire field, then complete area coverage is achieved, but power consumption increases and signal quality decreases
Solution Approach 1:
The patent extracts only the necessary projection region corresponding to the ROI from the complete projection field. By taking out only the relevant portion for depth detection, the system reduces the projection area, thereby reducing power consumption while maintaining sufficient depth detection precision for the target object.
Solution Approach 2:
The patent applies partial action by projecting patterns only in the selected projection region rather than the entire field. This partial projection provides sufficient depth information for the ROI while avoiding the excessive power consumption and signal degradation that would result from full-field projection.
3Reliability
If face recognition algorithms are improved to enhance security, then recognition accuracy increases, but the system introduces side effects and complexity
Solution Approach 1:
The patent transitions from traditional 2D face recognition to 3D depth-based recognition by projecting structured light patterns and capturing depth information. This dimensional change adds a new depth dimension to the recognition process, significantly improving security reliability without requiring complex algorithmic changes to existing 2D recognition systems.
Solution Approach 2:
The patent introduces depth map information as an intermediary layer between the captured image and the face recognition algorithm. This intermediary depth information enhances security performance by providing three-dimensional structural data, while the existing face recognition algorithms can remain relatively simple as they now work with enriched input data rather than requiring complex algorithmic modifications.
Data Source
AI summary
A method for performing region-of-interest (ROI)-based depth detection with aid of a pattern-adjustable projector and associated apparatus are provided. The method includes: utilizing a first camera to capture a first image, wherein the first image includes image contents indicating one or more objects; utilizing an image processing circuit to determine a ROI of the first image according to the image contents of the first image; utilizing the image processing circuit to perform projection region selection to determine a selected projection region corresponding to the ROI among multiple predetermined projection regions, wherein the selected projection region is selected from the multiple predetermined projection regions according to the ROI; utilizing the pattern-adjustable projector to project a predetermined pattern according to the selected projection region, for performing depth detection; utilizing a second camera to capture a second image; and performing the depth detection according to the second image to generate a depth map.