Signal Processing Device for Face Detection with Resolution-Based Size Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing face detection systems require generating multiple images with varying resolutions to accommodate different face sizes, leading to increased computational load and power consumption, especially in low-performance devices like smartphones.
Innovation Solution
A signal processing device that includes a storage unit for size information of a specific subject and a determination processing unit that acquires a second image with lower resolution than the first image, using size and distance information to estimate the subject's size and shape, thereby reducing unnecessary computational load and power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images with different resolutions are generated to accommodate various face sizes, then face detection accuracy is improved, but the amount of calculation and power consumption increase
Solution Approach 1:
The patent divides the image processing into two stages: first, a low-resolution distance image is processed to detect subject presence and estimate size; second, only if a subject is detected, a high-resolution luminance image is processed for detailed face recognition. This segmentation avoids processing high-resolution images unnecessarily, reducing power consumption while maintaining detection accuracy.
Solution Approach 2:
The patent performs preliminary detection using a low-resolution distance image before committing to resource-intensive high-resolution processing. By预先 detecting subject presence and estimating size from the distance image, the system determines whether further face recognition is necessary, preventing wasteful processing when no subject is present.
2Measurement precision
If multiple images with different resolutions are generated to accommodate various face sizes, then face detection accuracy is improved, but the amount of calculation increases
Solution Approach 1:
The patent segments the calculation workload by performing simple subject detection and size estimation on a low-resolution distance image first. Only when a subject is detected does the system proceed to the more complex high-resolution face recognition processing, thereby reducing the total amount of calculation required.
Solution Approach 2:
The patent introduces a low-resolution distance image as an intermediary between the captured scene and the final face recognition. This intermediary provides sufficient information for subject detection and size estimation without requiring the full computational resources needed for high-resolution processing, thus reducing overall calculation complexity.
3Measurement precision
If high-resolution images are processed for all subjects, then identification accuracy is improved, but power consumption and processing time increase
Solution Approach 1:
The patent performs preliminary subject detection and size estimation using a low-resolution distance image before initiating time-consuming high-resolution face recognition. This preliminary action filters out cases where no subject is present or where the subject is too small to be recognizable, reducing unnecessary processing time while maintaining identification accuracy for valid subjects.
Data Source
AI summary
A signal processing device according to the present technology includes a storage unit that stores size information of a specific subject and a determination processing unit that acquires a second image as a distance image having a resolution lower than that of a first image used in identification processing of determining whether or not a subject is the specific subject. The determination processing unit performs determination of whether or not the subject in the second image is a processing target subject in the identification processing using the first image on the basis of size information of the subject in the second image and the size information of the specific subject.