Robot Image Processing for Central Face and Peripheral Change Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing systems for robots face challenges in efficiently detecting faces in central areas while also monitoring changes in surrounding areas, as face detection processing is computationally intensive and requires significant resources, limiting the ability to detect and respond to changes in the environment effectively.
Innovation Solution
An image processing device and method that separates processing into two distinct tasks: first, face detection processing focused on a central area with higher computational load, and second, change detection processing in surrounding areas with a lower processing load per unit pixel, allowing for efficient detection and adjustment of the camera direction to center changes within the image frame.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If face detection processing is executed on the entire captured image, then detection accuracy is improved, but processing load increases significantly
Solution Approach 1:
The captured image is divided into a central area and a surrounding area. Face detection processing (first detection processing) is executed only on the central area where faces are likely to appear, while change detection processing (second detection processing) is executed on the surrounding area. This segmentation allows high-accuracy face detection without processing the entire image, thus reducing overall processing load while maintaining detection accuracy for faces.
2Adaptability or versatility
If change detection is executed on the entire captured image, then environmental monitoring capability is improved, but processing load increases
Solution Approach 1:
Different processing approaches are applied to different regions of the image based on their specific requirements. The central area receives intensive face detection processing with higher computational load, while the surrounding area receives lighter change detection processing. This local quality approach ensures that environmental monitoring is effective in the surrounding areas without unnecessarily increasing processing load across the entire image.
3Measurement precision
If high computational load processing is applied to all areas, then detection precision is improved, but energy consumption increases
Solution Approach 1:
The image processing is segmented into two distinct processing tasks applied to different regions: face detection in the central area and change detection in the surrounding area. This segmentation enables the system to maintain high detection precision where needed (central area for faces) while reducing energy consumption by applying lighter processing to the surrounding areas, thus optimizing the trade-off between detection precision and energy consumption.
Data Source
AI summary
An image processing device includes an acquirer, a face detector, and a change detector. The acquirer acquires images captured in succession by an image capturer. The face detector executes a first detection processing (face detection processing) that detects a certain target (face of a person) from a central area of the acquired image. The change detector executes second detection processing to detect a change over time in a surrounding area other than the central area in the acquired image. The second detection processing has a smaller processing load required per unit pixel in the image than that of the first detection processing.